Hacker News new | past | comments | ask | show | jobs | submit login
I regret my website redesign (mtlynch.io)
2011 points by mtlynch 17 days ago | hide | past | favorite | 749 comments



This thread has over 600 comments. If you want to read them all, you need to click "X more comments" at the bottom of each page, or like this:

https://news.ycombinator.com/item?id=32179563&p=2

https://news.ycombinator.com/item?id=32179563&p=3

I just mention this because for a thread this large, it is unusually good. (And yes, we're still going to solve this paging business and then there won't be any more of these annoying notices.)


You know how, during a big programming project, you have to keep the devs from going "this codebase has a lot of technical debt, let's rebuild the whole thing using Rust and Kubernetes and Deno and move the hosting to Azure and switch databases and use microservices and..."?

Designers are exactly the same way. Just as technical fiddling is fun and interesting, making new designs from scratch is just as fun and interesting. And, just as fixing bugs is tedious and boring, tweaking designs is tedious and boring.

I've been on both sides for a lot of years, and I have to keep a sharp eye on myself to keep from spinning my wheels on distractions.

Even the Pope had to keep Michelangelo focused on the Sistine Chapel and not wander off to work on his tomb.


I was ready to fire the agency the moment the author mentioned they were going off script and building whatever they wanted. This was a huge, blinking red flag for me. Why would you keep paying someone when they're doing work you didn't ask for, doing work that doesn't align with your goals, and ignoring what you say?


That seems to be the point where sunk cost fallacy strikes hard in the post to me. They had paid $X for "80% of the first milestone of work" and that sunk cost locks them in to everything else that happens. They seemed too fearful from that point onward that if they fired the agency and brought in a new agency or freelancer that they would start from square 1 and do all of the previous work (and spend) over again.

The sunk cost fallacy suggests that sometimes is better for you if you should just accept existing losses, accept you've already sunk those costs and won't get them back, and move on. I don't know what their contract looked like with the agency, but an 80% of a logo design sounds like a perfect deliverable that you can safely fire the existing team and take the 80% deliverable to a new designer, not start from scratch, and ask them to do a final polish step. I would have cut losses there, but of course it is much easier to armchair quarterback from hindsight and different perspectives than if you are in the middle of it fighting that gut feeling that you've already invested so much and can't "afford" to cut losses.


This is why I still like Agile and Scrum, as much as other devs might hate it. “Yeah I want REAL deliverables after the first or second week”.

Keeps me honest. And will keep me from working with architecture astronauts don’t really deliver anything but hot air and build ultra-extensible structures that are actually impossible to extend beyond the fantasy world of their maker. Or the equivalent for designers.


Agile and scrum are in practice such nebulous terms that while I don’t doubt that you have had good experience with the particular version of these concepts that you yourself use, the agency could well consider itself to be using the same concepts - yet here we are.

The older I get, the more I come to understand that the problem with projects is people, not methodology. 30% of IT projects still fail, despite “agile” now being widespread.

The first value of the agile manifesto is “Individuals and interactions over processes and tools”. Unfortunately, most implementations throw this one out first.


I agree that there's a lot of wiggle room, but there's still some ground rules, like being iterative, not having to plan for everything beforehand, deliver smaller things little by little, not being a perfectionist and turning the review process into a living hell.

But you are 100% right about the problem being people. The problems I see with agile are always because people try to find escape hatches in the process. Designers want to prepare every single detail before developers start working and hate iterating because fear of the "original vision" being a mistake, product owners don't want partial stuff in the beta channel so everything has to be in its final state, product owners don't know or don't want to reduce scope so developers end up with impossible sprints, developers want a perfect definition finished before starting work, so zero chance to iterate...

I also agree with people throwing out individual/interactions and picking a heavy-handed process instead :(


> 30% of IT projects still fail, despite “agile” now being widespread.

That's a very optimistic statistic, unless "fail" is defined to be something other than "the opposite of success".

IOW, a fail is any project that goes over time, over budget or delivers under scope.

If you're defining success as "well, we delivered eventually, and we delivered 80% of the requirements, and it only went over budget by 20%", then, sure, only 30% of IT projects "fail".


Ok, by that measure I guess over 90% of projects would fail. Which would probably be unfair in most cases, because unlike this project, where the scope as defined by the client was 100% clear from the start, most clients when starting out with a project can't really say exactly what they want, so the scope creep is not only driven by the agency, but also (sometimes mostly) by the client.


> most clients when starting out with a project can't really say exactly what they want, so the scope creep is not only driven by the agency, but also (sometimes mostly) by the client.

Most clients know exactly what their goal is ("To increase market recognition", "to increase conversion rate", etc). The scope creep can only happen in a requirements phase[1]. Once a client has signed off on exactly what will be delivered, and what the payments milestones are there's no wiggle room for the agency to continue experimenting.

If you're skipping the requirements phase because "agile"[1], then the work will be over budget, over time and under the scope.

There's very few projects that actually require the exploratory-driven process that agile uses. If you have one of those projects then farming the management of it out to an agency is exactly the wrong thing to do.

[1] There may be multiple requirements phases. For example a large "need to revamp our website" project can be split into multiple serial projects, each with their own requirements phase.

[2] I.e. We'll figure out what is acceptable to the client with constant feedback.


I think the goal is not to eliminate "failure", but reduce it. Failure will always exist but it always requires interpretation. Thus the credos of "failure is learning experience", "don't be afraid to fail" etc.

It's a very abstract term. To put in simpler terms you will always have some outcome mismatch and generally your goal should be to reduce it. But if you do not have outcome mismatch, then you should push the line so that you can look ahead and understand future areas for improvement. Or you can just stay content if "market" is not chasing you.


The problem as I see it is that projects often start without a vision, or even clear goals; even if there is a vision, it ends up getting lost and forgotten in a sea of feature requests.

Not only should we be asking if a project has achieved its original goals once it’s finished, we should be continually assessing progress against those goals throughout the life of the project - yet few projects I’ve worked on actually do this, even when I’ve explicitly requested it.


I've had someone tell me, with a straight face and serious voice, to "stick to the agile process from now on"


only 30%? I'd have guessed that most IT projects at least fail to achieve goals if they don't crater completely. the bigger the projects, the more they resemble interplanetary collisions.


Well, that’s according to the PMI. Finding the link will simply lower my level of happiness for the day, but you should be able to Google it (sorry).

The thing is that as far as I can tell, “success” is measured post-hoc, so all sorts of shenanigans might have taken place in order to determine that a project was “successful”. Having worked with my fair share of corporate project leaders, I’m quite certain that the actual number of failed projects would be much higher if measured by the original criteria used to justify the project in the first place.


Helmuth van Moltke (1880): "No plan of operations reaches with any certainty beyond the first encounter with the enemy's main force." It applies to software projects too, especially large ones. The enemy is the real world (or our lack of understanding of the real world.)

If we look at the original requirements every single project is a failure. If we look at the strategic goals and keep track of all trade offs maybe 30% is a fair guesstimate.


Scrum is really for low-trust environments where management fundamentally don't trust the developers to deliver - whether because the developers aren't experienced, not very good at self-management, or too easily distracted by shiny things, or because you're an archetypal MBA type who distrusts any underling who might be smarter than you.

If you trust your developers (whether in-house or contractors) to do the job you can pretty much let them manage themselves with whatever kanban-ish system that works for them. If you don't, you have a ready methodology that gives you meetings upon meetings and two week deliverables and every little thing broken down to the smallest unit under the guise of "empowerment".


While this is definitely true (I've worked in both kinds of environments), I think the scrum process isn't for low-trust per se, but for managing expectations, especially if you have a larger IT organization. Your direct product owner needs a way to determine how long the development of features can be done, so that they can coordinate and manage expectations with other product owners and other stakeholders (management, customers, etc). I see it as a way to abstract away the individual developers and dev work.

And you probably know that there's a BIG portion of developers who aren't as self-organized or disciplined. You mentioned the shiny things distractions, I think that was a factor in the original post. The other being mismanagement on the agency's side, where they allowed their designers to go out of scope. And I'm not even convinced it's mismanagement, because they turned a $7K project into a $46K project, a 6.5x increase in their revenue off of some small side project.


That's not entirely wrong, but Scrum is also great for high-trust environments where you want to have deliverables from sprint one.

And it is actually more useful when things are flipped, and the team doesn't really trust the client/customer/manager. Which is often what happens more frequently.

But you're totally right that Scrum can get kinda heavy on the meetings side. Which tends to be the rule, not the exception. The only times I had this exception it was when we had a technical person working on the requirements, though. No designer or product owner making nebulous Jira tickets.


I'd bet you a large sum of money this agency used Agile.

Agile is not a guarantee of deliverables you care about. There is a lot of useless stuff you can deliver.


> I'd bet you a large sum of money this agency used Agile.

They probably said the word "agile" a lot and then continued doing something else, that seems to be the usual pattern.

> Agile is not a guarantee of deliverables you care about. There is a lot of useless stuff you can deliver.

Which is why it's the product owner's responsibility to prioritise.


> Which is why it's the product owner's responsibility to prioritise.

The company didn't seem to have issues prioritizing the authors tasks into no mans land. So they seem to have that part down great already. However I find the consistent two week deliveries early on a bit disturbing, in my experience two weeks is exactly one sprint and fixing deliverables to that time window means that you can't really do any agile planing with it, your team just has to accept that this work has to be done and can't give random work that seems to pop up every now and then the attention it deserves. To be really agile and allow your sprint planing to reflect that you shouldn't promise anything in time spans less than two months.


The whole point of agile is that you deliver something every two weeks. Whatever it is you're suggesting, it's not agile.


Delivering "something" and delivering specific items at a fixed rate are different things.


Agile never promised to do the latter (and I've never seen a methodology that succeded at it in a way that was useful in practice).

Well, if you are doing it correctly, you are supposed to make a sprint goal of a deliverable that has value.


> I'd bet you a large sum of money this agency used Agile.

Probably not. Larger relatively successful design / marketing agencies are - based on my experience - poorly run, poorly managed, and poorly led. This gets exasperated by high turn-over (read: good people leave, dead weight stays behind).

Training, etc.? Not a priority. Revenue, it's all about revenue. They're "creatives", not business problem-solver types and presume mess and excessive friction are unavoidable.


Even if they're using agile, it's up to the customer to ask for deliverables that they care about if they want them.

Of course it would be much better if the agency were honest, but if for seven months the deliverables were only "promises", then it's definitely not agile...


Well, waterfall used to not deliver anything at all. Not delivering anything is pretty useless.


> Well, waterfall used to not deliver anything at all. Not delivering anything is pretty useless.

That's pretty much untrue; had this project been a waterfall project then:

a) The devs would not have gotten paid until they hit particular milestones,

b) The deliverables would have been scoped upfront,

c) The agency would have eaten the cost of developing useless things, not the customer.

Because this was obviously an agile/Agile project:

a) The payment milestones were not specified upfront,

b) The devliverables remained unscoped until feedback was needed (which was well into the development phase),

c) The client eats the cost of developers time, whether or not milestones are hit.


Which absolutely makes me think this was a well thought out con designed to prey on those types of fears and maximize the value extracted from the mark.

Those lottery phone scams out of Jamaica which target the elderly in the US are almost the exact same scam as what was described here.

Promise something, get them to send money, don’t deliver, tell them you need more time/more money. Keep repeating until the mark walks away.


Maybe "well thought out con" is a little strong here. I see it as more likely the consulting firm has learned this behavior over time and it has rewarded them well. I imagine that if they're normally dealing with large contracts, those companies footing the bill are probably easier to string along like this. Just do enough and promise enough that they keep you on and only budge when they get more serious about potentially terminating the contract.


I don’t buy this was some mistake in good faith. I think the author is a little naive which made him a good mark.

They started work on and subsequently billed for something that was explicitly out of scope.

My guess is that the founder of the agency knew early on that he could push this client around to extract $.


Also they finished the work in a brisk sprint-like pace towards the end, knocking out the pages in record time. I think that was more indicative of the actual effort required to finish the work, not all the hours they billed. There is a perverse incentive when the client has a retainer to bill hours and make up excuses to line their pockets. They finessed the author and preyed on his naivety


I agree, and it takes force to break them out of this behavior. I know of a guy, who forced the contractors to break down the wall they just built because they didn't let him inspect the place behind the wall before enclosing it, as promised in the contract. He knew that if he lets this slide, they'll keep doing it.

An analogy here would be forcing the agency designers to completely delete any speculative redesign work, without any backups, to force them to focus on the logo, and send a strong signal that I don't care about the redesign, and won't be paying that work.


Most of my career has been spent at agencies.

I think you're attributing malice to what was more than likely routine mismanagement.


> I think you're attributing malice to what was more than likely routine mismanagement.

I wouldn't call it mismanagement. Many agencies thrive despite regularly delivering these types of experiences. It's a conscious choice they can easily rationalize because of the money.

This agency turned a one off $7k job into $46k by smooth talking and scope creeping an actual developer. I'm sure they're absolutely killing it doing the same to non-technical folks.


I'm sure some do have that attitude but I don't believe, for most, that they're choosing to operate that way.

TinyPilot is clearly successful and growing. Any smart agency would nurture that relationship, deliver a great product, and secure an ongoing relationship.

By contrast, there's a huge amount of reputational risk in allowing things to spiral. They've walked away with more today but imagine for a moment the damage that would have been done were they named in this article. It doesn't take much for a brand to turn toxic and things to implode.


But hey, they were not named :) And that lines up with my experience in agencies - there is always an “Isaac” who’s job is to cool the customer down.


Yes, I’m very confused why he did not make the agency.


My guess - contractual obligations or some other liability?


What benefit would the author get from naming the agency involved?

There’s a big legal risk involved in doing so and no upside that I can see.


I was disappointed the author didn’t have the guts to name them.

I would hope he could get some satisfaction in knowing that he might be able to stop them from taking advantage of clients in the future.


Easy for an outsider to say, though I’d imagine it’s a perfectly reasonable precaution to be cautious in the first post they make. Also he’d then not need to worry about what he exactly says in the post or mince his words/self-censor.


Yeah my read is they were having some internal issues and were overloaded. On top of that they hadn't done many projects like this.

Management threw it out to the floor and the floor ran with it like every other job. Feedback went to management but the people doing the work didn't really understand that this client was not like their other projects.

I've been on the other side of this where one bigger client changed the game and we didn't adapt as we were used to things at a different scale. We didn't clean up the mess we inherited fast enough. We didn't put enough resources on the projects and we failed to live up to expectations. We were used to smaller tasks with more limited scope.


Perhaps, but there is an incentive to continue that mismanagement, rather than fixing it.


The agency is incentivized to draw out the work to maximize billable hours


Such things can happen very easily even when everybody has good intentions especially if nobody keeps the entire team focused.


Yep. It's not a situation where they're an indeterminate way through building a custom app using a custom framework and you really do lose everything if you start again, it's a case where an agency has designed and handed over a logo and some very conventionally-designed mockups which any competent freelancer with knowledge of Vue should be able to implement on your platform, and revise as requested. And they're citing lack of availability, so it's not even rude to walk away with the deliverables you paid for.


The sunk cost fallacy is perhaps the most useful thing you learn in business school. eg. Do I want to pay $9,600 to get the job finished. What would it cost if I hired someone else to finish it ?

The same thing if you are going to a concert and lost your tickets, do you want to pay $200 to go to the concert (not taking into account that you've already lost $100, you could have lost that $100 on whatever) eg. you are paying 200, not 200+100.


Pretty sure the agency is well aware of this fallacy, which is no doubt why they treated this guy so well in the very beginning.


You are right But it's not always a clear cut. Sometimes trades have "artistic integrity" for a lack of better words. Carpenters, architects and also programmers - The client tells them what to do, but they usually want some freedom to leave their mark.


I call this the being too nice mistake. It is like the customer that would watch an employee spit in their food and say thanks to them. Effective leaders do not fear confrontation.


In this case it might be a case of being intimidated by a big agency. Isaac telling him that he was their smallest client had a "you're lucky we are willing to work for you" feel.


> I thought I’d enjoy service normally reserved for large companies despite my limited budget.

Sounds like that's exactly what he got, tbh.


This feels even more dubious because if I figured out the correct company, they're fishing in the freelancer threads quite a lot. They know what they're offering to HN readers.


This would be first red flag for me. If someone from the getgo positions himself as doing you a favour, switch - it is not going to get better. At best its uncontrolled ego problem, at worst its a scheme to dry you up.


You're correct, and I expect the real issue is that working that was was completely foreign to WebAgency.

They typically are employed by businesses who are looking for a very specific result: a good website that works well.

Their team wasn't setup to, and didn't know how to just deliver a simple logo.

The correct answer would have been for WebAgency's CEO to say, "We'd love to work with you, but we're not really setup for this type of project. If you'd like to have us take on your whole website, here's what that would look like. Otherwise, I don't think we're a good fit."


If it's not part of the agreed upon scope and they didn't ask for permission, I see no obligation to pay them. They can put the logo in context, sure, but if they end up spending more time on the sketches for that context than the logo itself, that isn't what you ordered.


I think that at the time, the agency upsold the author and they went along with it; after that, sunk cost became a factor. I do believe this is what most agencies will try, upselling I mean.

That said, clients can also be problematic; I recall one where the client paid a fixed amount per two-week sprint, but then really started to stretch out how much work was done within that sprint by basically not signing off on tasks and reporting new features to be added as bugs. It's another case I think of a party not being firm enough about scope.


> Designers are exactly the same way. Just as technical fiddling is fun and interesting, making new designs from scratch is just as fun and interesting. And, just as fixing bugs is tedious and boring, tweaking designs is tedious and boring.

My god, that explains why modern designs and user interfaces suck so much. That, and the fact that many designers work off their gut feelings and personal subjective preferences, rather than systematic and evidence-based study.


Remembered me of a rumor (or fact?) that almost all Windows 10/11 designers use Mac. And it shows: the end results broke all the UX Windows users were used to.


Graphic design has been a Mac stronghold for basically forever. I remember them being the only exception to the Windows only” rules most companies had for a very long time.

Also as a Windows->Mac switcher myself, I’m not sure breaking Windows UX is a bad thing.


It is. If I didn't like Windows UX, I'd switch to something else on my own, thank you very much.


Don't you wish there was a vibrant ecosystem of commercial OSes, each providing a different choice regarding the UX or other aspect you want to prioritize, each being financially viable based on their merits and not getting crushed by a monopolistic gorilla ?

You'd have options to go to when Microsoft makes choices you don't agree with.

I sure wish we were in that world.


Google has decided that the clock on my phone must be HH\nMM and there is nothing I can do about it :(


There's LineageOS. Not sure it is worth the effort to switch just for this though.

Which is also the solution for "OS gorilla". Use Linux. You have a vibrant ecosystem of distros, they work (as much as any OS), and you are free to mix and match. No need to stick to OS that changes under your feet with no recourse.


I'm fine with there being a vibrant ecosystem of open source OSes, each providing a different choice regarding the UX or other aspect I want to prioritize. The monopolistic gorilla is there but the only really effect it has is that other commercial developers like to only support the monopolistic gorilla.


My interest in commercial OS es is on the support/ contracting parts. For instance if tomorrow Sony wants to make a non Windows VAIO machine and look for a partner with an OEM program, open source OSes are out of the picture.

They could take a LTS somewhere, pay another specialized company for additional expertise, but the would be no way to have an actual contract with the leading entity, and have any weight on the direction the OS will be developped (MS is also bigger than any single OEM, but they will still try to keep the majority of OEMs happy and not pull the rug under their feet. Ubuntu for instance doesn’t have such strong incentives)


We’ve had that. Turns out it just shifts the battleground from “I want my company to control the world” to “I want my ideas to proliferate through the whole world”.

Not only do we end up with https://xkcd.com/927/ (15 standards), but we end up with massive amounts of work being essentially trashed because it’s in one of the tiny tiny fragments that gets passed over.

Maybe I’ve just gotten old, but I’d even compromise by making the “power user” the most targeted demographic again. Power users want things to work well, they want the ability to use hardware and software on their terms and for all sorts of things the creators may have never intended, and they want to customize anything that they choose to.

Remember how XP’s theming engine was trying to win users back from all the replacement “shells” like litestep? Or how macs had wild extensions like one that could rotate the mouse cursor as it followed the point? Or mouse-based gesture applications? Or generative art screensavers? Or websites dedicated to showing of the beautiful artistry of user-created desktop UIs where each was entirely distinct from the other (even on Windows!)?


Even the Pope had to keep Michelangelo focused on the Sistine Chapel and not wander off to work on his tomb.

Not that it matters in this context but wasn't it the other way around? Didn't Donato Bramante try to sabotage Michelangelo by convincing Pope Julius II to give him the Sistine Chapel comission, assuming that Michelangelo would fail (due to his lack of experience in fresco painting) and ruin his reputation in the process?

So Donato Bramante was more a consultant recommending an overpaid design agency hoping to benefit from their failure.


I was actually thinking of "The Agony and the Ecstasy," which is on somewhat shaky historical accuracy footing.

Rex Harrison whacking Charlton Heston with a stick because he's slacking on the ceiling is how all project managers should handle both devs and creatives. I hear that's how Larry Ellison does it.


We’ve heard of Scrum. Now comes Stick


Yeah I sort of rolled my eyes at all the logo redesigns (especially as these seem to have come first in the process) but it seems Author was fine with it. It’s an impressive read because Author (and Isaac to some degree) seem quite even keeled in dealing with everything.


> seem quite even keeled in dealing with everything

I'm guessing the money they lost in this endeavor didn't materially impact them.


I know we don't know costs, but per the article it's about a month's revenue, and I assume it's pretty profitable. Not the first hire either, so the difference between 7k and 46k total over several months is not a lot in a way.


I think we do know costs in this case! The author is very open about the finances of his products. And in this case, while he did start generating profits from tinypilot in 2021, those were only around 17-18k$ IIRC. So this is still a massive expense for the business I'd guess!


I used to think this whole pattern was just Chasing the Shiny with a heavy dose of aversion, but something I noticed was that on a redesign, the management tends to give you the benefit of the doubt for a while. There’s a brief period where everything is easier socially before the ugliness starts up again.

Getting a redesign approved can be the difference between having a long project on your resume and people asking you why you job hopped so much.


I love fixing bugs. Sometimes more than writing new things. Seeing the system health increase as I delete old code, add new tests, and increase performance is very satisfying.


I hate fixing bugs, but I love having fixed them more. It's the same sort of relationship I have with playing hard games, like Elden Ring, or whatever.

The actual search for the problem, and sometimes even the implementation of the fix, can be really frustrating, but that is a vital part in the absolutely ecstatic feeling I get from having actually Fixed The Thing.


You are a blessed, holy person. Not even joking a little bit. It's people like you that keep the world ticking over.


This makes a lot of sense, I feel like this is more of a management issue, it should be the responsibility of the manager to keep everyone working on what is initially planned.


They kind of addressed this by saying that they try to keep management billing hours quite low. Perhaps this low percentage would have been "enough" had the size of the project been bigger? Although, I think it'd be fair to argue that the designers should have been "trained" (i.e., with time not billed as "management" for a client's project) to stick to the project scope a bit better.

I get it though: my only experience has been with internal, full-time employees, and I see is given a fair amount of latitude to do exciting things and to experience small failures. It's a different mindset for external contractors, it seems.


The agency would normally allocate 5% of the budget to management, but Isaac made the decision (since the budget was so small) to completely eliminate that. The budget ballooned to 6x its planned size and the management allocation was the same percentage of that as of the original budget: 0%.

Rather ironic, since that latter budget probably is sufficiently large not to eliminate management completely. (~$350 vs ~$2000 for mgmt)


> a lot of technical debt ... let's rebuild the whole thing using Rust and Kubernetes and Deno and move the hosting to Azure etc.

For me, I'm the dev who says: "This codebase has a lot of technical debt. Let's get rid of all of the containers and VMs and kubernetes and artificial servicification, take it off the cloud, and refactor it into smaller programs which do the work efficiently and which can be built and run on basically any machine(s) and cooperate peacefully."


But definitely rewrite it in rust ;)


I wish I knew more devs like you. My life would be so much easier if we could all think like this.


Oh not only a rewrite, but also a re-architecture. What could go wrong?


Well considering theyre removing tools like kubernetes but keeping the micro service architecture, probably a ton.

All the problems Kubernetes and the like solve they’d have to rediscover one by one


1. Things have already gone wrong...

2. I guess I would add "slowly and carefully".


Probably a lot less than when making the move in the other direction.


> which can be built and run on basically any machine(s)

So like containers?


Uh... no. Try: Build configuration and dependency management tools.


I think it's important to establish clear goals before reaching out to freelancers and agencies. The agenda from each one is different than yours.

I learned to get the team focus by writing a brief (commonly used for Brand Managers in big companies) to keep focus on the deliverables. That helps to avoid these type of issues.


The agency was a JS abusing their client so idk why you would just lash out at new great technologies when they weren't part of this.


It's a pet peeve and a common problem in software development, where developers don't want to do the work that is asked of them because it's not challenging enough (yay another REST API in front of a database), so they add complexity themselves - new, more difficult languages like Rust or Scala (not as much of an issue anymore I think), excessively complex architectures like microservices, DIY infrastructure configuration like k8s, etc.

See also magpie developers. It's not that the new technologies are not great, it's just that in most cases it's not what a client needs, but what the developers want to do. They will have a solution and look for a problem to match.

I mean I'm not innocent of that, I'm looking left and right for problems where I could use Go to solve them with because I enjoy it. But I am aware of the risks.

Anyway the comparison is apt because it happens with developers; I've seen it happen firsthand at multiple projects, and in almost ALL of those, after the consultants left, the company moved back to their own technology, picked something off-the-shelf, or did yet another rebuild.


I think the comment was fair; Too often have I witnessed clueless and/or conniving salespeople recommend businesses to move onto Azure or other cloud services when there was very little to gain from it, yet it meant that the cost compared to hosting their very simple ERP software on metal would bankrupt them in the first bad month.


> have to keep the devs from going "this codebase has a lot of technical debt, let's rebuild the whole thing using Rust and Kubernetes and Deno and move the hosting to Azure and switch databases and use microservices and..."?

Also known as the "oh look, a shiny thing" development paradigm.


I wonder how you judge the tipping point?

"Why do it in C when assembly language has been working so well for years?"


Following through the example, if assembly language works well, then there's no justification to do it in C. One needs to find (valid) reasons to justify such decisions.


The tipping point would be when you can come up with a compelling answer for the question you laid out. You don't just rewrite it in C as a reflex to that question.


I recently left a boutique agency of 5 years and I can definitely resonate with this one. Our agency aimed to catch big fish, and we did, but since they are hard to land we'd pick up small jobs in the meantime, just like the project that you're describing here. In my perspective, this isn't someone deliberately ripping you off. I imagine they intended to ship at the cost they quoted, but the team didn't adjust their working style to match your price point.

All the variations of the logo and design mocks are clearly overkill for a $15k project. The design team had time to fill and wanted to provide lots of options for you to pick from, as they typically would on a larger project. Those variations are an expectation for $100k clients, and you got the $100k customer treatment, but unfortunately not at a discount.

The reality is, small jobs like this are effectively make-work projects for an agency. They typically don't pay enough to be an effective use of time for the agency, but are a way to stay in the black between higher value projects. Small customers become "nuisance" customers as soon as a something better is landed. The team members being swapped out as they are needed elsewhere and newly joining team members then need to re-contextualize and regain momentum, all on your dime.

Your takeaway is correct, don't be a small fish for an agency. If they're busy they won't take your work, and if do show interest, they are desperate for work.


Interesting perspective. Thanks for sharing!

>All the variations of the logo and design mocks are clearly overkill for a $15k project. The design team had time to fill and wanted to provide lots of options for you to pick from, as they typically would on a larger project. Those variations are an expectation for $100k clients, and you got the $100k customer treatment, but unfortunately not at a discount.

Oh, huh.

I was thinking about this as I did the writeup. It didn't feel at the time that they were spending excessive time on the logo variations, but I went back to the notes I took on our first call and realized how out of line all that early work now feels relative to their 30-40 hour initial estimate of the rebrand.


It wasn’t scope creep it was a scam. Shady contractors in every industry pull stuff like this.

They didn’t deliver what you actually contracted for until you put your foot down because that was the hook — they couldn’t keep taking your money if they gave you what you wanted.

For anyone else reading this the best move would be to shut it down as soon as a single minute has been billed for out of contract work: “Hey this is not what I contracted & I won't be paying for any out of scope work.”


Absolutely this and you can take them to small claims court for breach of contract.


Here's the thing I don't get. Did you not have some guarantee of completed work? When they quoted you the $7k, I could see building timeline flexibility into that quote, but I have a hard time imagining building so much flex into the contract in terms of budget that they can hold your work hostage to where you sign the retainer contract. It seems like you should have had some sort of legal recourse to hold them accountable to delivering the promised work in a reasonable time frame before you went the retainer route and found yourself on the hook for an extra $40k. Was there not?

Also, I was surprised by how much you let Isaac get away with. He admitted he badly mismanaged the project, and made decisions that lead to that mismanagement with out consulting you. He badly blew his estimates for you, and was pretty clear that it was his fault. I would have pushed him to eat much more of the losses than he did.


What's 7k? One week of work + client comms ?

You can't really imagine anything being completed and deployed in that time.

It was a death trap to begin with. They're experience should have flagged this.

But both are equally at fault. This is what happens when clients have the lowest bidder mentality.

"the cheapest price is always the best option because it's the cheapest."


> This is what happens when clients have the lowest bidder mentality.

Your conclusion is completely at odds with this quote from the article:

> WebAgency quoted the highest rate of anyone I interviewed, but their portfolio best matched the style I wanted.


Yeah, it's more of a "highest bidder mentality". That they don't have any experience with hourly billing should have been a massive red flag as that means they need to be very conscious in how they approach the project. That they used this simply as an excuse to deprioritize the project is what caused the problems.


The agency I worked for used to offer a "special" for a brochure site. Basically it was a "website-in-a-day" deal. Customer comes to the agency in the morning, meets with designer and manager, and they agree a design (preliminary work has already been done on a logo). Customer goes for lunch, and the design is handed to two devs. Around 4PM the customer gets sight of what the devs have done, and goes home. The devs work on until say 8PM, and probably put in a couple of hours the next day.

Of course, there was then snagging; it wasn't really a one-day job. The whole deal was pretty inexpensive. We used Drupal with a custom theme.

I don't know whether we made money on these projects, but we did end up with a string of long-term customers.


Couple of questions:

1) Did you pay everything up front or was it a 50% up front and 50% on delivery basis? 2) Did you manage any deliverables on a google sheet (dates, owner etc) or something similar?


>The reality is, small jobs like this are effectively make-work projects for an agency. They typically don't pay enough to be an effective use of time for the agency, but are a way to stay in the black between higher value projects. Small customers become "nuisance" customers as soon as a something better is landed. The team members being swapped out as they are needed elsewhere and newly joining team members then need to re-contextualize and regain momentum, all on your dime.

Man. That 3rd paragraph resonates so well with me/my employer. We're an industrial automation company. Family owned. Started from the owner's shed and grew to what we are now. We were built on smalltime clients and our product quality got around through word of mouth and are now at the point where we have massive multi-million dollar clients.

We still support the little guys though. And we get more little guys under our umbrella every year. I think there's still a part of our company that recognizes we have roots in helping the farmers automated their cleaner processes. We also have the nearly identical issue that our OP is bitching about: we're married to these massive clients and we fill the gaps with the little projects. But when the timing is getting tight, the little guys are who loses out.

I'm starting to see the cracks. Clients who built our foundations are losing out on support and growth opportunities. We're more concerned with the next mining project or new facility build than we are selling small guys upgrades and ongoing modernization. It's fine as far as the pocketbook goes but I feel like we play a dangerous game allowing our work schedules to be dictated by the big guys. Eventually they all grow to realize the same thing: the controls part is crucial enough to the business that it needs to be brought in house. Once that happens our value falls off quickly. It's only bad because we're losing our core for the opportunity to play puppet to some truly massive clients.

I feel like I'm getting a bit lost in the weeds, but really my point is just how I haven't really thought clearly about what the perception of our business must be to the clients, both big and small. We play a critical service role among many industries but we also run the risk of alienating the business that's virtually guaranteed to be there in hopes of marrying ourselves to somebody who only needs us now, and probably not tomorrow.


If retaining the small clients is a business goal, you should have a (small) dedicated team devoted to them, that knows how to work in the way that these small clients need.


Don't you think that'd become the "B-team" pretty quickly, though? I think it might fall apart. What might also work is something more like the "pro bono" model of lawyers? Make it a marketing thing, maybe see if the big clients would want to get in on the PR benefits of helping the little clients, maybe rotate your rockstars through the smaller projects and help develop less-experienced talent that way?


It has the risk of being a B-team but it's also a good way to onboard new hires. You get a much higher turnover rate of new things to try, do, and learn. One place I worked did communications generation (PDF letters etc), we had massive clients and 300 hour projects and we had 3 hour simple letters. Churning those small ones out was the best way to get me to learn when I signed up. Just as long as someone sensible and experienced is watching. And some people just don't gel well with project marathons - I'm definitely a sprinter who loses focus on huge projects despite my years of experience. I'd rather do a lot of small ones.

It eventually got sold to a bigger more respectable company and those 300 hour quotes crept up north of 1000 hours and it was impossible to finish any job in under 10 hours but that's another story about internal red tape.


Other way around, in my experience. We run a team of ~20 at our studio—the "A-team" all want to work on the smaller projects because they're usually more interesting, they have more control, they get to move faster, etc.


As the owner of a control software company, your explanation about big companies bringing things in-house matches our experiences perfectly, and instead of assuming that they will never be able to bring things in-house due to the niche and a lack of experience, it has inspired me to put some more thought into how to solve this problem as we grow by focusing on having teams dedicated to dealing with and keeping our smaller recurring clients (i.e. have A and B teams as suggested).


> Eventually they all grow to realize the same thing: the controls part is crucial enough to the business that it needs to be brought in house.

I think that creates a narrow operating area for almost all agencies - they need to operate in a narrow gap between "work that isn't valuable enough, so nobody pays for it" and "work that is so important for the business, that they bring it in house".

And you get to play outside this envelope only by finding clients so slow that it takes them years to realize they need an in house team. Now you have a different problem - working for a slow client doesn't push you to get much better.


Why not have a sister company/branch for the small clients? One they grow, you'd be able to transfer them to the big block?


Yeah, I like the suggestion, I'll preface in saying I'm not in a "Decision Maker" position at my company though so ideas are only so good.

Ultimately it's a management issue. Internally among the techs/engineers we've talked about it but management keeps pushing to satisfy the client and the temporary compromises we make at one moment start to become permanent mainstays of how we operate. I don't have any confidence that if we had a "B-Team" or "Sub-branch" that when a big enough project came along they wouldn't just poach the needed labour and destroy the separation between them.

You could almost say we operate with an "A-Team" and "B-Team" as we have techs split primarily between the Mining and Agricultural clients but they take our techs often from the Agricultural side to fill in on needs of the mining side. There's no shame.


I've been at a small agency that's grown into a midsize agency over the last decade+.

Everything about this story rings true, and the author's conclusions are absolutely on point.

Now our agency is at a place where we can say no to work like this, both because we have a solid client base that supports us financially without projects like this, and because we've learned that no matter how good our intentions, neither we nor our clients are going to be happy with the end result.

All that said, my bigger question is: does the new website bring in more business than the old?

It's certainly better designed, but looking at the copy and IA, I'm not entirely sure that the new site is going to convert better than the old.

To me, the old version was distinctive and unique, while the new looks like basically every other SAAS site designed since the year 2020.


FTA sales increased 66%


Yes. But what we don’t know is

1) How would the old design perform during that time period

2) How would a hypothetical $7k design perform during that time period

#1 is of course why a/b-testing was invented, while #2 is even more speculative. But we don’t know how much of the 66% can be attributed to the new design.


Also the earnings chart shows a clear upwards trend, so 72k is not really an outlier. Fwiw, I liked the previous design better. No idea how the codebase looked like though.


I've worked at small agencies for the better part of a decade and couldn't agree more with this summary.

A lot of the comments seem to take the view that this was some deliberate ploy to overcharge. In reality, it was just poorly managed.

> Small customers become "nuisance" customers as soon as a something better is landed.

My guess would be that when quoted, the project was expected to be completed by a certain date. Because the team failed to adapt, the project overran, new projects took over leaving this one to languish.


As a small business owner myself, this resonates. I would love to be in a position to be able to confidently pay money to other professionals to make problems go away.

Unfortunately, my overall experience has been that hiring any “expert” in a field that I am not also an expert in has a 50% chance of working, and a 50% chance of blowing up in my face and creating more problems.

I recently attempted to get a new accountant to help me handle some business growth. It was a person from a well regarded local firm, initial meetings were good, and then they proceeded to deliver none of the agreed-upon work, take 2-3 weeks to respond to emails (multiple times, I had to call their office and schedule an in-person meeting just to get a response), and then de-prioritize my business relative to other clients so badly that I wasn’t able to submit my taxes until June.

If anyone can successfully build a service that lets me reliably pick professional-services providers with the same level of confidence that I pick an AirBnb (not 100%, but pretty good, with an expectation for reasonable mediation and fallback coverage if the offering is radically different than what’s described), I would happily pay a 20% premium on those services versus the existing “ask friends for referrals and cross your fingers” status quo.


If it’s anything like the market for home services, the middleman service will be even worse than dealing with vendors directly: a middleman adding a layer of indirection but little value.


Depends on how much work you are having done. On a remodel you need a GC unless you have a lot of time and expertise to hire all the trades and such yourself. They really do provide value.


I've used marketplace apps for home services, and just as airbnb, the results were often stellar and always at least pretty good.


Upwork? I have okay experiences with people there - not perfect but in range with your stated confidence levels.


Is this an example of the Market for Lemons? If I hire a developer I have a pretty good idea what to look for (and to a lesser degree, a designer, as I have worked with enough designers to know how the good ones work). But I would have no idea what makes for a good accountant or corporate lawyer or plumber, so would need to depend on referrals and a certain amount of blind trust.


I am building this for M&A and investor due diligence. DueDilio is an online managed marketplace focused on due diligence. We connect business buyers, sellers, intermediaries, and private investors with pre-vetted due diligence service providers. Our large and growing network of verified independent professionals, boutique, and mid-size firms specialize in finance, legal, technology, commercial, and other key areas of business diligence.

We've connected clients with accountants previously with good results.

Our website is www.duedilio.com


As someone who have been on both sides of the table I can completely resonate with your statements. But I think that there is an expectation by small business owners that every expert will completely understand their business and will deliver exceptional results from day 1. So I think there should be some learning curve with very clear results expected on each and every step.


Is answering email at least this week exceptional? What learning curve a small business accountant has even? It’s not a factory with amortizing equipment, worker schedules, rolling production and distribution of expenses. I work with accountants all my life, and for small businesses it’s a matter of loading contracts, bank statements, invoices and salaries into an app, sorting expenses by legal category and that’s it. There is nothing to learn in a small business accounting, it is entry-level for this job title.



As someone who has worked at a content agency for a decade, let me just say: I feel really bad that this happened to you, and that scope creep is real.

I almost feel like you needed to ask for three separate things: A brand identity, a marketing strategy refresh, and then (finally) a website redesign. That all three were combined into one process likely caused this problem to drag on. The agency had its problems, but to be honest to me as someone who is familiar with this space, it sounds like they were combining a lot of disciplines into one project without considering that it would have been better to chew smaller bites.

There are times where you do need to bend the rules. At the beginning of the pandemic I sort of broke protocol to get a COVID-19 landing page on a client’s site online because I knew that it would take weeks done the normal way and possibly would have led us to charge the client for something that a skilled designer only needed a couple of hours to build in WordPress. While the landing page wasn’t perfect, it held up for nearly a year, and showed that we were taking things seriously at a time we needed to. A lot of agencies aren’t wired for doing right beyond billable hours, so be mindful of the risks.

Either way, I feel bad that you paid so much for a site that looks way better but doesn’t feel like $46k worth of work.


Don’t feel bad. The end result was positive. Sales increased enough that the cost will be made up in a few months. Person actually made out well considering the effect of the work.


It's not at all guaranteed that increase in sales was due to the redesign.

But yeah considering that revenues have only gone up, might as well just call it a cost of doing business. Learn from it and move on. (It sounds like this is exactly what he's doing.)


But imagine how much further he would be ahead if they came in at time and on budget. He arguably lost half a year of positive benefits because they dragged their feet for so long.

Things can be positive and cost-effective, too.


On the whole I agree with you.

That said, margins on this kind of hardware product aren't great. The author has shared them before. https://mtlynch.io/solo-developer-year-3/


I live through this every 2 years:

- Marketing team decides they want a new site.

- I tell them when/how we can schedule it.

- They decide they want to go outside so it can get done "quicker" by "professionals."

- It costs 5-10x what it would in house, the product is harder to work with, using some WordPress plugins no one has ever heard of, it's not responsive on mobile nor usable on our demographic's primary resolution.

- It takes 6-10 months of "clean up" to make the site usable.

- Web traffic, shockingly, has remained completely constant even after spending half of our annual marketing budget on a web site.

- My team is brought in when the agency becomes too slow because the entire team over there has turned over since the project inception.

- We eventually migrate everything over to squarespace or weebly or similar so that the marketing team can just edit things on their own.

- Every lesson above is forgotten in the ensuing 12-18 months.

We are an early stage startup. We've burned through almost 20% of the revenue we've ever brought in on this cycle. Thankfully, finally, we've grown enough to bring on a marketing manager who will I hope put an end to this madness.


A tale as old as (Internet) time - I've seen this cycle happen, too.

Tangentially, I have to wonder to what extent misapplication of Agile, and similar, project management processes is to blame.

You'd think for most relatively simple sites, like we're talking about here, it ought to be planned once and built once, but something about the mindset that the goal posts can be moved during planning and development seems to drag everything out at length.


I think that's pretty backwards. Non-agile is how you get these things being rebuilt every other year because by the time it's built the requirements that were originally gathered are obsolete.


Wow this one is spot on. Had exactly the same experience as a contract PM working with Marketeers who lived in fairy-land.

Two hilarious moments: 1. A big fuss was made of a launch of a new site, marketeers demanded that we had plenty of extra servers for the demand. On launch day I had the live stats of visitors and it occasionally flickered to above 0. 2. We finally needed to refactor and do some maintenance to the duct-taped code that had resulted in years of 'everything is urgent' 'one extra feature by tomorrow`. When I told the Head of Marketing they were doing this vital work, she couldn't understand that because no new graphics or UI were being produced that they were doing actual work - "But I can't see any new work"

It was not worth the small fortune in daily rates they were paying me


> - Marketing team decides they want a new site.

That's your problem right there. And the fact that their decision is the company's decision.


Ideally, who should decide what the public face of the company should be if not for marketing?


Founder / CEO. A lot of people will respond "no the founder needs to back off and let the marketing team do its job" but on a decision to allocate 50% of the annual marketing budget, the CEO should be involved.


I suppose it depends on the expertise of the the founder or CEO in question. On balance, something that takes 50% of budget absolutely should involve oversight greater than a single team or executive, but I'm not sure to what level without knowing more about the team dynamics at play.

There are plenty of engineer founder types who have all sorts of opinions on how infrastructure spend should be divvied up...they're probably on average more well informed there than they might be on the value of good consistent branding and website. On the flip side, a sales-heavy type might have strong opinions on how much engineering is spending on keeping the ship afloat, but that doesn't necessarily make them right.

Of course it's possible to be conversant within multiple modalities, so I am just talking to the average here.


Bezos famously had iron grip on the Amazon home page - I guess it worked for them. So yes I agree


The Amazon homepage is dogshit, people used Amazon because it was a service that they really wanted. Amazon's success is in spite of its web design.


But it's not in spite of Bezos being a control freak. It's a good quality in a founder.


Let's not get into ideals, but: The different stakeholders in the website should agree on it, rather than just one of them deciding - as marketing is just one of a website's purposes; it's not merely a marketing tool.


Perhaps it's my own bias but through the lens of founding and running a SaaS business, but the website often just IS marketing, while the core offering(s) are hosted somewhere else and have their own team/owners/stakeholders. It's certainly different if your product is a physical good or service, but I'm reading the GP's comment as 'tech startup = software company'.


> We are an early stage startup

It's no different at massive corporate behemoths.


"Managed $XX,XXX site redesign that resulted in [cherry-picked numbers to make it look like it improved things when it didn't at all]" looks good on a résumé.

Your managers may not understand much, but they can understand (by which I mean get a thrill out of looking at) before-and-after screenshots on a powerpoint, which may matter when promotions are available.

That's at least as true in bigco as in startups and small companies.


But these kinds of mistakes can kill an early stage startup.


Do we work together? haha


> For most of the project, I was sitting on a bunch of partially-complete tasks. The cost of reassigning half-done work and spinning up a new vendor would be almost as expensive as starting over from scratch.

The original project budget was 7k, and you ended up spending 46k, for something that didn't really deliver on what you asked for originally.

It is really hard to imagine that it would have been more expensive to start over. It seemed clear by the middle of the story, that web agency wasn't doing what you were paying them for, at which point you increased the amount you were paying them in hope that would somehow change things, and then were shocked when they behaved exactly the same as they had before.

There's times when starting over is more expensive, this seems like the polar opposite of that and a clear example of sunk cost fallacy.


Yeah, I think the article lacks of a single, clear statement on why things it went wrong. It gets caught up in myriads of little reasons which I think distracts the author from arriving at a painful insight.

His shortest section, left towards the bottom of the page, seems like an accidental example of rationalizing the sunk cost fallacy.

> Firing WebAgency and searching for a replacement would have burned 30-60 hours of management time. And there was no guarantee that I’d find someone better. For most of the project, I was sitting on a bunch of partially-complete tasks. The cost of reassigning half-done work and spinning up a new vendor would be almost as expensive as starting over from scratch.


The author also mentioned that there is a cost to selecting an alternative firm and managing them. Also, I'm not sure the author would have learned the same lessons outlined in this blog post had they not seen the process through to the end.


There always is - that is why the sunk cost fallacy is so attractive. But still, this seems one of the most clear cut example of the switching costs being worth it, i have ever seen.

I'm doubtful that the author learned the correct lesson even by the end. I suspect if a similar thing happened again the author would fall victim to it again.


That is very expansive, for a $100 a lot of teenagers could make it.


> That is very expansive, for a $100 a lot of teenagers could make it

These are the $4 devs. Look, a teen coder can hack up a storm. But I distinctly remember a brilliant friend of mine taking a gig in college and not understanding why the small business who contracted him was pissed he’d rebuilt their site in Silverlight. He wouldn’t make that mistake now. But he doesn’t charge what he billed then either.


Sorry but I'm in a few communities with VCs, designers that aren't even 16yo. In the Rust or Deno communities: https://lcas.dev/ is very young for example.


Do you think you could hire Luca for $100?


It depends who I am, I believe he would do it for free.

Digital Agency Owner Here... I have to say, this is difficult to read and it gives a bad name to other agencies.

- I think the first "mistake" was not having it as a fixed price. We spend a lot of time up-front defining the scope and get sign off before committing to the work. After that, we roll it through the relevant departments. We take the risk on for bugs/issues (that are within scope) and do the project as a fixed price. Anything out of scope is a different conversation but will always be priced additions.

- This agency got the best of both worlds, hourly rate and no risk on their side.

- I disagree with some of the comments on here about agencies not being your "friend", a "good" agency will treat you like a partner and be focused on helping you meet those goals, regardless of if there's more money to be made. Ultimately, our job is to make you look good and reach goals. If we don't do that, we don't need to exist.

- They should have said no at the beginning. We turn down clients and it's hard to do sometimes, but ultimately you're setting everyone up to fail. Honesty is key for all parts.

- The onus should be on the agency, not on you. If there's issues with the build, that's the agency's problem not yours. That's why the initial scoping is so important.

- A good agency IS better than a freelancer(or freelancers). You have a team of people conducting the work, pulling in the right people at the right time, expertise and experience of doing this day-in-day-out, what to avoid etc. On top of all that, ensuring it runs on time (it's in our interest to do so). A rolling freelancer is incentivised to keep the project running as long as possible, but on top of that you're limited to a knowledge of one.

- The CEO unfortunately took full advantage of the situation here. This is a short-term approach that may work for a few clients but after that, the word spreads. There's few things you can keep hold of these days, your name and reputation is one of them. I'd advise actually naming this agency as the anonymity will only promote them to keep doing it.

Sorry this happened to you and I admire your positive approach to the situation, sometimes what's done is done and you have to learn and move on.


Fixed price upon scoped deliverables is the way for such a small project, when using new freelancers always use fixed to trial how they are, then possibly move on to hourly. On those freelancing websites you can cap hours so they don't overbill for out of scope stuff, this is a rough experience but not that uncommon especially for first time.


I agree with most, especially that OP left the tap open. No risk for an agency and de facto unrestricted costs for OP.

With one exception of a blanket statement:

> - A good agency IS better than a freelancer

If there is a single freelancer:

- There is a single person responsible. If it gets diffused, it may easily end in an unending project, as no one is incentivized to finalize the project.

- The number of people working on the project is capped at one. You won't ever get charged for two people talking with each other.

- For a single freelancer, they may be some "sanity check" regarding the cost. For any well-known agency, it wouldn't raise one's eyebrows if you said [This Well Known Agency] charged you [any number] $. Logo design alone could have costed OP $50k.


For me, the 3 key downsides to a single freelancer is:

* You’re totally dependent on a single point of failure. If they get ill, go on holiday or simply disappear, take on another project or go full time somewhere else, then you’re left in the lurch - particularly post-live when you may need additional support/amends/bug fixes.

* I’m yet to find a single freelancer that can project manage, UX, design, write copy and build to a high standard in all areas. You’ll probably need several freelancers.

* Lack of code review. Usually an agency will have several devs on their team, so there should be some form of code review whereas this is less likely with a single freelancer.


As the agency owner, how would you have felt reading this if the author had named the agency?

It really bugs me how people take a "protect the innocent" approach to these articles when in reality they should be saying loud and clear who the offender is.


I'm not sure I understand your comment? Who's the offender here?


The author of the article didn’t name the agency he was engaged with.

In other words, he essentially wrote a really long, detailed review but never mentioned the actual product.

My question to you was simply how would you take it if someone named you in an article like this.


I believe he should have named the agency.

Obviously you never want that to happen to your own agency and would like to think any responsible CEO / owner wouldn’t allow these situations to happen (don’t get me wrong, I appreciate mistakes do occur - but this smells very different), so unless they suffer some consequence, they’ll do it to the next unsuspecting client and keep doing so.


> I believe he should have named the agency.

I disagree. (1) There is evidence that his company may profit from the redesign, and (2) he risks being sued for even the slightest mischaracterization in his blog, which requires tens of thousands to defend (at least). I don't see the point in risking the family business over this. Both parties learned a lesson. Perhaps the WebAgency will decide never to have hourly clients again.


Respectfully disagree. There is only one side that suffered the consequences, and it's not the agency. There is no incentive for them to change anything.


I find it odd that there are people here who believe that the agency regrets anything,

Or that it learnt anything more than "yes this really works for us to maximize revenue and we can get away with it".

(I do guess it's better to not name them though -- because naming them risks making actually well behaved agencies or freelancers nervous about working with the client in the future?)


How would this be different from a review? and could one be sued for posting a opinionated review?

I guess the title intro and conclusion should be changed as a review..


> ultimately you're setting everyone up to fail.

The agency didn't fail? They made lots of money and can do the same scammy thing again and again

More like a success story for them

> Honesty is key for all parts

I think it's not, not for agencies that want to maximize revenue

(Unfortunately.)

Seems you're a bit ideologically driven though, which is nice I think :-)


what percent is charged up front?

This reminds me of when I worked for a B2C company in the healthcare space. We hired a freelance designer to redesign our checkout flow and we wanted it done in time for Black Friday, which was by far our biggest day of the year.

Of course the project ran long and we crunched so that we could ship the redesign exactly on Black Friday. I think we shipped the Tuesday before (because Thanksgiving) and everything seemed normal. Black Friday rolls around and we go into the office and they have our internal dashboard monitors set up with our Black Friday unit sales counts. Spoiler alert: it did not go well. We were something like 25% off of our goal and 10-15% off of our previous year's sales. Exec team is freaking out and they order us to revert the design change ASAP in the early afternoon. We do and sure enough, we see our sales start to increase.

Nobody considered that rearranging the layout and colors of the checkout buttons would have such an impact but they did.


Customers/users learn a bad design and get accustomed to it. Any changes, even ones that ostensibly improve it, add cognitive effort and contribute to their aggregate cognitive overload (taking into account everything else they have to learn and remember on a daily basis). The original design achieved “don’t make me think”, and any changes, even improvements, reset that.


Have to admit, when I saw the two screenshots, I thought the OP's problem would be exactly that, not the agency process. Original design not great but has a big picture of a hardware device, an unmissable order button and some explainer videos. New design much more visually appealing, but looks like a different company, potentially even a different class of product and whilst the order button isn't exactly difficult to find, it's not shouting as loudly to act.


The bottom of the post mentioned that sales have increased by 40% with the new design. It will probably take some time to know if that sticks, but it seems like it works from that point of view, even if it was maybe overpriced.


The 40% increase is ~$18k a month based on the numbers in the article. That means that redesign pays for itself in three months. That's the type of "regret" that I want.


Yeah, the old design could perform better if it loads faster and has an easier to find CTA (order button).

Wonder if OP could A/B test the two designs?


God, I wish this were printed on the wall of every software design office. Mediocre designs are fine if people know them, because they learn to work around the rough edges to the point where they often don't notice them. But a new design (probably also mediocre!) requires way more cognitive load. Tech as an industry is horrible on this front.

Just to pick on one example: Android. Google absolutely loves changing the settings and UX on each major version. People use these controls so much they eventually get habituated... until they change and have to go hunting around and learn the new workflows to get back to par. Each one of these redesigns probably wastes millions of cumulative user hours.


Android 12 was a complete disaster.


It also now has a bug where sometimes unless I reboot the phone it will refuse to show more than 4 of those bluetooth/wifi buttons…

The other day I had to cross a little tunnel and I couldn't put on the light on the phone until I rebooted it -_-'


This is why I hate all these websites that keep A/B testing.

Just when I get used to a layout, they pull out a new design, completely disorienting me.


> Just when I get used to a layout, they pull out a new design, completely disorienting me.

Honestly, I feel like the only way of working around this is having multiple different interface options available.

For example, the new Reddit look is more app-like and certainly has improvements to the user profile pages and whatnot. Yet for certain types of browsing content, or wanting to do it without your browser slowing down as much, the old interface is still available:

  https://www.reddit.com/
  https://old.reddit.com/
Many out there will stop using the site the day when the old interface goes down and for now can just use the old one despite the new one being available - thus allowing them to stick to the user experience that they're used to.

Of course, not many out there want to deal with something like this on the development side, such as CRUD systems that would need to move fields around, add new business process steps etc. There, maintaining two separate versions would be a massive pain.


I would argue that the most important factor when considering old reddit vs new reddit UI/UX isn't a matter of preference based upon performance, certain content, or habit. Old.reddit is actually just better for the end user experience overall and new reddit UI is better for Reddit's ad revenue.

Many times a user not wanting to switch to a new UI isn't based completely in effort/adaptability but a history of experience with product life-cycles weighing more towards business objectives over time. e.g. Facebook calls users lazy for not trying out "improvements" and blame old soccer moms for being inflexible when they're just trying to extract more money. Not that businesses spending effort to get more money doesn't make sense, because it does, but businesses love to lie about this common user complaint.


The fact that new reddit defaults to showing only a few comments on the post, followed by recommending 20 other unrelated posts, just shows how badly aligned that design is with their goals.

Reddit is a glorified web forum. Period. Making comments hidden and difficult to browse basically negates 50% of it's function (the other being media + content discoverability).


I imagine it's quite well aligned with their goals of getting increased user engagement metrics from increased clicks to read stuff from casual browsers to the site, and convincing regular users they should download the app

Of course it's extremely badly aligned with their regular user's goals of reading comments, but that's solved by using the old.reddit urls if not the app, whilst the casual browser coming in from Google or a link gets the full on contempt for users' desire to actually read threads UX until they've bumped the user engagement metrics up by clicking on more stuff.


old design is easier to process. not sure if its just me but seems like the new design wants to tell me what's important and I have to fight it spending precious brain cycles


If you want old then there's also

  https://i.reddit.com/


That's because more often than not you're not the target audience. Growth > retention in many cases, so it's more important to give a good first experience than a keeping a good continued experience.


I have a rule now that when designing a page, any "money screens" get at least 1.5X to 2X the estimate. I define a "money screen" as anything that leads a company to land a client or land a sale, things like checkout flows, signup flows, etc. Usually that extra time gets sucked up in A/B testing setup and setting up a staggered deployment per region that the biz operates.

Whenever customers push back I tell them the story of Knight Capital [1]. You pay extra for extra assurance that you won't loose a shit load of money in the future.

[1]: https://www.henricodolfing.com/2019/06/project-failure-case-...


In the early 2000s I was a professional day trader and Knight was a market-making firm. EVERY single person I dealt with was an absolute crook, happy to break rules and do disgustingly horrible things to enrich themselves, because they were truly incompetent traders.


Apple gets a lot of flack for keeping design constant over the years, but this is the reason why they do it.

People hate changes to their workflow.


People crave consistency. McDonald’s isn’t popular because it’s good, but because the burger you eat in Santa Monica is the same you’d get in Pigeon Forge, TN.


McDonalds’ fries are actually good. But point taken


They're hot fries.

Warm greasy potatoes with a bit of crunch are always good.

In 30 minutes? Not good fries.


Wow, you are very lucky if you've never had really bad fries. I've been to places and had undercooked fries, burnt fries, fries with almost no potato in them, soggy fries. McDonald's fries are very okay but they are always okay. They are very rarely hot though, usually quite old, but at least not stone cold like KFC fries (in the UK and Europe we have fries instead of mash with fried chicken). The fries in Belgium are by far the best, but there are some great ones here in Germany.


I've eaten some potato based food crimes in my day. Agreed that Belgium has consistently good fries everywhere.

Consistently mediocre (3/5, thoroughly passable) is the value prop of fast food.


What they do in the UK is horrifying. The soggy oily mass that you eat with a prong thing.

I even went to the current winner of ‘best fish and chips’ that year, in Whitby. Argh.


UK chips are my favourite in the world, but they are qualitatively different in every way. I always get annoyed when other countries claim to copy fish and chips but serve them with fries. There isn't anywhere else in the world that you can get chips like that, so inevitably the fish and chips that try and copy it are always disappointing. I know it's controversial, but are supposed to be soggy and oily, not crunchy. It's supposed to be like eating oily potatoes. There is a reason why fish and chips is so renowned and loved, and the chips are a big part of it. They are the best in the world bar none. I only didn't mention it earlier because people find it quite offensive, because they are so much different from other chips.


I understand everything you're saying, but I somehow keep reading it as "they're supposed to be bad, that's why fish and chips are so loved, they're the best in the world at being bad".

I mean, yes, but still, I want my potatoes crunchy, not soggy and oily.


I'm a sucker for thin, salty fries. I know they're low quality but my mouth enjoys them


I don't know where Santa Monica and Pigeon Forge are. But traveling around Asia... McDonalds doesn't taste the same between Thailand, Singapore, Malaysia, Japan, Taiwan, Australia, New Zealand.

Infact Thailand burgers are VERY salty...


I wonder if that's particular to Asia, because I've noticed McDonalds is the same as the US in Latin America (gringo safe space), aside from a couple extra menu items


I wonder if this is an intentional choice to adapt to the tastes of local markets?


I would assume so, Singapore for some reason HATES salt. They don't put salt on fries from mcdonalds, so you always end up with a bag of soggy fries.

https://www.straitstimes.com/singapore/health/singapore-hear...

Taiwan has burger king burgers with peanut butter on them...

https://www.burgerking.com.tw/jps9805


Throw some sriracha on that BK PB burger and that would be _magnificent_.


*Adds "peanut butter and sriracha" to list of improbable combinations to investigate*


It is beautiful. Do it. Works best with sweetish-savory items. I first stumbled upon it when I put peanut butter & Sriracha on an omelette filled with pork sausage, bacon, and browned onions / mushrooms.


You could pretend it was satay?


But unfortunately not true in other countries! In Hong Kong the menu is very different and even the fries are different!


mc donald in italy has espresso and croissants


Quite to the contrary, Apple makes gratuitous changes to the UI of their OSes on an annual basis.


Not on things that makes people feel like they can find what they are looking for. The Apple drop down menu hasn't changed its place since the original Mac OS.

The UI skin might change a bit from year to year, but the UX of Key components haven't changed much for 40 years.

Used to work at AppleCare special programs support in Ireland, and when we did change something like hide certain menues in the Server appbecasue we wanted to disincentive usage of parts that were going to be depricated, all hell brekas loose with the older sysadmins calling in furious over having to read change notes. Thing is Apple makes sure to communicate a lot of these changes in emails sent to users, in change notes, and in popups, and people just refuse to take notice.

Or when Final Cut Pro X changed it's library management. There were good reasons for these changes, but people at first hated them since it disturbed an otherwise uninterrupted workflow. But when these changes are made, it's usually for a very good reason based on user feedback.

This did however highlight a serious behavioural issue with the users. Some users rarely update, so they never get the incremental changes, and instead opt to jump 3-5 major release versions which breaks everything. This in turn makes the user wary of ANY updates as they now associate it with breaking everything and changing everything. Which is a problem that they would not have if they did follow best update practices.

I saw the exact same issue at HPE, Nikon, Salesforce, Microsoft, when I worked there.


> Apple gets a lot of flack for keeping design constant over the years, but this is the reason why they do it.

I am not so sure, they are turning system preferences into an iPhone app on macOS.


Not sure what you mean by that?

The UI/UX design of the System Preferences hasn't really changed since Snow Leopard imho. Cant speak to much about before Snow Leopard as that predates my personal experience.

I'd love to get an example of what you mean?


You can eventually learn to use a bad UI, but you'll never learn a constantly changing one.


Very true. This is why I rage quit Using Windows at Windows 8... adn barely accept the horrific state of Win 10. Im particularly pissed off about having system preference settings in at least 4 different places that are all counter intuitive.

I really liked windows 7, and was grumpy about first using Mac OS, but what won me over was that I didn't have to fight the OS in finding what I was looking for in settings.

Also the reason why I like Linux and BSD. there is a standard to where you find the setting files.


> Nobody considered that rearranging the layout and colors of the checkout buttons would have such an impact but they did.

I hope this was two decades ago because people absolutely should know by now that doing this has the potential for a huge impact.


This is a good example of why some people have zero business being anywhere near management. They win the birth lottery and have it easy all their lives by sheer luck, but when a decision has real consequences - this happens.

Anyone in the trenches could tell you that rolling out a huge change to a money-making project on a "round" date is suicidal. They just have no idea of what they are doing in the first place - just playing darts - and it usually works due to the ants killing themselves, to make it all happen. Because health insurance.


Was there ever a follow up to help determine why? Were repeat customers returning and panicked when they saw a new, and unexpected, layout change?


That is why I was hoping the article would have the conversion rate of the website before and after the redesign.


There is a graph at the bottom that shows the conversion before and after the redesign.


Thanks! Sorry I had missed that.


"But despite all the missteps and stress, the results might justify all the pain. I expected the new website to increase sales by 10-20%, but it’s been closer to 40%."

Should have put this in the beginning of the article.

As for all the issues mentioned in the article, trust me on this, it's always like this. I've been that "small client" hiring externals at all tiers: mechanical turk, freelancers, agencies.

You ask for A but get B. You agree on a timeline but none are respected. You can put your foot down but that does absolutely nothing, they don't need you. You're more like a hobby on the side.


Wow yeah I missed this. Read through most if it and was looking to see if anyone else had already commented my thinking:

“And the new design is WAY worse in every way!”

Honestly it’s hard to tell what it even is with the new design: SaaS product? Contract agency? Flight tracker?

I found the first design to be significantly clearer. I wonder how the author distinguished between revenue increase coming from natural growth vs. the redesign.


40% increase in sales is phenomenal. If that’s down to the redesign and not because of existing trends then it was money well spent


Yeah I'm a little surprised he regrets it. I guess he assumes that a freelancer would have done the same work since his takeaway says so. He may have indeed gotten "service normally reserved for large companies despite my limited budget".

My heart goes out to this guy ...

For those unaware, the blogger left a high-paying job at Google 4-years ago [1] to start his own business and over those last 4-years, unfortunately, hasn't really made any money/profit [2]

His blog is a treasure trove of insights and lessons learned along the way.

Highly recommend for others to read.

[1] https://mtlynch.io/why-i-quit-google/

[2] https://mtlynch.io/solo-developer-year-4/


Thanks for linking!


I find it hard to feel bad for him. Big N jobs are cushy but there's so much operational headache. Not that owning your own business is much better in that regard, but at least you feel like you're accomplishing something.


Wait what. He worked for Google and is not able to build a website like this himself?


I did a lot of Python and C++ at Google. I'd be able to implement a nice ETL pipeline for the website, but Google didn't pay me for my design skills.


I think OP intentionally offloads some work to others, to focus on his managerial role of his company. Not like he can't do it himself (after some learning processes) to certain degree, but it takes time and his time is more precious than potential financial cost.


What does being a google dev teach you about marketing?


Well he did build a website like this, but thought (incorrectly) that a professional designer/marketer (which is not what a Google engineer specializes in) could build something better.


Why incorrectly? Worked out well considering the resulting revenue increase.


As the owner of a digital product agency, this is really hard to read. It is such a shame that an agency would do this to you. I know you felt like they "did their best", but by setting you up with unrealistic expectations out of the gate they essentially guaranteed that everyone was going to walk away unhappy. Besides, when you go to an agency communication/transparency/team of experts is what you are paying for! You're paying for accountability! There really are good agencies out there that care about their customers and bend over backwards to deliver what they promise, but they are hard to find. For the size project you were looking at though, I do agree that a freelancer could have been a better option, but you'll run into some of the same challenges. Finding good freelancers can be just as hard as finding a good agency.


The agency shouldn't have taken the work to begin with -- the part where the lead admitted that he killed project governance entirely was a bit of a painful moment. Agencies are optimized to work at a specific scale, and it's risky for them to scale up or down for a specific project; in this case, their client fell through the cracks because they were using him as fillable hours and didn't ride herd on their designers. Considering they were working outside of the SoW, those were disputable invoices, but that's cold comfort when you don't have the free cash flow to take this to legal.


The difference in size between companies imply different operation trajectories. If you are their smallest customer, you are bound to be deprioritized and treated as second-class. If you are their biggest customer, you may well destroy them with procurement and governance processes.


Thanks for reading!

Yeah, I don't think the agency is blameless here, but I also don't think they're malicious or dishonest. I think they just overestimated their ability to scale down their workflows to a project smaller than their typical gig.


Well they know they've done an awful job, you know it too, but they happily kept the $45K. If they were as honest as you say, they would have refunded some of it.


That's exactly what does not leave my mind! It would have been a question of honour for them to stomach it and deliver close to the original estimate.

Everything else leaves me with a sore feeling, that it, while eventually not a scam, at least a careless, profiteering way to handle this, on the back of the customer. I hate this agency :-)


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: