Tracking feature requests is a huge pain for most companies. A few pieces of feedback after dealing with this for years:
The Customer Support teams of your customers are going to be one of the main channels for capturing feature requests. We use Zendesk and have custom fields for categorizing and sub-categorizing requests along with a brief description. We do this so that the PMs don't have to read the entire Support ticket if one sentence is about the feature request out of 10 pages of text. If you could build an integration with Zendesk and other major ticketing tools that tie support tickets directly into your product it'd be an easy way to get your customers to have a huge influx of important data.
Sales will also provide lots of feature requests as they are also talking to many customers. You have to make this as seamless and as easy as possible otherwise they won't use it. Setting up a Salesforce integration is probably your only way to do that.
Tracking feature requests for strategic customers need to be treated differently than for other customers. There needs to be a system to categorize tickets by customer cohorts. If you can tie ARR, deal-breakers or other metrics to feature requests, that will help PMs prioritize them.
For your customers, publicly exposing feature requests is a double-edged sword. You're going to capture a lot of feedback. Some of it is good, some is bad. Not taking action on customer feature requests can build negative sentiment in the community.
Hey, thanks so much for this. Zendesk + intercom + salesforce integrations, plus segmentation (by MRR, etc) are all on the roadmap.
Feedback that internal teams track (opposed to users voting) is kept internal, so users can't see stuff that sales/support adds. A few companies are using us entirely internally.
The internal use case is definitely interesting but then we lose the nice organic channel.
Thanks again for your insight! Really appreciate it. Say hi on our intercom chat sometime.
Nice. BTW, I do not know what you guys are selling (the home page did not tell me much - you may want to improve on that), but the blog post itself was kind of engaging.
Nice story - $100 in December, then we do this, then $100 in March, then we dont cry and despair but do that, then $3,500, "but how do you possibly live on $3,500/mo?" - answered right away.
So I thought it was a nice read. Also, good luck :)
Nice product, just gave it a quick try. Some off the cuff thoughts,
- Consider a Zapier integration, might save you some time and make delivery/adoption faster
- In step 3 of onboarding you ask me to share a url with team members. Can I make this url private/sso? If yes, make the discovery of that immediate. If no, that's going to be a blocker for any internal adoption.
- PMs tend to organize their internal priorities and roadmaps in specific ways. I like that I can sort by "coming soon" "done", etc, in your list BUT, what I really want is a consistent visual delivery, this seems to usually be a kanban style board.
- Internal teams and stakeholders need to understand why what you're working on right now is the single most important thing that the teams need to be working on. It's important that this assessment criteria is clearly communicated, visually, and understand.
Interesting, we actually consciously didn't promote because of your interview with Eoghan. He said the more you try to make content convert, the worse it does.
The blog post wasn't about us getting customers, it was about helping people startup.
We still got 25% of blog traffic to our homepage anyway. Would you still make the change?
I'm with you. If I read a blog and am interested in the product, please don't make me work to find it. I expect that clicking on the logo will take me to the product website.
THE Andrew Warner is giving you product advice.... listen up!
One of the primary goals of any new startup should be to get Andrew Warner to notice you.
I agree with the comment re logo... it just drives me crazy that people don't understand the mechanism of blog post lures people to the product site via the logo link.
Tip: add a link somewhere on your blog to https://canny.io. I had to manually enter the URL in a new tab to look at your site. For some strange reason many blogs forget to include a link to the Real Thing.
Lol yes. I've often seen the logo on a blog post link to the blog homepage rather than the product homepage. Sort of defeats the purpose of the blog post (which is presumably to drive traffic to the product)!
On the contrary, if I'm done reading a blog post and I want to read more from that blog, it infuriates me to no end when clicking the logo on the top sends me to a product landing page and I have to hunt around to find my way back to the blog.
Also, make sure that there is a clickable home page link easily visible in any email you send, whether promotional or merely informative. Even places like Amazon and eBay get this wrong in some contexts.
Sales funnel feedback: I want to know what this will look like on my website to my users. That screenshot on the top isn't quite enough--I'm not sure which parts of that are my website and which are your widget. Is your widget always a big brick like that, or does it collapse into a little icon (like intercom + others)? Why not just put the actual widget on your homepage?
So I started the trial setup, but I bounced at step 1/3. Didn't want to write 5 feature ideas. And I still don't know what it'll look like on my site.
(Small side thing: I wanted to use the sample "board" name that you gave, but I couldn't just hit enter--I had to re-type it into the text field.)
Good luck! Looks interesting, from what I can understand so far.
I know it's my fault and probably obvious but that site doesn't load on Chrome if you have disabled third-party cookies. You might want to show an informative message instead of an infinite spinner :(
Could you build a "show me what this would look like on my site" feature that takes their site, and fakes a version on your server (maybe in an iframe or using javascript)?
I have raised money for a company and I have boostrapped a company. Both have their perks, but bootstrapping makes me feel so damn proud. Here are my numbers: https://i.imgur.com/rm5ChOD.png
Solo founder. Saas product. 70k users all word of mouth.
We use ChartMogul at Canny. It's a huge relief not to have to keep track of customers + revenue + churn in a spreadsheet. And it's FREE for under $10k MRR. Such a no brainer.
I cannot stress enough how much I enjoy your product. When you offered the freemium model, I was thrilled. Happy to convert to paying once I cross that $10k MRR mark!
Community marketplace for amateur YouTube talent. I'd been an amateur voice actor for 10 years so I had a lot of contacts and felt like I was intimate enough with the industry to build something they wanted.
It's been awesome to follow Canny's story from the start. The founders faced incredible skepticism from "experts" early on but they soldiered on and against all odds they're making it!
Repl.it has been one of the earliest users of the product and we couldn't be happier. It's embedded here: https://repl.it/feedback
Quick piece of feedback (and yea, I'll submit it), is that when I go to repl.it/feedback, the only visual clue that I am on the feedback page is the URL. Maybe it should say "REPL.IT Feedback" in the header, as a visual clue?
Congrats, that's awesome! I think $3,500 per month might be ramen-profitable in the Bay Area, but that's pretty much retirement money if you're living in Spain or Thailand.
Keep in mind we have 2 co-founders, so this is about $10/hr working full time. Definitely not a crazy amount of money, just a big milestone for our startup.
Such a humane way to describe your startup experience. I really enjoyed reading it!
I also noticed in a linked article that you are using Notion. (I am the co-founder). Happy to give you guys a free account to help you bootstrap. You should receive a message from us shortly ;-)
Interesting post! About living nomad, is there any pain points regarding the management of the company from anywhere?
I read the post linked[1] in this blog post which talks about day to day life, even if work-life balance can be tough, it must be awesome for you guys! But what about administrative paperwork or hurdles (IRS, legal mail, and whatnot)? Was there anything different due to the fact that you guys are digital nomads?
Superb write-up Andrew,
I would recommend you to read this awesome book called Traction, it has definitive ways to market any digital product.
Would love to read more about your journey as you progress, keep us posted.
Happy Canny user here. Do you have any plans to introduce a free tier (or almost free) plan? We have a low amount of traffic and compared to the other services we currently get for <$10 a month (hosting, auth0, sentry, github) I'm struggling to justify the cost we are currently spending on Canny.
Offering a free tier, just invites freeloaders, who are THE WORST end users to deal with. They are the most demanding, and will tie up valuable support time. Do not do free anything for SaaS. If you're product is good enough, make them pay for it.
How long will have you been in Valencia/will you stay around? I can give some tips if you are interested, both about just nice living/food places or the tech community: public@francisco.io
We've been here for a few weeks, and are leaving in a week. Headed back to North America for an undetermined amount of time. We'll definitely be back though! Valencia has been great.
This cafe, Cafe ArtySana, is one of my favorite spots to work in the whole world! Great coffee + food, great wifi, plenty of outlets, and they don't mind if you work all day.
awesome I'm in seville getting my own product of the ground http://documize.com, if you every come around it would be great to catch up , or if you make it to the WebSummit in Lisbon in November I'd love to trade notes with you, or any other startup that goes there.
Awesome growth graphs and congrats on the early success
I'm guessing by the result I obtained trying to read your post (see below), that somewhere in your article there's a paragraph about running the entire operation on a single 2GB DigitalOcean Instance.
Funny how I have been recently thinking we will need a way to open up feature requests to our customers for a bootstrapped project I'm working on. We can't take the $50/mo. rate just yet. Canny will be in when we get there.
We're probably not ready for this just yet. We're currently in a good place with a) solving a problem we have and b) a couple dozen paying customers who are helping us see the cracks in our MVP. Some of whom we can work with face-to-face. That's huge considering our target market isn't the best at articulating the challenges they are facing, so watching them use the app is a great insight vector. It's also really motivating to hang out with people who are excited about using our product.
Thank you, and now that I can read it -- nice post. I like how you covered a lot of the "wow that didn't work at all" stages. You didn't downplay it, which I see a lot of posts of this sort do.
Thanks! It was 1.5 years of making no money, so it's definitely been a grind... Feels good that something's working, and would love to save others the trouble we went through!
Congratulations on your early success and thanks for sharing your story. It's inspiring and I love your storytelling style. I don't think people realize that it takes awhile to get everything started and your timeline/decisions clearly show that.
It's also a great product for us but as a fellow bootstrapper, the price range is out of our reach at the moment. We will be looking into when get more scale.
We're building a simple CRM tool that tracks calls, meetings and all kinds of details for companies. It would be interesting if you already have an API available that we can integrate with.
If you're ever in Edinburgh/Glasgow, be happy to take you and your significant other out for a coffee/beer.
Yeah! Literally 18 months of making $0. Such a grind. At some point you just decide you're tired of making $0 and do anything and everything to not do that anymore.
Reach out on our Intercom and we may be able to make something work.
Would love to visit Scotland! Probably next year when it's warmer again. :)
S!=aS. I can't think of many instances where I'd pay a recurring fee for the use of a piece of code. Code to me is an end-product, meaning, I buy it then I own it in perpetuity. I still have my copy of XP from the Paleozoic era but it's still mine and if I choose, I could spin it up. I can only think of two instances where I'd agree S[is]aS; signing up for a full class, and buying tickets. I thought both of these were clever ideas but I'd still be hard pressed to agree to pay anybody else for this "service."
That said, hat's off to those of you who make this work.
Is the corporation listed as a C-Corp? Doesn't the business have to pay taxes on profits, then pay employment taxes on what it pays to employees as well, on top of what the employee must pay?
C corps pay taxes on operating profits which are gross revenues (that is, revenue less cost of goods sold, which for software companies is usually 0) minus operating expenses. Operating expenses include salaried employees. Then you take out interest from any debt you have and pay a percentage on that towards taxes, to get your net profit. Net profit is the amount available to actually take home for shareholders. Once you make your distributions to shareholders, you are left with retained earnings.
No, I mean the fee you need to pay on salaries. 30% in my country. So if you pay $100 gross salary you need to pay an additional $30 to the state. On the $100 the employee will pay normal taxes but the cost for the company will be $130. I'm presuming they "take out" the money as salaries here. Or is it done by dividends? 15% sounds low for that since you, in my country, need to pay corporate tax first and then taxes on the actual dividends themselves.
Chrome says your site is using SFUI (which is beautiful) so I did some digging, it looks like that is the Apple San Francisco font, how the hell did you get a license for that as web font?.
This is very meta but I want to post a feature request to Canny about allowing anonymous features requests but I can't since I don't want to create an account (hence the request).
A lot of our customers are SaaS companies. To them, a feature request is worthless if they don't know who gave it. We don't plan on supporting anonymous feedback until we expand our focus to consumer apps.
Also, if you don't leave your email, how will you know once the feature is built?
Well, I mean, an article entitled "How we started a business" could involve how they raised money. Given that this is HackerNews, it's maybe even more likely than not. The term "bootstrapping" precludes that. And "business" could mean anything from a desktop app to a physical t-shirt factory to a copper-mining concern. "SaaS Startup" means something much more specific. And yes, they could have said "... to enough profitability to cover our living expenses," but "ramen profitable" seems more concise in a context (like HN) where the audience knows what it means.
I'm all for calling out buzzwords, but each word here seems to do a decent job of communicating a lot more than "How we started a business."
For the average person, meaning someone getting a business loan or mortgage, this just isn't true. Take a look at historical bank rates. They're much cheaper now than the 80's.
My recollection of the 80's and 90's was that almost all businesses required a solid chunk of cash to start.
The idea of bootstrapping is only valid in recent 1-2 decades because it is now possible to start businesses without a chunk of cash, thus "bootstrapping".
I'm not convinced bootstrapping existing under another name in the 80s.
For the pedants, I'm not saying it was impossible to start a business in the 80s and 90s without cash, just saying that such a path was rare enough to not have a common name such as "bootstrapping".
In fact, prior to Amazon ec2, even Internet businesses required typically the cash to buy/house a server in a data centre which was often thousands of dollars, and even that was considered incredibly cheap for access to a global market compared to what it used to cost prior to the web.
The definition of bootstrapping includes saving up ones own money to start a business, generally from a previous or other job. Most companies are started that way. Venture is the exception, not the rule.
And prior to EC2 a lot of people either just rented servers (in someone elses colo), or got a leased line and hosted servers in their office. Both were very, very common in the 90's.
Sorry, you are associating aws and ec2 with the advent of non capex intensive web hosting or dedicated servers. Prior to ec2 there were plenty of cost effective hosting options beg rackspace, rackshack, softlayer, digitalnation, verio, theplanet, serverbeach, and many many others.
There was pretty cheap competitive web hosting available from 2000+, not just 2006 when ec2 was released.
You are wrong. Many, many engineering companies started without requiring a solid chunk of cash to start.
In Germany it is quite common to start that way even today. Many startup folks consider this old fashioned. But it is the most engaging and valuable way to start a companies. Because then your about skills and providing real value and not for that fast money with a fast exit. You don't want an exit at all. 3 friends of mine started in the 90s a tools and services companies for the automotive world. They are now all 3 retired and moved to company to a trust fund so that no other company can ever own that, so that this company can exists for ever. Compared to the US. Remember DEC, HP, SUN, Silicon Graphics, and alike? All companies got sold at on time to another and basically disappeared with that.
Yes, "bootstrapping" is the old fashioned way in the old fashioned world to start companies.
> In fact, prior to Amazon ec2, even Internet businesses required typically the cash to buy/house a server in a data centre which was often thousands of dollars
You seem to imply that shared servers (shell or web) did not exist prior to AWS?
For the stuff that I was building, we always needed enough control of the server to need our own. Perhaps people were building "real web applications" on shared PHP/CGI hosts but my impression was that sort of thing was used for lightweight websites with a submission form.
I do recall that there was typically shared PHP hosting.
ec-2 changed the game by making complete control over your own server simple, cheap, software controlled.
Shared hosting existing before but the ec2 formula was the one that worked.
I think you may have missed that you could rent servers online for a long time. Dedicated or shared, full control and much cheaper than EC2.
As far as I know EC2 introduced auto scaling, load balancing and spinning up instances in such a lightweight way. Not a revolution but still nice and necessary steps.
EC2 offers scaling up and down the number of instances as you grow or shrink.
If you built a website 20 years ago and it exploded in traffic, shared hosting without easy scaling would not be a solution. You would have to plan your growth and pre-order shared instances. Thus making your business cash dependent again without proven growth.
You really think business grows so fast that you need instant servers?
Google ran for many years on piles of commodity boards and they were expanding pretty fast...
The amount of marketing hype around cloud services is insane because all the big tech companies have a vested interest in your renting their platforms. It doesn't allow you to do anything you couldn't 15 years ago, just makes it easy.
Just makes it cheaper to get started and automate. But compared to setting up a server and running something on it, modern cloud platforms are about as complex IMO (to do right at least).
True, but it would still take more time/effort/knowledge to spin up your servers on commodity boards.
Plus, the level of resources for configuring EC2-like servers, and Heroku-esque preconfigured servers reduce the barrier to entry pretty significantly.
I think you over estimate how many people were online 20 years ago and what the expectations were.
You could host a reasonably successful web site on a relatively low end server (say, a Sparc 10) off a T1 line (1.5 megabits) and serve a ton of users. They all had crappy connections and were used to waiting.
AWS was the first place where a joe-developer could dial up a machine to use without a heavy sales process and involving purchasing, legal etc.
This all played out slowly though : at first AWS was only used for things like running tests, replacing developer desktop hosted VMWare. It took a few years to get to the point where it was used in the way we are familiar with today.
fwiw when AWS (EC2 really) started, Internet/Web services weren't even typically run on x86 hardware or Linux. Solaris, AIX, HPUX et al ruled.
This is just not really true. Shared hosting and leased servers were widely available from dozens of providers and very popular back in 2002 when AWS started. Also, FOSS operating systems were very popular for servers back then as well. Sure, bigger players owned lots of servers they sourced from the bigger vendors, but smaller players and upstarts largely did not (and certainly didn't need to).
You are correct. Somehow my brain transported the launch of EC2 back to around 2000. I _think_ I was conflating it with a similar service that was provided by a company I worked for a the time. Since we were doing on-demand machine provisioning then, I guess over the years my brain has assumed Amazon must have been too. fwiw EC2 wasn't launched as a beta until 2006.
Absolutely not true. I remember getting servers from one of the colo providers in the early 2000's (The Planet, which became EV1 Servers, I think?) This was all available with a credit card, no legal or purchasing department required.
AWS didn't launch until 2006, which was actually pretty late in the internet game. x86 Linux hosting was ubiquitous then. If you're talking Solaris, AIX, and HPUX, you're remembering the late 90's...
You're correct in some ways but, I think, missing a few key points. For roughly 15 years before AWS, inexpensive shared hosting was available.
From a bootstrap perspective, there was little cost difference at the beginning. The main difference was in the middle stage: scaling. After a bit off success it was still possible and practical to scale until you outgrew the ability of a single dedicated server to handle a specific task. Until then, you could grow and scale with little difficulty from a shared server, then dedicated, then shunting functionality to individual dedicated servers, e.g., web sever, database, etc. Yes these cost money, but ig you were bootstrapping then you only got to that point after you'd passed through the nearly free-per-month shared stage and, abd long after revenue flow began.
Scaling beyond that was absolutely way more difficult in terms of tech knowledge and cost. Modern AWS makes that scaling much easier, but the pre-ramen profitable stage was not much if any more expensive.
The key cost that was then and still is the biggest, in my opinion, is opportunity cost. A hundred bucks a month then or now, with or without adjustment for inflation, is nothing compared to the opportunity cost.
As a more general rebuttal of the sentimemt expressed, I'd offer up "started in my garage" businesses of all sorts, then and now. Those whose primary initial input is never cash but sweat equity and grinding out the details and execution of an idea.
Plenty of business started out of a garage in the 80's and earlier, many of them are well known names today like apple. A lot of video game companies grew from bedroom coders. Software could be sold through direct mail order. In the 90's a lot of companies bootstrapped themselves (the term was in existence) by building access apps for businesses or building small bits of custom software for large local business.
If anything it's probably harder to get started today, your competition is worldwide, not just local and customer expect more than a text based terminal app.
The joke is that "start small and keep expenses low; focus on making enough money to live" was known in the past (and outside of the tech bubble) as simply "starting a business".
The title is a bit buzzword-laden, and GP is calling that out.
Not much of a joke that I think.. For some VC's have a stereotype of looking down on ramen profitability because it gets in the way of convincing young people to burn / spitball / "innovate" a decade of their lives away their 20's swinging for low percentage 100x returns in exchange for burn out, etc... The alternative of say, 10x, can be very profitable and stable and enable just as much.
Said differently..
For some, largely on the funded side, everyone's a statistic, and after generally "going hard", they inevitably "go home" because they're told there's no more money from investors, and not enough paying customers.
On the self-funded / lean / boot-strapped / ramen profitable side, you get to earn your keep to stay in business so no one can tell you to go home except your customers who stop paying you.
I'm not sure there was a "joke". Basically GP's saying that the marketing speak laden title is an artifact of our times. Really what does all those words mean that "we built a business" doesn't communicate?
The Customer Support teams of your customers are going to be one of the main channels for capturing feature requests. We use Zendesk and have custom fields for categorizing and sub-categorizing requests along with a brief description. We do this so that the PMs don't have to read the entire Support ticket if one sentence is about the feature request out of 10 pages of text. If you could build an integration with Zendesk and other major ticketing tools that tie support tickets directly into your product it'd be an easy way to get your customers to have a huge influx of important data.
Sales will also provide lots of feature requests as they are also talking to many customers. You have to make this as seamless and as easy as possible otherwise they won't use it. Setting up a Salesforce integration is probably your only way to do that.
Tracking feature requests for strategic customers need to be treated differently than for other customers. There needs to be a system to categorize tickets by customer cohorts. If you can tie ARR, deal-breakers or other metrics to feature requests, that will help PMs prioritize them.
For your customers, publicly exposing feature requests is a double-edged sword. You're going to capture a lot of feedback. Some of it is good, some is bad. Not taking action on customer feature requests can build negative sentiment in the community.