Hacker News new | past | comments | ask | show | jobs | submit login
How we build an MVP on a budget with BizSpark and Azure (apirise.com)
52 points by geopsist on May 8, 2014 | hide | past | web | favorite | 28 comments

BizSpark is really great. In the article's case, they're not using Microsoft technologies, but in my case, I use Microsoft at work and I'm very comfortable with their tools.

For a side project, we'd started out with an unfamiliar stack (Grails + Postgre hosted on Amazon AWS), and kept running into roadblocks (not the tech stack's fault), but we missed some very nice enabling technologies with Microsoft software.

So, when we signed up with BizSpark, we started using Azure, and we're now leveraging the SQL geolocation capabilities (which Postgre also has), and our knowledge of .NET.

As the article says, you don't need to use the Microsoft stack, but if you're used to it, it's a HUGE enabler (we had discarded using Microsoft at first due to the perceived cost).

I'm a believer that, for an MVP, you have to keep to whichever technology you're comfortable with. In my case, it was the Microsoft stack.

Just be careful down the line. Those costs you are talking about are still there, they are only delayed because of BizSpark.

Having BizSpark did wonders to the company I'm in, but when the 3 years ended, there was a lot of hard work and even more tears because of it. BizSpark's entire reason for existence is to lock you into Microsoft products, especially their SQL server.

What were the problems you encountered?

As a potential BizSpark user my impression was that:

- after 3 years I get to keep using the licenses I acquired, leaving only hosting costs to be concerned about

- if I don't have enough capital/revenue to cover my hosting costs at that point it's highly likely my business is not worthwhile pursuing further anyway

- for my particular solution I don't envision needing much server power (and what's provided is pretty powerful) so additional licensing costs are not a problem (I can see this being an issue for other businesses)

The problem is that there is always going to be a new Windows Server release, plus a new SQL Server release. Even if you keep your licenses, are you going to stay on that version for the lifespan of the system? Likely not. Then you start to pay. For the database, the OS plus SQL Server Standard is still going to be around $15K (this was around 4-5 years ago when I was involved with this). Now this is just one instance. Multiple that by multiple servers and instances and you will reach upwards of 6 figures for your costs.

I guess the numbers work out differently for cloud but I don't imagine they would be that far off. For example, if my Azure costs are $10K a month, from that cost is likely $2K-$3K the Microsoft tax. Keep scaling your VMs and servers and that margin starts to look a bit less palatable.

If you know the limits of your business, then it's probably not an issue. I think the sweet spot would be to use Postgresql for the database and Microsoft for the rest, as most of the cost is SQL Server for systems that aren't in the full Microsoft eco-system (BizTalk, SharePoint,etc.)

Licensing costs are not really an issue when you're on Azure. If your app ran fine under the BizSpark program subsidy of $150/month of azure credit then when you're out of the program you'd just have to pick up the $150/month hosting cost, not suddenly have to pay $15k+ for a MSSQL or Windows Server license.

I know that. I didn't make it explicit in my comment. The Microsoft tax I mentioned with Azure is the licensing costs. Microsoft isn't just going to give out SQL Server for free and just charge for the VMs and bandwidth. There is premium pricing for SQL Server and BizTalk and custom licensing pricing for SharePoint and Oracle software.

Being on Azure doesn't mean you get a break on the licensing costs. They just amortize the cost over the hour the VM is up.

EDIT: Fair point on the hosting cost being known and acceptable right from the start. I also don't mean to imply that you are not aware of the licensing costs, just wanted to make it clear. For real businesses, at some point it makes more sense to build their own "private cloud" and at this point these issues become relevant but for most start-ups, it's likely just not going to be an issue.

While SQL Server has and will continue to be an excellent technical option, the costs can be a drag.

That said, the work various .net-related people have out or are working on have the very real availability or promise to lift that dependency, if not provide portability, with DB choice. See RavenDB (document DB) & biggy (Postgres 1st class). Supporting this for more traditional takes on relational store is that there is increasing experience in the .net world for applying coding patterns that can help projects pivot/abstract out the data tool dependencies.

If you have a sensitivity in this area, there are options which, when calculated with the ancillary benefits of applying existing .net experience and the great MS free + cloud tools, would pan out very well for many problem spaces.

Obviously, YMMV, but if you are doing a wide spectrum of cloud + mobile, Azure is very compelling and Xamarin can become nearly as large a burden cost-wise. That total burden still can be extremely competitive and the ramp for existing .net shops a pretty clear winner. But being a C# developer has never been so promising - and it wasn't ever that awful to begin with.

If anything, the problem of coming from the old MS-dev platform-planning and having to understand the range of options and optimizing for costs around those is the weak spot.

> are you going to stay on that version for the lifespan of the system? Likely not.

Not sure why you say that -- if it works, it works. I've just spent the last year working on a multi-million dollar project budget building a new system on a SQL Server 2005 cluster. Why? Because there was no need to budget for new hardware. (SQL Server 2008 R2 would have been a big step up, after that not really worthwhile upgrades, in our case.)

> For example, if my Azure costs are $10K a month, from that cost is likely $2K-$3K the Microsoft tax. Keep scaling your VMs and servers and that margin starts to look a bit less palatable.

Let's say my hosting costs are 10k p.m, 120k p.a. and represent 15% of revenue (this is just how my business works - if I need that much SQL power I'm doing a hell of a lot of business). 120/0.15 = 800k in revenue. Keeping things really simple, let's say my gross margin is 50% and corp tax is 20%. That leaves the business with a nice 320k profit. Sure I don't like spending 4.5% of my revenue on MS licensing, but at this point I can quit my job and port over to Postgresql (or more likely pay somebody else to). The point is that nobody wants to give away 4.5% but if it's the line between success and failure then you probably have bigger issues with your business model/environment.

Don't get me wrong - for my current side project I'd love to use Postgresql as the backend and .NET for the rest but the problem with Postgresql is two-fold and really specific to me:

1. I'd spend a lot of my time learning a new set of tools for zero benefit when time is my most expensive resource right now. I'm very familiar with T-SQL and the MSSQL platform (even simple things like having MERGE available has saved me lots of time given my problem domain, nevermind the ability to reuse a lot of code I've written and optimised over the past 10 years and knowledge of various issues and errors I will encounter). Porting in 3 years will be easier because we'll have a working system to port from, and the risk to the project will be less because we aren't doing two things at once (learn a new platform + solve a new problem -- only the later is going to make me money).

2. I plan on using some of the MS BI tools for the first time as part of this project -- this represents quite valuable experience to have as a .NET developer/solution designer; if I go the Postgresql route I will spend a lot of time exploring and evaluating open source options with zero benefit to my day job career progression over the next 3 years - I don't have the luxury of neglecting my day job while I work on side projects, so I need to be careful I don't lose value

Definitely all valid points. I don't have any side projects that can potentially turn into a full business at the moment, so I do have the time to learn. I'm focusing on Django and python for the moment. I think Postgresql is a great hybrid solution as you could live in both worlds (.NET and otherwise). Would be much harder to do with SQL Server.

I also felt a similar way in terms of wanting to keep my day job skills in alignment with my side projects. I think its actually the opposite. Picking up python and Django (or RoR) will make me much more valuable for my own work and otherwise. Microsoft certainly has its place for desktop applications but that's not my focus. Heck, even for that, I am seeing more applications in Adobe Air.

Also, it's just fun to learn these new stacks. I don't need to be an expert, just enough to get the job done and know when to ask for help. Another motivator for me is the ability to hire developers remotely to take on development. I would imagine this is going to be easier to do with an open-source stack versus Microsoft.

Don't get me wrong. Microsoft in the last 3-4 years is doing great work. If you happen to stay out of the Silverlight, WPF, SharePoint side of things it's not been all bad. ASP.NET MVC is now mature and there is enough momentum with nuget and open-source .NET projects that you can replace components if need be.

I'm fortunate to be on both sides of the fence. I don't intend to never use Microsoft tech for my side work but there is a lot to gain from seeing what's out there that isn't Microsoft.

a 150 GB business Azure SQL instance is $225/month, sounds like a steal to me.

You're right on. I am using Bizspark. The software you get is free to keep. When you start out it feels like hosting costs are expensive. But if you are making money (and money is the validator in my opinion) then it becomes negligible.

I thought about learning a new language when I started, but figured the time would be better spent learning the language/skills of sales and marketing. (My day job at a small profitable startup and my side SaaS product are both .net

If you are an MSDN subscriber, you also get discounts on the Azure cloud for testing and dev porposes http://azure.microsoft.com/en-us/pricing/member-offers/msdn-...

TLDR? 33% on VM's, 25% on cloud and then according your MSDN subscription 40 - 115 €/ month.

I used bizspark at my last job but didn't use Azure. We were on AWS, and it applied just fine to EC2 instances running Windows Server(you don't pay the additional costs for ms licensing). Only thing it didn't work on was the MSSQL RDS instances we had, we ended up running our own Mssql server. One thing to note, you can also get credits from AWS if you're a startup. I've seen $25k to $50k being thrown around and my current company received enough credits that i didn't need to pay for service for over a year.

Hey guys, Felix from Microsoft here. Some of you may know me as the guy who's working with YC companies. A quick PSA:

We sponsor YC companies with $60,000 in free Azure usage.

Get in touch with me at felix.rieseberg@microsoft.com if you have any questions!

We used BizSpark and a MSFT stack for our first iteration of the MVP (C#, ASP.NET MVC3, ServiceStack, etc.). It always raised eye-brows in a negative way when taking meetings with VCs and Angel investors. I'm not sure why, but they almost always asked a question "why MSFT? stack" and not XYZ that all their other companies were building in. I always felt like we should leave the tech stack off our slides... And they almostnever liked the answer I gave which was something along the lines "because I could build it fastest this way"

We've since transitioned into a Python/Java/Postrgres stack. With no real complaints.

I wonder, though, if there's any stats on companies not liking to aquire tech built on MSFT vs companies built with Java. I did a project once for one of the biggest names in enterprise software security, mostly a Java shop, and they had acquired a C#-based product, and I heard nothing but gripes from the tech team about that...

At the time, one thing annoying was licensing Bizspark to run SQL server on AWS. From what I understand, this is fixed now, but back then we had to pay for a SQL Server instance, despite being part of Bizspark plan.

Another thing that I wonder, if anti-MSFT bias is a regional thing. LIke if we were based in Seattle, would it be beneficial?

For the record, I still like a lot of the MSFT stack, and it's great to see them embracing some other technologies, like say Node.JS on Azure and feel they'll do a lot to promote some of these newer OSS technologies into the enterprise world.

EDIT: one other thing, we didn't have access to Azure at the time (it was just starting out, and I don't think it was part of the program). But Azure as a hosting option for managing the MSFT server licenses and such is a pretty enticing option.

Thanks for writing this. A small feedback: "The power of transformation over APIs... without the hassle." does not really convey what you do to a potential customer. People have very short attention spans and you have only a few seconds at most to convey what your service does. Most people will not scroll down to read more otherwise.

I agree. No idea what they do based on that description.

Same here. Thank you for the feedback. We will try to make it clearer! :)

what we do is to offer a configurable proxy for APIs where someone can configure various aspects of them, e.g. transform and filter request & response bodies, headers etc, simplify and accelerate the work a developer has to do in order to interact with APIs and integrate them with her apps. leting us know that we dont communicate well what we do is probably the most valuable feedback we can get at this point and it is much appreciated!!!

thanks a lot for your feedback! We are trying to make it better. Got your point ;)

PS. I hope the post info will prove useful :)

I'm also using Bizspark and the fact that you can do everything with a 115 € (Europe) monthly budget is probably not well-known :)

Yep this was my original thought. Thank you for the reply :)

That's cool, but you could also build an MVP using PHP and MySQL on cheap shared hosting or a cheap $5/month Digital Ocean instance. Unless you are on the Microsoft stack, you probably aren't saving a ton of money that you couldn't save by making different decisions.

Probably the more valuable opportunity in BizSpark would be to get any extra marketing out of the relationship with Microsoft. Being a "success story" can get you press, links, etc. which is very much more valuable than saving money on hosting.

You'd still be saving money, you can run PHP and MySQL on Azure and get enough server power to host your production app for free instead of paying a shared hosting company $5 for something only sufficient for development.

yes I agree

I agree... There may be many solutions available! In our case we found the best value for money case.

A bit of topic but I can't seem access apirise.com from Turkey. Connecting via VPN works.

this is strange, we'll have a look at it although you should not have any problems accessing it from anywhere. Thanks for letting us know though.

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