Failed SaaS Post-Mortem: Validate Your Pricing Model (Like I Didn't)

April 30, 2020

I just spent about three weeks bootstrapping a SaaS as a solo founder. For context, you can read my description of it here, and I’ve talked about the inspiration behind it and motivations here and here.

The basic concept was solid: provide an insanely simple way for business intelligence consultants to manage project data.

Spoiler alert: it failed miserably, despite the fact that I did a lot of things right:

  • I came up with an idea as a hypothetical solution to a problem, instead of the other way around (you can read my post on coming up with SaaS ideas to understand why this is important)
  • I built this solution with a specific end user in mind, a la Paul Graham’s legendary essay Do Things That Don’t Scale. Specifically, I built it for my dad, who is himself a BI consultant on the UX side of things and actually helped me design and develop the tool
  • I built a quick and lean prototype (front end only) in just seven days, thereby giving myself emotional space to analyze the idea objectively without getting too attached to it. This was a TON of work, by the way — it was a 90 hour week, and the resulting software was around 3,000 lines of code (you can play with it here).
  • After building the POC, I approached people in my target market about it by messaging around thirty BI engineers and consultants on LinkedIn. I was able to have a number of conversations about it to clarify my messaging and understand the market’s need better.
  • I even came up with a sick-awesome name and domain: Bicycle (B I for Business Intelligence, and “bicycle” referencing Steve Jobs’ quote that a computer should be a bicycle for the mind).

I was so excited at the start of this project, and I really had a GREAT time building it out and fantasizing about how it would change my life.

But I got one thing wrong, and it doomed my project from the very beginning, despite those auspicious yet humble beginnings:

I failed to validate my pricing model before I began.

The exact moment I realized this was while I watched Jason Cohen’s talk on the formula for bootstrapping a successful SaaS. Jason’s a funny guy and he has a great speaking voice, but neither of those things prevented the point he made in that talk from making me realize I had to ditch my SaaS.

So, let’s dive in. I really hope this experience and post will prevent you from making the same mistake that I did. Killing a project is not fun.

cover image

What was so horribly wrong with my pricing?

Given that I had just put a bunch of time and effort into building some totally awesome software, why did I decide to quit working on it, cold turkey?

It all came down to some pesky numbers.

My initial plan was to charge around $20/mo per customer, where my ideal customer is anybody involved in a BI project (well, it was a little more sophisticated than that, but you can check out my description of Bicycle to learn more). In my ideal world, they’d whip out their wallet and fork over the cash on the spot. Then they’d expense the monthly bill to their company and presto — paying client!

This seemed fine for me and was “good enough” in my mind to justify spending the time and effort that I did without digging in further.

But then Jason’s video rocked my world: he said you needed to be charging at least $50, closer to $100, per month, per user, in order to make it as a bootstrapped solo SaaS founder.

I mean, if you charge less you might be able to hobble along, drowning in support requests and working 90 hrs/wk, but you’ll never really hit escape velocity, and that’s not really what I’m after. I’m after freedom, not just a different and more demanding boss.

I won’t go into all the math here, as he’s done that brilliantly already in his talk, but needless to say, I was convinced.

Problem is, you won’t find many BI developers who are keen on trying to expense a monthly $75 bill. Most employees don’t have the kind of expense bandwidth to do something like that.

Well, I did try to pivot… key word: try

I realized this pretty quickly, and decided to try a pivot: maybe I could sell the software directly to businesses who do BI consulting, and let them provide it for their employees?

But after a bit more digging, it became apparent that this really looks a lot like enterprise sales (because it, uh, IS enterprise sales), which is a no-no for first-time or self-funded solo entrepreneurs. Takes way too long to close a sale — enterprise sales cycles often run twelve to eighteen months.

If I went that route, there’s virtually no way I’d be able to iterate quickly enough on my product with a feedback loop that long and no runway. I’d have to wait at least a year before my first sale, minimum — and that’s assuming I got everything right the first time around. And given my background in enterprise sales (or more specifically, my lack thereof), the odds of that happening are roughly equivalent to a monkey sitting down at a typewriter and bashing out Shakespeare’s Hamlet in one go.

Also, it turns out that the other tools in my market charge absurdly low amounts of money, anywhere from $3/user/mo to $10/user/mo, and those other tools are often addons to existing enterprise software which goes for a lot more.

Economies of scale are much friendlier with those larger companies, because they can afford to spend a lot on marketing and even take a loss on the product if it increases their main offering’s value-add.

There was no way I could compete with that.

Plus, my goal is to hit $2k MRR in the next two months, so if I persisted in the enterprise sales route, I’d have to totally throw that goal out.

So, after having eliminated BI firm employees, and the BI firms themselves, the only remaining target market was independent business intelligence consultants, who are both rare and also don’t have a need for this tool (much of this is confirmed by my market research).

Given all of that, I decided to pull the plug on Bicycle.

I just wish I had known these numbers before I built the darn thing.

How can you avoid making the same mistake?

I didn’t think through my pricing model thoroughly before beginning, nor did I know what my target numbers should be. My pricing model was as follows: find “a bunch” of users for $10 or $20/mo. Ish. And hope it just worked out.

At that rate, I’d need to get between 500 and 1,000 users in order to reach an MRR that would let me go full-time. With that many users, as a solo bootstrapper, I’d be spending all of my time on support, instead of marketing, product development, and learning to play Bach’s Well-Tempered Clavier on the Yamaha P-515 I just ordered.

I know now that I probably need around 150 users paying an average of $70/mo to land ~$10k/mo in revenue, without drowning in support requests.

I just didn’t do the math before I started writing code.

And it cost me!

So, while you’re vetting your next startup idea, include pricing as a critical success metric from the very beginning. You can have a great idea, even a valid problem and solution, but if the pricing doesn’t work, it’s worthless.

Hope this helps!

By the way, if this was interesting or valuable to you, you can follow my journey on Twitter. My goal is to hit $2k MRR in the next two months, before my first child is born in July.