Carly Buxton, PhD

Even though I’d read all of the “lean startup” fodder, I got something majorly wrong about building my first tech MVP. Maybe if I write about it here, other founders will save themselves from making the same mistake.

Here’s what I’d advise, as you consider your MVP. Interestingly enough, I KNEW these principles when I started Nessle, but there’s a key nuance I got wrong, which I’ll explain.

STEP 1. Distill your product’s #1 “job to be done.” You have a product idea. So what, ultimately, are you expecting a customer to pay you to do? If you have 2+ customer sides / a multi-sided marketplace, etc., just start with the customer segment you think is ultimately PAYING the money for what you’re putting into the world. You may need to tackle the other side(s) later.

STEP 2. Build the hackiest, lowest fidelity way of completing the task your software (or product) is designed to complete. It may be a Google Sheet. It may be a “done for you” approach that looks like tech but is really just smoke and mirrors of a website while you do the work behind the scenes. It is not a coded website. Or at least, if it is, it took you like two seconds and you used AI and spent less than $10. See: GPT Engineer, Replit, etc.

STEP 3. Now, go try to sell that thing. Don’t tell people about the “idea” of it being in development, etc. Tell them that the service itself–the actual JOB TO BE DONE–is for sale, and you are taking money at this very moment for it. Do they give you their money?

I misunderstood this. I thought that because I was building a tech platform, I needed to have an MVP version of a tech platform. And that if it didn’t exist on the internet–and look halfway good and “do” certain things–people wouldn’t recognize it as a tech platform, so of course it wouldn’t grow. In short, I wrongly assumed that my MVP would need to be several stages further along to get any sort of “real” sense for whether people would pay money for it. So we built a tech platform.

In the early days, I lied to myself: of course it’s not taking off yet. That’s not because it’s not a rocketship idea. It’s because: it’s not pretty yet, or fully formed yet, or HIPAA-compliant yet, or enough people just haven’t heard about it yet.

The truth is: if you’re trying to build a startup beyond any kind of “lifestyle business” scale, ie., something that you want people to back with capital, it WILL need to be the sort of thing that people pay you for, even if it’s janky and hacky and awful. It will start to spread, to spike, to take off on its own… even without any advertising spend, and without a pretty interface (case in point: chatGPT). If it doesn’t, then sure, you may still have a backable, sellable, profitable business someday, but you’ll likely pivot a ton–and waste money and time along the way–if these signs aren’t appearing in the earliest days of janky MVP.

I could have saved myself SO much time and money on the front end of what we built as Nessle 1.0 if I had just focused on ONE side: the (paying) suppliers, and their ONE chief reason for paying to be a part of what we were selling: a way to make more money as a parent-support entrepreneur. (This, ultimately, is what our company Parentswarm has become, many a pivot later).

My MVP could have looked like this: a PDF called the Nessle “fill your client calendar” service (or something like that). A pretty little, Canva-designed flyer (or wix splash page) that says something like: “I’ll fill the dates/times you want next month with clients if you give me 20% of what you make.” OR “Join this Doulas-in-Business support community for $10/month” OR “Pay me $199 for a 2-hour strategy session where we audit your business website, rewrite your bio, and create a product/service strategy plan that gets you more revenue.” And then a Stripe checkout button. I could have tested the would-be customers’ appetite for the actual thing I wanted to sell them. I could have tested price sensitivity. I could have learned about their objections. I could have seen that this target market may SAY that they want regular virtual events but that they can never actually come to them (because they’re busy).

SO much learning ultimately had to happen years and tens of thousands of dollars down the line, because I convinced myself that these things had to happen on an actual tech platform.

Fortunately, I learned my lesson. I *did* follow my own janky rules with last year’s Nessle CareSheet system. Michelle and I had started talking about building a feature into the Nessle software that would do something people always SAID they’d pay me to do for them: get them childcare coverage when they needed it. Having learned the hard way about MVPs from several Nessle pivots, I decided to test the payment appetite for this first. 

I built a hacky system of babysitter recruitment and shift notification via an integration of tools from Mailchimp, Canva, and Google Sheets. (This is actually how I organize my own babysitters in real life, and it’s magical). I offered to set this up for people for $19: Pay me $19 and I’ll set up a system for you that ensures you have babysitters when you need them for the rest of your parenting life. I had ZERO TAKERS.

And so I learned that CareSheet could NOT be a business, and it would not be worth our time to build it as a feature on our platform. At least not now. Because no one bought it. Even without a massive audience learning about it, even without the “right” marketing language, even without the “real” tech features working, if there were a “there there,” somebody would have spent $19 to have their babysitter shifts covered for the rest of their lives. It’s just not the right solution–or the right job-to-be-done, or the right price, or the right method of delivery. And in this example, my next step would be to dive into these details, to talk to the target customer and find out why they didn’t buy it and iterate from there.

In sum, to all of you founders out there: if you want to build a tech platform and are convincing yourself that you can’t “release” your MVP until it actually looks something like a tech platform–or if you’ve released a non-tech MVP and it’s not doing much (and you’re making excuses about that), you are falling into the same trap that I fell into. Keep everything as low-fidelity as possible and ask for people to pay for it. And if they don’t, pivot now.