Looking for a market - Start (Part 1)

Introduction

Alright, nearly 3 years after, I’ve had my learnings from the Airtable marketplace and how it is to build apps on it. Long story short, the lead gen there is pretty amazing, but the ability to actually build anything useful on there is severely hindered. And if what you can build on it isn’t much of a thing to write home about, then you also cannot make any serious money on it. Some people have done it, but they’ve been there from the very start, capitalizing on the biggest pain point.

Now that I’ve saved up some runway, it’s time for me to do something fresh and exciting. Because I have the know-how to build full-blown SaaS, then I thought it’s probably best not to limit myself to marketplaces anymore. If a marketplace can serve it as an extension, then that’s amazing, but it won’t be the starting point anymore. At least until I learn of a marketplace where it makes sense and does not limit me this strongly.

Process of generating ideas

There’s a lot of common wisdom around how you find out what you should build as a business in the first place. The most obnoxious sitting-on-a-high-horse idea is that you shouldn’t build one unless there’s already people screaming at you, and throwing money at you to make the thing. That, of course, isn’t the reality almost any of us live in. And if it is, you’ve likely been blind to that opportunity for way too long anyway.

If you put that same idea on an after-school programme, with some rigorous diet and exercise, what you’re left with is this: build something people need. This, in turn, has the disease that almost all problems are already solved by pretty good solutions. This holds true to almost all ideas at the scale that you can make the MVP happen at.

What I took from this is that as long as I build something that already exists in my own way, then with enough time, I can talk to my customers, and iterate towards solving their even deeper needs. Basically having the product would give me a window into their problem spaces that’s adjacent to that product itself. I thought that if I get my first set of customers, I can just double down on that, and move the product closer to the thing they actually need. I just need to prove that I can get these first customers.

I had a lot of thoughts of what kind of product was it that I would actually love to work on. I had a few aspects that I cared about more, one of which was the focus on being able to work on beautiful things. Based on multiple of these kind of properties that were important to me, I simply picked one. It was something I thought I would be able to work on for a long time without getting tired. A proposal software for small agencies was born: ProposalKit.io, knowing very well that this was just one of many plausible options, and I might have to pick again.

Process of validating ideas

So what is validation, really? It’s really whatever you think it is, there’s no authority on what’s “valid” other than the test you yourself write for it. For me, it was that very threshold of having an initial set of customers. Not “a great ROI on ad spend”, not significant demand, but just any demand I can use as a starting point.

So what I did is set up a Fake Door for the product. I used Nano Banana Pro to generate product screenshots that looked real. I wrote the copy as if the product already exists. I did not make any claims I wouldn’t be able to actually back in the real product. Other than, maybe, the overly complex screenshots. And I tested how many people would actually click on to sign up for a free trial for a product that would later cost them $49/mo. Surprisingly, I got about 8 clicks, of which 3 actually included their email and their industry. And that was only for a couple of hundred dollars of ad spend on Reddit that resulted in 53k impressions. At $49/mo pricing level, that sounded like enough of a justification to start building. No money in hand at that point is obviously not proof of a market either.

The positive side to this is that in the future I know the benchmark to compare against. I know what to expect.

Process of building & launching

The notable part of building the product was how many amazing tools are out there. For one, I didn’t know how much you can just do with serverless functions. Another thing that surprised me was the amount that I would actually learn about my target market.

Then, of course, I added the site to all listicles I could think of, except for some, where there’s more danger of technological farms just copying your product one-to-one, than getting a valuable backlink. I announced it on LinkedIn and got a few nice comments, as expected. All up to plan at that point.

Then, it was time to actually go back to the channel that I tested the idea in the first place! Why not just let the same kind of people come on board again that were interested before? So I created a bunch of good-looking Reddit ad images. The CTR was even greater than before, but the quality of the people landing on the site was way worse. It looked like it was mostly bots, almost all on mobile, and just scrolling back and forth. On round two, when I tried the same ads, only for desktop users, Reddit couldn’t properly even deliver these ads at that volume. So something’s gotten way worse on the platform since, because it worked fine on similar spend during the Fake Door test. A couple of hundred in ad spend, and no sign-ups from this channel. Note that the ads themselves were better than before. More people actually clicked on them. But with mobile on, Reddit sent me worse leads, and the website didn’t convert as well.

Then I tried going on Meta ads, but it self-destructed on signup because it doesn’t know how to charge credit cards (they blocked my payments). Now attempting Google Ads, so we’ll see how this goes. Since going full-time on this and building the product, it’s been less than 2 months, so we’re still in the early days. There’s lots to learn, and there will be a part 2.