Stories about Software


You Don’t Pick Niches — You Discover Them

As I continue along this meandering odyssey of posts about freelancing, I’ve arrived at the point where I should talk about marketing.  After all, I just talked about sales, and then about how not to market.

But the trouble is that, until you pick some kind of niche, your marketing is essentially just platitudinous bullshit.

Marketing is really about making consumers aware of how your offering could (or couldn’t) help them.  What, then, does marketing generalist labor look like?

  • I’m passionate and I believe in quality!
  • I have good communication skills and I believe in under-promising and over-delivering.
  • My credit score is 800!

It quickly descends into the farcical, particularly since there is only one true, accurate piece of generalist marketing: “I’m cheaper than the other people you’re talking to.”

Don’t believe me?  Consider the only two generalist arguments about hourly rates:

  1. I’m the cheapest!
  2. Sure, the other dude is cheaper, but I’m higher quality, which means you’ll actually pay less in the long run, which — plot twist! — actually makes me the cheapest.

So before we can talk about marketing, we need something worth marketing.  We need to get you a niche.

The Ocean-Boiling of Niche Selection

I talk about picking niches a lot.  Here, in fact, are some past articles on the subject:

In that first one, I actually provided a walk-through of what niche selection might look like for a hypothetical generalist.  The reason I did that, and the reason that I talk about it so much is that you — the readers — ask about this all the time.  For most aspiring free agents, picking a niche is the Holy Grail of professional life.

The trouble with creating content on this subject is that it’s so situationally specific.  It’s almost as ocean-boil-y as if people wrote in to me, asking “how can I find happiness?”

All of the bromides in the world about mindfulness and gratitude aren’t going to help you.  Not meaningfully.

But, in spite of this challenge, I’m going to take a stab at it anyway.  I’m thinking I might do it through case studies from user questions, and tackling the subject in more depth.

So beware.  We’re entering weird, uncharted waters here.

Let’s Get Pedantic About Niches

I’m going to start by making a subtle, but important, point.  And I’m going to drive that point home as hard as I can.

You don’t pick a niche.  You discover one.

I have failed to make this distinction in the past, myself, when talking to all of you.  But I want to remedy that, going forward.  It’s important to how you frame the activity in your head.

A niche exists with or without you.

Let’s say that there’s a good bit of demand for a service that creates small microservices as a way to incrementally move fortune 500 businesses away from legacy, mainframe systems.  That need exists whether you supply (or anyone supplies) the service or not.

Thus, at its core, your personal journey of finding a niche does NOT look like this:

  1. Introspect.
  2. Decide what you want to do in your professional life.
  3. See if anyone is interested.

Formulating Niche Hypotheses

Rather, it looks like this:

  1. Go out niche hunting, aka market research.
  2. Form a list (ideally, a large one) of niche hypotheses.  (I’d argue that you can’t truly know a niche exists until you try to fill it.)
  3. Force rank that list on the basis of your interest, your ability to fill the hypothetical niche, and your confidence in the hypothesis.
  4. Start with the top entry and attempt to validate or disprove the niche hypothesis.
  5. Iterate until one works.

To understand what I mean here, let’s go back to this hypothetical niche of low-risk migrations away from legacy services.  Based on all of my years of management consulting in the enterprise, I feel pretty confident in that niche hypothesis.  But I don’t know that this niche exists without trying to fill it.

There are countless things that could go wrong by trying to sell this service into the enterprise:

  • Enterprise sales cycles occur on glacial timelines, so I might go bankrupt before I learn whether an enterprise would bring me in.
  • It could be that large consulting firms are already filling this niche.
  • Maybe there just aren’t that many mainframes anymore.
  • Maybe microservices have faded as an enterprise architecture fad, and monoliths are becoming all the rage again.

So it’s important that I not get too attached to any one niche hypothesis.  I need to retain the ability to tune the hypothesis or move onto a different one with relative ease.

Validating Niche Hypotheses

I imagine that this naturally leads you to the question of how you validate a niche hypothesis.  How do you, as the folksy song goes, know when to hold ’em or when to fold ’em?

I’m not going to elaborate too much on validating niche hypotheses in this post.  If you really want a deep dive on this, read The Lean Startup (affiliate link, if you want to buy me a coffee).  The entire book is about applying the scientific method to nascent business models and, as a bonus, if you read it, you’ll stop using “minimum viable product,” incorrectly in conversations.

The gist of the book and my message here is that you want to figure out the cheapest and easiest way to validate the niche hypothesis.

Take the thematic microservices niche, for instance.  I could stand up a website, put together some offerings, give a bunch of conference talks, run paid ads, and hope that the business started rolling in.  Or, I could just call a bunch of execs in my rolodex and ask if my offering is something they’d buy.

Neither of these methods is perfect, but one doesn’t waste months and thousands of dollars, if my niche hypothesis is wrong.

So as you make your list of niche hypotheses and rank them, channel Karl Popper and have notes about how you could falsify (or validate) each niche hypothesis.  You not only need the niche itself, but a cheap strategy for deciding whether to proceed.

Hedging Your Risk with Content and Market Research

There are probably innumerable ways to get intel on your niche hypotheses.  Some will work better than others and, I’d argue that the only foolproof method to validate a niche is actually to serve it successfully.  You’ll know your niche is valid when people give you actual money to fill it.

But I’ll close out here with the two I would strongly advise defaulting to:

  1. Create valuable content around the niche in question.
  2. Schedule market research conversations around the niche in question.

Both of these strategies cost nothing but your labor, and both of these strategies hedge against the cost of your labor with other potential upsides.

In the case of creating valuable content, you have the content, whether or not your idea works out.  Even if content about microservices migration doesn’t help you sell a boutique productized services, it might help you land generalized app dev roles or a job or something.  The content remains potentially useful and repurpose-able, niche notwithstanding.

And in the case of market research conversations, you might learn about other potential opportunities or niche hypotheses.  From personal experience, I can tell you that the overwhelming majority of ideas I have for offerings arises directly out of conversations with other humans.  All you need is half an hour and a notebook.

So as we explore the idea of niches, understand a fundamental and axiomatic distinction.  You aren’t gearing up to invent niches — you’re gearing up to go hunting for them in the wild.

Newest Most Voted
Inline Feedbacks
View all comments
3 years ago

I think the topic of networking could use a bit of expansion, and by networking I explicitly don’t mean LinkedIn (or really any digital platform). It’s easy to get stuck in a tech bubble of sorts, while you really want to connect with business owners or at least people who can decide to spend money – and likely not Fortune 500 to start.

Erik Dietrich
3 years ago
Reply to  Nils

That’s a great point. I’m creating a future draft for prospecting among the non-technical set. Thanks for bringing it up 🙂