The model of a product-led company

“Don’t make me think”.

Steve Krug¹

There’s something I say all the time when advising startups on their product strategy: solve problems for people, and the money will come². It still amazes me that this often comes as a shock.

All too often companies try to ‘game’ their customers: bait ’n switches, enshittification, bullshit marketing. These are all signs of a company focussed more on itself than its customers. They somehow convince themselves that being customer focussed is a lesser path to profits ³.

I can hear you snorting now though - “you’re being idealistic and unrealistic!” Yet throughout my 30+ year career, I’ve seen behaviours that prove the point: make a product that solves a customer problem, make it easy for them to love you for solving the problem, and the revenue comes. Even if you’re not charging for the product.

A couple of weeks ago, I saw another proof point of this, with a product I have only recently discovered, yet already love enough to want to pay for it, although I don’t have to pay for it. Tailscale.

What makes a great product

Here’s how to make a great product that will sell: solve a problem and make it incredibly easy for other people to adopt and use your product to solve their problem. That’s it (again, see ²).

The automatic thought is Tailscale is a VPN, but really, it’s not. It’s a simple way to connect devices securely and privately. Trying to get yourself off the cloud drug and repatriate your own data to a NAS you own, at home, yet want to access it from your phone when away? You need a way to securely and privately connect those devices without exposing your private NAS to the internet. Tailscale is the answer.

However, implementing technical products like this is often hard. Which is where Tailscale hits my simple requirement. I connected a handful of devices with barely any thought whatsoever. There was no need to mess about with firewalls; I didn’t need to think about or understand routing and IPv4 versus IPv6. I just followed their simple instructions and it all just worked.

As I’ve used Tailscale more, I’ve realised they also tick a few more of my required checkboxes. Whilst looking at the interface the other day, I noticed something about “an exit node” — with a small (i) icon for more information. The information box showed me how to make a device an exit node; what switch to run the command with. No need to branch off to the documentation to find out how to do this; the minimal information required was there, in a helpful little box.

Don’t make me think

To make a product really easy to use, the complicated has to be simplified somewhere. A lot of companies don’t put enough effort into that simplification, leaving their users to struggle. This is a perennial problem in the tech industry. However, if you, the company developing a product, put the hard work into that simplification process, the user benefits. Which ultimately means you benefit, because they’re more willing to use a product that has little friction to adopting it.

If I can’t be productive with a product within 15 minutes, I will give up on it. If a piece of information isn’t clear, or even worse, there is deliberate obfuscation that makes me work to understand what I’m looking at, I won’t trust the company. I don’t think I’m alone in this thinking, either. People are tired of being used, trapped and ’experimented on’ just to use a product. The hard work of simplifying has to happen somewhere — and it should be you, not your users, doing it. Make it so easy to use your product that I don’t have to think. Thinking is literally costly to our bodies (it takes more energy to use “System 2” than it does “System 1”⁴).

It starts with company culture

“Company culture” is a woolly phrase, too often used as a euphemism to create tribal loyalty. That’s not what I’m talking about here. I’m talking about leadership’s appreciation and respect for people, not least their customers. When the leadership of a company is driven by the desire to solve problems for people, it more often than not flows through the company to create respectful products.

Which brings me back to Tailscale. Recently they changed their pricing and made their free tier even better. In the live chat on Discord (you can watch the recording here) Tailscale’s CEO, Avery Pennarun⁵, actually said at one point (10 minutes in to the recording) that 99% of customers on the paid-for Plus product don’t use any of its features, they just want to give the company money.

I’ve seen that behaviour before. In the early days of Ansible, we had a customer pay for the commercial product, which they didn’t actually need. They loved the open source part so much they just wanted to give the company some money.

Make life easier, faster, or cheaper

We humans dislike change. Even if we have a problem, there can be resistance to adopting a new product to solve it. Especially since historically products haven’t always solved the problems they purport to. A product needs to solve a problem that, in doing so, makes the user’s life easier, faster, or cheaper. Even better if it nails all three. A person’s inertia is far easier to overcome if trying your product is as simple as possible. Don’t make me think.

Any third-rate engineer can increase complexity; but it takes a certain flair of real insight to make things simple again.

E F Schumacher