• Z-Axis
  • Posts
  • Doing Things That Don't Scale

Doing Things That Don't Scale

Exploring a classic piece of startup advice

The Rundown

You may have heard the expression “Do things that don’t scale.” The idea is that early stage startups are able to leverage highly inefficient ways of accomplishing goals, ways that larger companies cannot leverage.

My favorite personal example of doing things that don’t scale comes from the early days of my own startup, Anchor. Prior to being a podcasting platform, Anchor was an audio social platform that allowed people both to create and consume audio stations.

Anchor in the early days

One day, after digging into our data, I approached my co-founder, Michael Mignano. “If we were to have a human being record just 200 names,” I told Mike, “we’d be able to greet over half of our users by name every time they launch Anchor.” We had a quickly growing user base, but it was still small in absolute terms and primarily English speaking. And first names tend to follow a power law distribution, with a high concentration among a few and the remainder in a long tail. Therefore, a day of work by a human being recording 200 names could lead to a personal greeting by the platform each time half of our users launched it.

This was 2016. AI-powered synthetic voice was not yet a viable option for this type of feature. Nor were LLMs a technology that existed beyond academia. But what good was just greeting people by name? We began to take this idea further.

Every month, day, and year could similarly be recorded by a human with relative (albeit tedious) ease. Geo-location, we realized could be used to estimate whether a user lived in a major city, and we could record fifty of those major city names. Weather could be thought of the same way. A person recording a handful of weather conditions (“cloudy”, “sunny”, “rainy”, etc.) and the numbers between -10 and 105 could cover nearly all weather conditions in natural voice. We could then use a weather API to give us information about the forecast where the user lived and convert it into a stitched together audio file:

Hey + {Annie} + Today is + {September} + {5} + {2017} + Here’s the weather in + {Miami} + It will be + {sunny} + with a high of + {79} + degrees and a low of + {55} degrees.

You’d be amazed how good this actually sounded. Simply stitching together bits of human-recorded audio was enough to start your Anchor experience on a personalized high note.

So we branded it The Rundown, launched it, and our users loved it. Each day, that non-scalable human-read audio, plus a little bit of programming logic for stitching it together, delighted our users in an unexpected way.

Here’s the thing about The Rundown. It didn’t scale. Only half of our users benefited from it. The method we used to create the content could only get us so far. But that “so far” was quite far for an up-and-coming startup.

I can vividly recall how, shortly after we released The Rundown, we demoed it at a meeting with a big tech company. We were told afterwards that some of the engineers in the room were amazed at how powerful and realistic our AI voice sounded, and they were shocked that a company with only a handful of employees and limited funding could achieve such high quality.

Little did they know The Rundown was simply a guy in a recording booth reading “Hey John”, “Hey Mary”, “Hey Linda”, and so on from a spreadsheet I’d printed out.

On Friction

So back to the cliché. “Do things that don’t scale” was once something I rolled my eyes at. Now, having seen its success, I’m now part of the choir that champions this advice. “Do things that don’t scale” means taking steps early on in the entrepreneurial journey that help you make progress even though they’re not sustainable in the long term.

We can understand why this works with an analogy to concepts we all learned about in middle school science: static friction and kinetic friction.

Any time you move something, the contact of the material with whatever it’s traversing (a surface, water, or even air) causes some energy to be lost via conversion to heat. We call this concept friction. And intuitively, it means that you can’t accomplish anything without having to overcome some resistance along the way.

There are two different types of friction. Consider an object on a table. It takes more force to move the object in the first place than it takes to keep it moving. That’s because an object’s static friction is greater than its kinetic friction.

Starting an object moving meets more resistance (the red) than does keeping it moving (the green)

Amazingly, once you get to the kinetic phase (the green) and the object starts moving, an increase in speed does not result in an increase in friction (hence the flat green line).

Generalizing these phenomena beyond science, here are a few universal truths we can deduce:

  • It’s impossible to move something forward without encountering resistance

  • There will be more resistance in starting moving it forward than in keeping it moving forward

  • Once you get going, the faster you go, the less the friction will matter

On Startups

Which brings me back to startups.

Prior to the magical state where things suddenly get going (often referred to as Product Market Fit), a startup is like that object on the table. There are significant resistive forces that prevent it from moving. At that red phase, the job of the employees (and particularly of the leader) is to focus, above all else, on entering the green.

And so we can re-label that physics chart above with terms that fit startup physics.

The Rundown would likely never have been built in the green phase of the startup journey. But that’s precisely why it worked, and why it helped us overcome the red.

“Do things that don’t scale” is another way of saying “do things in the red zone that won’t work in the green zone”. Why? Because the forces that are keeping you from starting are not the same forces that will keep you from moving. In other words, the laws of startup physics are different prior to achieving product market fit. Use that to your advantage.

Don’t reject solutions that won’t work in the green when you’re still in the red. Realizing that is quite liberating. It means that the solution space for any problem is substantially bigger the earlier on you are in your journey.

And that isn’t just the case for startups, but a valuable realization for life more generally. Yes, it’s harder to get started than to keep moving. So use tricks to get you started, even if your approach might change down the road. You only get down the road if you start moving in the first place.

P.S. If you enjoyed this article and know someone who might like it, please consider sharing it. Forwarding these emails is the primary way Z-Axis is able to grow.

P.P.S. If you aren’t yet a subscriber yourself, consider subscribing to this free weekly newsletter by clicking here.