Our Bet

Starting a new project is like placing an informed bet. Especially for tech products today because things are changing so fast. The bet is what you think the future will be like. Placing the bet means building towards that future, so that your project converges with it. Here I want to explain the bet behind sessionvision.

A small piece of the pie

First, sessionvision is competing in an already-crowded space. Product analytics has big players with enterprise-grade product suites: PostHog, Amplitude, Mixpanel, Fullstory. (My favorite is PostHog - mad respect for them.) This is good and bad.

Good because there is an established need and we don't have to do any work establishing a category. These companies have a combined ARR of ~$700M¹. If two indie builders can take 0.1% of that pie we will be among the happiest builders in the world.

Of course, the "bad" side is that we are two dudes going up against teams of well-paid experts shipping features at extreme velocity (another shout out to PostHog). But surely the pie is big enough for us all.

Two ways to make money

In our modern era we still live under a certain constraint: when you build stuff you have to sell it (not so fun) to be able to keep building (the fun part IMO). Typically there are two ways to do that: create value or extract value.

We all can think of businesses that exist to extract value from their users. Think gambling software, and social media that optimizes for how much time you spend on their platform so they can sell you (oops, I meant to say ads) to others.

In the product analytics category, extracting value looks like making dark patterns more effective. Like better hiding a "cancel subscription" button or putting up a paywall at the exact right place so that users at peak frustration give you a few more dollars. That doesn't feel good, we won't build for those use cases.

For sessionvision creating value looks like real people's daily lives becoming net better because our solution makes their experience of software net better. Example: A doctor's record-keeping software is so ergonomic that she spends more time with patients and less time clicking buttons. That'd be awesome.

The reward for success is seeing one's work help others and become really useful to them, making a living along the way so you can keep doing it. That means we can only have success by creating value, not extracting it.

The main bet: more software, not less

To empower builders, they have to exist. Some think that software builders might not exist for much longer.

Software development is moving into the future perhaps faster than ever. In my own daily work, I have experienced a complete transformation in the last 8 or so months, going from using AI like a better stackoverflow to "orchestrating" parallel end-to-end coding agents on top of Conductor. And this workflow could evolve completely again in the next X months based on new tech emerging.

Lot's of people are making predictions. We must because we need to skate to where the puck is going. Saaspocalypse is one such prediction - Saas is dead and personal AI agents are eating it.

I don't think Saas will die but I do think the bar for Saas (and software in general) is going up. There's no more room for simple CRUD apps that can be spun up in a day or less, and no more room for apps that operate in silos, cut off from interacting with the world outside. AI has unlocked a whole new realm of what software can do. Software can now react to context, work creatively with language, video and audio, and do stuff that daily makes me say out loud "I cannot believe we are living in this world right now." The expectations for software are skyrocketing and the tolerance for bad software is plummeting.

If we apply Jevons paradox we can see a future where there is more software doing more stuff in more domains, more builders, maybe even more Saas.

Analytics fatigue, smaller teams doing more

If we take the above as likely, then demand won't be the problem. Building something truly useful that competes against incumbents remains. As well as convincing people to try it (hence this blog post).

A trend I've experienced first-hand with PostHog, Amplitude and the like goes like this: founding teams set up analytics, build dashboards, start shipping and measuring. The product evolves, people need to update event tracking and fix dashboards. Eventually teams just keep shipping and engineers + PMs don't necessarily have the bandwidth to keep up with their analytics. Dashboards rot and people learn implicitly to not set up event tracking for the next features because it won't matter after you ship it. The end result is a $300/month subscription to an unused tool. Plus low or no visibility into the impact of what you've built.

This is no one's fault, it is just the result of teams that ship fast and have more work than they can manage, due to real world constraints.

The time between data collection, insight, and action becomes a feedback loop that keeps getting longer until the loop breaks and never goes beyond the data collection part. We want to shorten the time and improve each step as much as we can.

Product teams feeling like they don't have enough time or energy to do everything is not going away, even with the rise of AI agents who build and execute their own software. And the solution is not just adding more human or AI teammates.

Part of the solution is better tooling. Analytics without fatigue. If sessionvision becomes a tool that helps teams build better in the right direction, then we would be happy.

That would be success to us.

---

Footnotes

1. I literally asked Claude deep research estimate combined ARR of PostHog, Amplitude, Mixpanel, Fullstory and did not verify the results at all - yolo ↩