Analytics matter. Maintaining them doesn't scale.
Asher Smith-Rose spent a decade as a product manager before co-founding Solo AI. He'd seen this play out at every company: analytics always made the priority list in theory, and almost never got done in practice.
The issue wasn't motivation. It was overhead. Deciding what to track, building a tagging taxonomy, coordinating with engineers to understand what events meant — then waiting a week for data to accumulate, only to find you'd set something up wrong and needed to start over. At one previous company, getting clean data for a single question took roughly a week. Many initiatives never survived that gauntlet.
At Solo AI, he ran Amplitude and PostHog with Slack alerts for key events. It worked when everything was tagged correctly, which wasn't always the case. When the team launched their new pricing page, Asher assumed tracking was in place. He checked a week later and found there was no tracking at all.
“If you don't tag correctly, you've lost that data and you don't have that signal in time.”
That meant two weeks of user behavior insights, non-existent. By the time they fixed the tagging, the data was usable but the window to act on it early had already closed. Running a startup means making hard calls about where to spend your time. What Asher needed wasn't another tool to maintain. He needed a third member of the team.
From analytics skeptic to true believer in 30 minutes.
When the Novus team reached out, Asher's first instinct was skepticism. He'd been through enough analytics implementations to know that “easy setup” usually meant a week of engineering time, a tagging taxonomy that needed constant upkeep, and data you still couldn't fully trust.
He got on a call anyway. Within 30 minutes, everything was connected and the first outputs were already on screen.
“Novus came back with everything already tagged — buttons, interactions, things I wouldn't even have thought to add. I scanned through it and thought: that's pretty revolutionary.”
What hit Asher wasn't just the speed — it was what the speed unlocked. If instrumentation was already there and staying current automatically, then every question that came up next week, next month, whenever a launch happened — the data would be ready. That shifted how he thought about what was now possible.
For the first time, the constraint wasn't “did we remember to tag this?” It was simply: what do we want to know?
A marketing channel decision, answered in minutes.
Setup was just the beginning. What came next was what changed how Asher built.
With everyone talking about AEO — ranking in AI-generated answers — Asher wanted to know how much of Solo's traffic was already coming from ChatGPT and other LLMs. While he hadn't set up any tracking for this, he asked Novus directly, in natural language, and had a daily update running immediately.
LLM traffic was low.
The data came back: LLM-driven traffic was a small fraction of Solo's overall inbound. That single proactive answer informed a real business decision — Solo committed to a month-long initiative to improve their AI search rankings, with Novus tracking the results to determine whether to keep investing or move on.
“Whatever question happens to pop into my head that morning, we have data on it. And I don't have to spend a lot of effort trying to figure out how to answer that.”
A UX breakdown, caught before it compounded.
The second discovery was one Asher never would have found on his own. Novus flagged that Solo's AI had returned a “we don't have that information” response six times in a single week — a pattern that signaled something was broken, not just missing.
A bug nobody knew existed.
No one had built a tag for this scenario. No Slack alert would have caught it. But because Novus was proactively monitoring behavior, it surfaced the pattern before users started churning over it. When the team investigated, they found a bug and fixed it.
Both discoveries shared the same defining feature: Novus brought them to Asher. He didn't have to go looking.
A third PM that never clocks out.
The change Novus made to Solo's workflow wasn't just speed. It was the nature of the loop between signal and action. In his PM career, Asher's job was to find the insight and hand it to an engineering team. Now, as a founder who's also shipping code, he can go from a Novus signal to a front-end fix himself without waiting on anyone.
The weekly trend digests changed how he stays informed too. Instead of logging into dashboards and hunting for signals, Asher gets a summary of what changed: conversations up or down, features nobody's using, friction appearing where it shouldn't. For a two-person team, it's continuous coverage that would otherwise require a dedicated hire.
“As a founder, you know analytics matter. But without the right system, the work of maintaining them crowds out everything else. Novus cuts through that noise and it pulls you into the conversation when something actually matters.”
If you're not tracking everything you should be, you're making decisions without the full picture. For Solo AI, Novus makes sure they always have it.