Our AI VP of Marketing “10K” Said To Drop Prices. Our 10+ Years of Data Said Otherwise. Our AI VP Marketing Was Right.

SaaStr AI Annual 2026 Attendance Is Up +41%. Don’t thank the humans on our team. Thank 10K. Our AI VP of Marketing.

For the better part of a decade, we had a working theory about SaaStr AI Annual ticket pricing. All our spreadsheets and human analysis confirmed it.

It went something like this: ticket prices were already cheap compared to other major B2B + AI conferences. Founders and execs knew it. And the data seemed to confirm it. Folks buying at the last minute especially didn’t seem to care what the price was, as long as it was below the threshold of “feels like a lot.”

This year we wanted to go bigger. Things are going well with SaaStr AI. People do care about our AI Agentic journey. So how do you go bigger? One “obvious” idea — drop prices, drive more attendance — felt like a waste to even discuss. If buyers already weren’t price-sensitive in any meaningful way, why leave revenue on the table? And I’ve never found lowering prices that already aren’t very high really incents any buying behavior from high-value buyers.

We believed this for a long time. Years.

Then we built 10K. Our AI VP of Marketing. And 10K said … Stop. You’re doing it wrong.

10K is our AI VP of Marketing. We vibe coded it ourselves after spending months looking for something off-the-shelf that could actually run marketing for a niche community, media and events business. Not a mediocre AI app to generate AI content, but one that could truly orchestrate marketing. Every campaign, every offer, every channel, every day. Nothing on the market did that for real, so we built it.

Under the hood, 10K uses Claude Opus to do deep analysis of all of our data across 5+ years. Every campaign. Every conversion rate. Every registration pattern. Every sponsor interaction. That analysis feeds into an app that has designed every single marketing action for each day of the year through mid-2026. And it updates in real-time as new data comes in.

It doesn’t carry our assumptions. That’s the whole point.

One of the first things 10K did when we gave it access to the full dataset was build a real pricing model. Not vibes. Not gut. A model.

Its conclusion: “Drop pricing 15%, and attendance will go up 40%.”

We pushed back hard. Oh 10K, we’ve had that conversation so many times before.

We thought we had years of data that seemed to say price didn’t move the needle. We’d watched last-minute buyers pay almost anything. We’d convinced ourselves the constraint on attendance wasn’t price. It was awareness, or intent, or logistics. What difference does $80, or $40, or $51.23 really make for a C-level exec to a well funded founder?

10K disagreed. She showed her work. She argued that our pricing model was suppressing a layer of buyers we weren’t even seeing. The ones who looked, calculated, and quietly decided Not Now. They never showed up in our data as price-sensitive because they never showed up at all.

We eventually agreed to try it.

And … as of right now, SaaStr AI Annual 2026 attendance is up +41% year over year. And ticket revenue is still up +6%.

Our AI VP Marketing was right. Our goal was to maximize high value attendees while still covering our costs.

When you’ve run something for a long time, your assumptions calcify. You’ve seen enough data to feel confident — and that confidence becomes its own kind of blindspot. You stop questioning the conclusions you drew three years ago.

An AI that has no stake in being right, no memory of why you made the original decision, no agenda, no politics, and no attachment to the current model will just… look at the data. Clean. No history to protect. No campaign she built last year that she’s defending. No sunk cost fallacy.

That’s the thing we underestimated most about 10K. Not the automation. The honesty.

That’s genuinely useful. Not because AI is smarter. But because it has no ego invested in the old answer.