An inside look at the ideas rising to the top at MAU 2026
May 21, 2026
Retention and engagement benchmarks are still holding up in 2026 — but the platforms producing them can't see the half of the journey that increasingly decides them. Here's what stood out from MAU Las Vegas this week, in sessions, over lunches, and on the exhibit floor.
MAU Vegas tends to surface the tension the app industry is living with. Sessions fill up on growth and monetization, but many of the conversations that linger are about measurement — what teams can and can't see when they try to understand how users actually behave, and what to do with the gap.
This year was no different. The RealityMine® team was on the ground in Las Vegas this week, and a few themes kept surfacing across keynotes, panels, and the conversations in between.
One of the most consistent threads running through the event was the gap between what internal dashboards report and what's actually happening at a category level.
Teams kept returning to the same point: discovery, conversion, and ultimately, attribution are being obscured by new surfaces and the rise of AI - and the signals available to act on aren't keeping up.
A related point that kept coming back: the importance of high-quality data for training AI systems and agents. AI agents and automation are only as good as the signal feeding them. Bad data scaled by AI is still bad data, just faster. Too often, data is fragmented, incomplete, poorly sourced, and possibly compromised by bots. Whatever the flaws in your data, AI will magnify them, making them nearly inescapable. One line from the floor that stuck: "AI will just scale your fragmentation."
Tentpole moments like the Super Bowl deliver install spikes that look like a campaign working — the dashboard lights up; the team gets the win. The harder question, DraftKings raised from the stage: how many of those peak-season installs would have happened anyway? It's the kind of question first-party data structurally can't answer. The signal lives outside the app — in what competitors were running, what users were seeing across other platforms, and where attention was already pointed before the campaign started. Without that view, "our campaigns are doing really well" is a hypothesis dressed up as a metric.
The framing that kept coming up: Retention metrics can look healthy while share of category attention shifts in the background. An app's engagement numbers don't tell you whether users are spending more time with a competitor, whether they're splitting attention across three platforms instead of one, or whether a new surface — AI assistants, for instance — is intercepting the journey before it even starts.
An underlying argument behind several key sessions — that unified signals across mobile, web, and CTV are increasingly necessary for accurate performance measurement — maps to a broader structural shift the industry is navigating in real time.
Kalshi CMO Allan Maman, speaking on stage about how the prediction markets platform manages its channel mix, put it plainly: the team scales up or pulls back day-to-day based on which channel is delivering the best quality user at the lowest cost — not on fixed allocations. For a brand operating in real time around cultural moments, that kind of flexibility is only possible if your measurement is fast and granular enough to act on.
Walled gardens like Snap can show advertisers what happens inside their own platform, but not what users do once they've left it. That gap has existed for years. What's changed is that discovery is shifting away from traditional search, and the gap is widening faster than the tooling is closing it.
Deb Gabor's Executive Forum keynote on "irrational loyalty" made the point that the apps that retain users aren't necessarily the ones with the best features. They're the ones that understand the emotional role they play in a user's life — and design for it consistently.
The strongest apps are the ones that fit themselves into a user's everyday habits. If your go-to office lunch order is on DoorDash, the app learns the time you usually order and nudges you before you remember to think about it. The aim is to make the service feel like the default — to strip out friction and become the thing you reach for without weighing alternatives.
That's a harder argument to make with in-app data alone. App publishers and platforms need to know what users are doing outside their apps — which competing platforms they're reaching for, how sticky their habits are, which moments lead to drop-offs, where their product sits in users’ daily digital stack.
The through-line across all of it: App discovery retention benchmarks and engagement metrics are necessary -- but as the market continues to evolve, they aren’t sufficient. The teams that will pull ahead aren't just measuring better within their own platforms; they're building a view of how users move across the entire category.
That's the problem RealityMine® was built to solve. Our technology captures real-world app usage across the platforms clients don't own — giving growth, strategy, and insights teams session-level visibility into competitor mechanics, where attention and spend sits across the category, and the demand shifts first-party data can't surface on its own.
We wrote about why this matters for AI model training earlier this month, and Kate Jacobs recently published data on how search behavior is shifting across the US and UK, which speaks directly to the changing discovery landscape many MAU sessions circled around.
