We didn't build another GPT wrapper strapped to a dashboard. We built the analyst marketers actually deserve.
 

Every martech vendor on the planet is shipping an "AI assistant" right now.

Open LinkedIn. Scroll for 30 seconds. You'll find at least three "agentic copilots," two "AI-native analytics layers," and one "autonomous growth agent." Spoiler: most of them are the same thing. A chat box bolted onto a dashboard, wired to GPT, trained to answer questions from a pre-written knowledge base about the underlying platform. Some added chatbots where you can ask your analytical dashboards, or even attribution dashboard with marketing insights. 

There’s a saying we use at INCRMNTAL. It’s not very “PC”, but it is the truth: Garbage in, eloquent garbage out.

We didn't want to add to that pile.

So today, we're genuinely excited to announce something different: AURORA - the marketing industry's first conversational AI growth analyst built natively on causal incrementality data. AURORA is now in BETA, and it's already doing things in customer accounts that frankly surprised even us.
 

What AURORA Actually Does

AURORA runs on top of INCRMNTAL's always-on incrementality engine - the platform that does not require user level data, nor running of any 4 week planned experiments to measure incrementality. The same platform that DraftKings, Paramount, eToro, and Supercell already trust to measure true incremental impact.

The difference is the interface.

Instead of opening the Explorer, picking dimensions, filtering by country, sorting by marginal ROAS, and reading the chart - you just ask.

Real questions our beta users are asking right now:

• "I need to scale ad spend by $1M without raising my CPA. Where should it go?"
• "How can I grow my iOS revenues in the US?"
• "What are my top-performing channels by iROAS?"
• "Which channels are cannibalizing each other?"
• "If I had to cut spend by 50%, what should I reduce first - and why?"
• "Should we keep investing in Brand?"

AURORA pulls the contribution data, runs the marginal cost modelling, considers diminishing returns and your individual platform constraints, and comes back with a recommendation. Not a chart. A recommendation. With reasoning. In seconds.

And here's the part I love most: AURORA knows you can’t scale indefinitely. 

If Brand is saturating, it'll say so. If Applovin has higher marginal ROAS than your biggest channel, it'll point you there. If your Meta CPA is 45% above marginal cost, it'll flag that you're overpaying for users the model says should be cheaper. It doesn't optimize for what makes you feel good. It optimizes for what's true.

A Real Example From BETA

One of our BETA users asked AURORA to scale $2M/month in ad spend while keeping CPA constant. AURORA didn't just spread the budget evenly. It identified that allocating the full $2M into three specific channels would actually improve combined CPA by 61.8% - bringing it from $131.76 down to $80.46

That's not a small efficiency gain. That's the kind of move that, in beta testing, surfaced budget reallocation opportunities delivering up to 29x improvement in marginal efficiency - by identifying spend trapped in low-incrementality channels, or channels that reached the point of diminishing returns and any additional dollar spent was simply not yielding any return.

This is what happens when conversational AI meets causal measurement instead of attribution theatre.

What Customers Are Saying

Shad Alter, Head of Programmatic at eToro, said it best:

"Programmatic moves fast, and the questions move faster - where to scale, where to cut, where the next dollar earns the highest incremental return? Having a conversational interface sitting on top of real causal measurement changes how we operate. Instead of waiting on a report or building another dashboard, I can ask a direct question and get a direct answer backed by incrementality data. That's a meaningful step forward for how performance teams work day-to-day."

That's the bar we're holding ourselves to. Not "cool demo." Actually changes how performance teams work.

AURORA functions both as a side bar across the Explorer dashboards, as a full screen mode, supporting threads, and pre-set prompts to inspire insight seek.

How to Get In

AURORA entered BETA in May 2026 for select INCRMNTAL clients. General availability is planned later this year.

Two ways to get involved:

If you're already an INCRMNTAL customer: Reach out to your customer success manager. Spots are limited, and we're prioritizing accounts that can give us strong feedback across different verticals and spend profiles.

If you're not a customer yet: Book a demo and we'll show you AURORA running on a real incrementality dataset. Fair warning - once you've asked a question in natural language and gotten a marginal ROAS answer back in 18 seconds, opening a BI dashboard will start feeling like using dial-up internet.

The truth doesn't lie. Marketing measurement has spent the last decade telling marketers what they wanted to hear - last-click wins, every dollar is incremental, scale is always good. Causal measurement broke that comfortable lie. Now AURORA makes the truth easy to ask for.

We didn't build another assistant. We built an analyst.

INCRMNTAL

AURORA is part of INCRMNTAL, the always-on, privacy-first incrementality platform. INCRMNTAL was acquired and is now part of Smartly. Learn more at incrmntal.com or smartly.io.