Is your tech stack ready for 2026 and beyond? Download our new eBook and get a blueprint for the composable AI data stack.
AI decisioning is one of the most promising technologies available to marketing teams today. But adopting it isn't so easy. We break down the three most common hurdles teams face when implementing a decisioning tool and how they can overcome them.
In the video series, From Data to Decisioning, we explore the concepts of causation and correlation in marketing, and why causality data is the missing ingredient your team needs to drive meaningful growth with agentic AI.
Watch episode 7 below, and scroll to the bottom for a full list of episodes.
Tobi:
One of the most promising tools in the emergence of generative AI that we've seen for marketing is what we call AI decisioning. Yet, we know for most of our customers, it's very hard to adopt AI decisioning with measurable results that move the needle. Why is that the case?
What we understand from our customers is that there are certain hurdles for adoption.
I think one of the biggest hurdles is to be able to pass over to the decisioning system everything there is to know about each customer in the moment. So, a true customer 360, as opposed to yesterday's data that is limited to certain characteristics that define customer context.
Anthony:
I think this is a key piece, Tobi. When you think about what GrowthLoop does differently, instead of sending data to the decisioning system, we bring GrowthLoop, the decisioning system, to the data.
So on top of all of the customer data in the data cloud, there's no loading data out to GrowthLoop, which gives you real-time context to be able to make better decisions faster on top of that data.
Tobi:
Which is a good point and I think that right away translates to the second big hurdle that we see.
The second big hurdle we see is that it's very hard to select an outcome that can be optimized by most of these systems.
Most of the tools that we see in the market push you towards fast wins. Those are what we call vanity metrics, right? Optimizing decisioning maybe based on measurable clicks and measurable conversion.
The problem is these vanity metrics hardly relate to the long-term goals that you as a marketer are measured on. Think: customer lifetime value, long-term retention, customer loyalty.
Anthony:
Yeah, this can drive bad behavior in the system, right?
We've seen this case before, where you try to optimize for retention, but then that spits offers out to everybody, which is probably bad for long-term value and your bottom line, right?
With GrowthLoop, we work with you to set up experimentation and decisioning against long-term results that matter for your business, for long-term outcomes for your customers. Things like lifetime value, revenue, profitability, and loyalty.
Tobi:
Great, so now we're past those two hurdles, which leaves a third hurdle.
All of these decisioning systems, by definition, start cold, meaning they don't know what actual intervention drives the needle.
And even worse, starting cold means that some of these decisioning trees can actually net a negative effect for your long-term goals, and there is no way to control that situation.
Anthony:
Yeah, that's right. So we see the cold start problem out there a lot and the boomerang effect, as you have mentioned, where they take a long time to find a solution and maybe that solution is actually driving negative outcomes.
With GrowthLoop, we help you snapshot long-term causality data. This is what experiments you've run, what states your customers are in when the experiment runs, and what actual outcomes were driven. That way, you can start with the best first intervention, avoid long-term boomerang effect, and avoid negative outcomes through measuring what happens through this causality data.
Watch more From Data to Decisioning episodes:
Episode 1: Overview of correlation vs. causation in marketing
Episode 3: The way everyone is building AI decisioning is wrong
Episode 4: Why causal data is the key to unlocking compound growth
Episode 5: Why composability is a non-negotiable for AI decisioning
Episode 6: The AI decisioning resource you need to start building today
Learn why marketing teams must start building an agentic context graph in their data cloud, and how it helps AI decisioning platforms deliver smarter results.
Unlock AI's full potential with an agentic context graph — a structured memory layer that tracks causal customer journeys and improves every campaign.
Marketers who shift from correlative data to causal AI decisioning can move beyond unreliable A/B testing to deliver true personalization at scale.
