80% of digital experience pros say AI has increased review workload

AI has fundamentally changed how digital experience teams work, with faster execution, more output, and more capability across every stage of generative AI product development.

But that speed has introduced new challenges around time, with 80%  of digital experience professionals saying AI has increased the time their team spends reviewing and verifying outputs before they ship, and 51% saying the increase is significant. Without the right behavioral data behind it, a forms between what AI produces and what teams can actually ship.

This survey of 719 U.S. professionals explores where that gap shows up in AI product development, why behavioral data is the missing input most teams aren't using, and what happens when teams close it.

“AI is changing how fast teams can execute, but when that comes at the cost of increasing review workloads, it's usually a data problem. Without a complete picture of how users experience the product, AI can't make decisions you can actually trust.”

- Claire Fang, CPTO, Fullstory

32% of teams say their biggest obstacle today is reviewing AI work, not building

Verification has quietly become one of the most time-consuming parts of the product workflow. AI has accelerated execution and output volume, but without the behavioral context to make reliable decisions, teams are spending more time checking work than building it. That tradeoff is costing teams the time AI was supposed to give back.

Two years ago, the top constraints looked very different:

• Engineering capacity was the #1 bottleneck at 20%

• Data access followed closely at 23%

Only 18% say their biggest time shift is more hands-on building. Where teams once struggled with resources and access, they're now struggling to give AI the behavioral context it needs to make decisions they can trust.

The builder's role has always been about understanding users and making good decisions on their behalf. AI accelerates that process, but only when it has the behavioral context to make those decisions reliably. That's what separates teams that are moving faster from teams that are just doing more.

- Lane Greer, Solutions Engineering Lead, Fullstory

The review burden isn't equal: Mid-market teams lead at 56%

Not every organization is feeling the review burden equally. Mid-market companies ($50M to $500M) are feeling it hardest, with 56% saying review time has increased significantly. Compare that to

• 46% at enterprise companies and 

• 41% at smaller businesses

Mid-market teams are also the least likely to have behavioral data infrastructure in their stack, with only 47% using session recordings or behavioral data as a primary input compared to 52% at enterprise. 

For teams carrying the heaviest review burden, adding behavioral data as a primary input is one of the most direct ways to start reducing it.

Without that data foundation, teams that haven't closed the insight-to-action gap between what AI produces and what users actually do are absorbing the cost of that in review time.

Mid-market teams are in a tough spot. They've invested enough in AI to feel where it's falling short, but haven't yet built the data foundation to fix it. At that stage, AI is making decisions without a complete picture of how users actually behave, and the review burden reflects that.

- Addison Price, Head of Global Solutions Engineering, Fullstory

Only 44% of teams feed AI the behavioral data that 83% say is essential

Eighty-three percent of digital experience professionals say behavioral data meaningfully improves AI decision quality, but only 44% are actually using it as a primary input. Without that signal, AI is missing the it needs to navigate your product reliably.

Instead, more familiar data sources are leading the stack:

• CRM: 57%

• Product analytics: 56%

• Customer feedback: 56%

Teams that do use behavioral data as a primary input are more confident (55%) that their AI understands real user behavior than those who don’t (45%). Higher confidence in outputs means less time reviewing them before you ship.

The teams that trust their AI outputs are the ones that have given it the clearest picture of their users. Behavioral data isn't just another input. It's the one that tells AI what actually happens when a real person uses your product.

- Peter Antoniou, Staff Sales Engineer, Fullstory

Case study: Behavioral context drove a 6% lift in AI resolution rates for Ninety.io

Ninety.io is a SaaS platform that helps organizations align teams and manage execution. As their AI support workflows scaled, the AI was answering questions, but conversations felt impersonal. Customers repeated steps they had already taken, and agents had to reconstruct the journey before they could start resolving issues.

To help, Ninety.io used , Fullstory's tool for delivering behavioral session summaries into external platforms. Ninety.io gave Intercom Fin, their AI support agent, a complete picture of each user before the conversation began.

StoryAI, Fullstory's AI agent layer, put that behavioral context to work, enabling Fin to understand, predict, and act on real user behavior rather than starting every interaction from scratch.

We all experience AI fatigue; AI interactions can feel less friendly. They lack the conversational warmth of a human touch. This is something we didn't realize we were missing. Fullstory adds a layer that makes Fin a more personable and friendly experience for our clients.

- Taylor Paletta, Director of Support and Digital Success, Ninety.io

Since adding behavioral context, 79% of tickets exposed to Fullstory data are resolved, and Ninety.io saw a 6% increase in support tickets resolved overall.

Create intelligent digital experiences with Fullstory

Teams that give AI a complete picture of how users actually behave are more confident in their outputs, spend less time reviewing them, and achieve higher digital experience optimization and .

Fullcapture automatically indexes every user interaction without manual instrumentation, giving AI the complete behavioral signal it needs from the start.

To see how that connects into the AI workflows your team is already using, start with Fullstory MCP.

Give your AI the behavioral context it needs.

Learn more ➜

Methodology

The survey was conducted by Centiment for Fullstory. The results are based on 719 completed surveys among U.S. professionals working in product management, engineering, UX and design, growth and experimentation, and executive leadership roles. Respondents represented a range of company sizes and AI maturity levels, from early-stage experimentation to fully agentic systems, across industries including retail and ecommerce, financial services, travel and hospitality, SaaS and tech.