Havas has identified a gap that many brand teams may recognize but struggle to measure: AI can make marketing operations faster without making the customer experience more valuable. Its new Valuable AI Experience framework evaluates whether AI-enabled touchpoints deliver utility, credibility, and appeal, rather than simply whether they reduce cost or production time.
The distinction changes the question marketers should ask. An AI roadmap can document where the technology appears across a journey. It does not show whether customers find those moments useful, trustworthy, or worth returning to. Havas’s findings suggest that the most visible AI application is rarely the most valuable one.
Table of contents
Jump to each section:
- What Havas measured
- Why utility is beating visibility
- The AI efficiency trap changes the brief
- What marketers should know about AI brand experiences
What Havas measured
Havas developed the Valuable AI Experience Score to assess AI-enabled brand touchpoints through three lenses. Utility measures whether an experience is genuinely useful and relevant. Credibility captures trust, while appeal considers whether the interaction is engaging. The score also penalizes experiences that consumers perceive as manipulative or deceptive.
The research combined qualitative and quantitative methods across seven markets: Brazil, China, France, Germany, India, the UK, and the US.
10,558 respondents and 190 qualitative interviews informed the research across seven global markets.
Havas then applied its framework across a large set of consumer judgments.
More than 21,000 evaluations assessed AI-enabled brand touchpoints using the Valuable AI Experience framework.
Only a small share reached the framework’s highest level.
9% of evaluated AI brand experiences achieved the highest score for consumer value.
That result does not mean consumers reject AI. It means adoption is a weak proxy for experience quality. A brand can deploy AI across content, commerce, and service while still failing to create a touchpoint that customers consider meaningfully better.
AI presence is easy to count. AI value is harder to prove.
Why utility is beating visibility
The category results make the pattern clearer. AI-assisted education and support earned the strongest Valuable AI Experience Score, followed by product comparison and brand integration inside conversational tools.
56.7 out of 100 was the score for AI-assisted education and support, the highest-performing use case in the study.55.4 and 54.9 out of 100 were the respective scores for AI-assisted product comparison and brand integration within conversational tools.
These use cases give AI a defined job. They help someone understand an issue, compare options, make a decision, or seek support. The customer benefit is legible before the technology becomes part of the brand story.
By contrast, AI-generated influencer content sat at the bottom of the ranking, followed by traditional advertising and shoppable formats.
45.9 out of 100 was the score for AI-generated influencer content, the weakest use case measured.48.4 and 49.6 out of 100 were the respective scores for traditional brand advertising and shoppable AI formats.
The conventional assumption is that more creative output creates more opportunities to engage. The contrasting reality is that consumers rewarded AI most when it reduced decision friction, not when it increased the volume of branded material. The strategic implication is straightforward: the strongest AI brief may begin with a customer task rather than a content format.
The best AI experience does not announce the technology. It makes the customer’s next decision easier.
The AI efficiency trap changes the brief
Havas calls the gap between internal optimization and customer value the AI efficiency trap. The phrase is useful because it separates two outcomes that often share one business case. Faster production can improve margins, but it does not automatically improve the experience being produced.
Consumer perceptions in the study show where that gap becomes risky.
50% of respondents said brands’ use of AI makes touchpoints look, feel, or sound identical.56% of respondents said brand AI can feel misleading.29% of respondents believed brand AI interactions feel human.
Scale can therefore magnify sameness as efficiently as it magnifies a good idea. When teams evaluate an AI initiative mainly through output, turnaround, or cost, they can miss whether the resulting touchpoint weakens distinctiveness or trust.
Havas also found that high-scoring experiences were associated with stronger commercial and brand outcomes than low-scoring implementations.
4.4 times higher engagement was associated with high-scoring AI experiences compared with low-scoring implementations.1.6 times stronger brand desire and 1.5 times greater growth potential were associated with high-scoring AI experiences compared with low-scoring implementations.
Those relationships do not establish that a higher score caused the outcomes. They do, however, give marketers a more demanding basis for prioritization. Efficiency remains a valid goal, but customer value needs its own measure and its own owner.
AI can scale an experience before a brand has proved that the experience deserves to scale.
What marketers should know about AI brand experiences
The framework points toward a more disciplined way to judge where AI belongs in the customer journey. The question is less about how much AI a brand can deploy and more about which moments become more useful because AI is present.
Start with the customer task. Education, support, and comparison led the study because each begins with a recognizable need. A defined task also makes it easier to decide what success should look like.
Separate operating gains from experience gains. Production speed and cost savings matter, but they measure the brand’s benefit. Utility, credibility, and appeal measure what the customer receives.
Protect distinctiveness as AI scales. When half of respondents detect growing sameness, brand voice and experience design become control points, not finishing touches. Automation should preserve recognizable choices rather than average them away.
Treat trust as a design input. The weak scores for AI-generated influencer content and traditional advertising suggest that passive persuasion gives consumers fewer reasons to accept automated involvement. Clear function and credible context matter more when the commercial intent is obvious.
The larger change is organizational. AI strategy is moving out of a narrow technology discussion and into the same territory as customer experience, brand governance, and portfolio prioritization. That makes marketing responsible for more than finding use cases. It must distinguish useful scale from merely available scale.
As AI becomes less expensive and more accessible, deployment itself will stop signaling sophistication. The harder advantage will come from choosing the right moments, defining the value exchanged, and declining automation where it makes the experience more generic or less believable.
The next phase of AI marketing may be measured less by how much a brand automates and more by how carefully it decides where not to.