Employee Login

Enter your login information to access the intranet

Enter your credentials to access your email

Reset employee password

Article

Live From AMEC 2026: AI Is Making Communications Measurement More Important Than Ever 

November 16, 2026
By Ines Schumacher

It has never been easier to produce a dashboard, generate a research summary, or put a number against communications activity. The harder question is whether that number tells us anything meaningful, and whether it’s reliable enough to guide a decision.

That question stayed with me after taking part in AI Day hosted by AMEC, the international association for communications measurement and evaluation, in London last week. The discussions brought together communications measurement professionals working through what AI makes possible and what our profession must do to use it well.

From Left to Right: Ben Levine, Peter Lakel, EJ Kim, Ines Schumacher, Liam Heaton and Nicole Harris at the 2026 AMEC Global Communication Effectiveness Awards.

There was plenty to celebrate, too. FleishmanHillard TRUE Global Intelligence won 20 awards at the annual AMEC Awards, recognizing work that combined innovation, rigorous research, and meaningful outcomes. I was proud of our team and encouraged by the ambition across the industry.

Big Wins for FleishmanHillard in the News: ICCO Large Consultancy of the Year | PRovoke Global Campaign of the Year

A point made by Dr Tina McCorkindale, President and CEO of the Institute for Public Relations and recipient of the Don Bartholomew Award for advancing the profession, captured the responsibility ahead: being able to measure something does not make it meaningful. Our contribution depends on whether we help organizations make a difference and can demonstrate that difference credibly.

AI gives that responsibility new urgency.

When more people can build their own analyses, the expertise required to evaluate those analyses becomes more consequential. Our role as communications measurement professionals is to establish what the evidence supports, what remains uncertain, and what an organization should do next.

The risk of measuring the wrong thing, faster

Generative engine optimization, or GEO, offers a useful example. The AMEC AI Day agenda was packed with GEO-everything. As organizations examine how they appear in AI-generated answers, they need to understand whether potential customers and other stakeholders encounter them, how they’re represented, and what that might mean for their business.

The questions selected for testing shape everything that follows.

A program built entirely around prompts a brand hopes its audience is asking may produce a reassuring visibility score while overlooking important topics. The brand could be appearing in answers about an unexpected use case, a misunderstood product feature, or a concern the communications team has not considered.

Examining observed AI answers about the brand and its category can help uncover those contexts and inform a more relevant testing program.

That requires care about what the evidence represents: which platforms, markets and interactions were observed, and where coverage is incomplete. A share of collected answers should not casually become a claim about the share of people seeing a brand.

Claims about GEO need to withstand the same scrutiny we would apply to any other measurement approach. So the next time you’re talking GEO within your business, ask whether the prompts that ran are based on actual AI answers your brand showed up in, or whether a machine (or your colleague) generated them for you.

Innovation needs investigation

The same discipline applies when AI helps generate research findings.

In one piece of AI-assisted work I reviewed, data had been included that didn’t exist. Catching that problem required someone to question the output and check the evidence behind it.

There’s also considerable opportunity when new methods are used thoughtfully. Among the work recognized at the AMEC Awards was a project in which a synthetic audience surfaced a finding that helped inform a larger strategic decision about a launch. Pairing that work with primary research was essential to understanding the finding and the differences across audience segments.

That’s where curiosity matters. An unexpected result should prompt investigation: Why are we seeing this? Does another source support it? Could the method itself be producing the pattern? What would we need to know before recommending action?

Those questions make innovation useful.

A standard for decisions

Earlier this year, FleishmanHillard developed its five principles of decision-ready intelligence to address the challenge of using AI in communications intelligence. They connect the quality of the evidence to the decisions people make with it.

The conversations at AMEC reinforced the importance of carrying that discipline through the entire process. Sound data foundations are essential. The work continues through interpretation, recommendation, and action.

For communications leaders, this means expecting clear answers from their teams and partners. Where did the evidence come from? Why is this the right method? What could change the conclusion? What decision does the finding help us make?

For our profession, it means making our judgment visible and useful at the point where decisions are being shaped.

I left AMEC encouraged by the people committed to that work. The future of communications measurement holds considerable opportunity, and realizing it will require us to keep connecting new capabilities to meaningful questions and defensible decisions.

AI expands what we can measure. We remain responsible for establishing what matters and helping organizations act on it.

From Left to Right: Catalin Anton Hemming, Caroline Wright, Ben Levine, Ines Schumacher, EJ Kim, Jordan Downey, Liam Heaton, Roxana Tintea, Nicole Harris and Peter Lakel