Beyond Gut Feeling: What 20,000+ Analyst Interactions Teach Us About Measuring AR

Twenty thousand analyst interactions give you perspective. They do not give you permission to invent a benchmark.

Kea has completed more than 20,000 analyst interactions across clients, technology markets and regions. That creates a substantial body of operating experience: different analysts, different objectives, different markets, different levels of relationship maturity and very different outcomes. It gives us context when we assess whether an interaction was useful, whether a pattern is emerging and what a programme should do next.

What it does not give us is one clean statistical study in which every briefing had the same objective, every analyst started with the same level of knowledge and every company faced the same market problem. Pretending otherwise would create false precision. The useful benchmark is therefore not an industry average invented after the fact. It is a disciplined way of comparing like with like, separating activity from progress and asking whether the analyst’s understanding of the company actually moved.

That distinction matters because Analyst Relations is unusually vulnerable to reporting theatre. A calendar full of briefings looks busy. A friendly analyst sounds positive. A dashboard with percentages looks precise. None of those things automatically tells leadership whether the market understands the company better, whether the evidence is stronger or whether the programme is helping the business make better decisions.

What Analyst Relations benchmarking should actually measure

At its simplest, Analyst Relations benchmarking is the comparison of AR performance against a useful reference point. That reference point may be your own previous interactions, a defined objective, another relationship at a similar stage or a wider body of relevant experience. The purpose is not to produce a prettier score. It is to make judgement more disciplined.

The first mistake is to confuse activity with performance. Briefings completed, inquiries arranged, analysts contacted, reports tracked and follow-ups sent are all useful operational metrics. They tell you whether the programme is doing the work it committed to do and where capacity is going. They do not tell you whether the work changed anything.

A company can complete twenty briefings and still leave analysts unclear about where it fits. Another can run a smaller programme and make significantly more progress because the right analysts received the right evidence at the right moment. Operational metrics measure what the programme does. Performance metrics should measure what the programme accomplishes. Both matter. Treating them as the same thing is where weak AR reporting starts.

The second mistake is trying to measure an interaction without first defining its purpose. “Brief Gartner” is an activity. “Make sure the analyst understands that the company now supports enterprise deployments in regulated environments, with current customer evidence to substantiate the claim” is an objective. The second can be tested afterwards. The first can only be counted.

That objective may be to correct an outdated understanding, establish a new use case, demonstrate that a previously planned capability is now generally available, give the analyst stronger customer evidence, understand why the analyst sees the market differently from leadership, or prepare the business for a defined research process. The objective does not guarantee the outcome. It gives the programme something meaningful to assess.

This is why a good benchmark starts before the call. You need to know what the analyst already understands, what remains unclear, what evidence you intend to provide and what movement would count as progress. Without that baseline, the post-call discussion tends to collapse into instinct: “I thought it went well”, “the analyst seemed engaged”, “they asked good questions”. Those observations are not useless. They are simply insufficient.

What 20,000+ interactions can—and cannot—tell you

A large body of experience becomes valuable when it helps you recognise context. A first briefing with an analyst who barely knows the company should not be judged against a strategic update with someone who has followed the business for five years. A product demonstration should not be assessed against the same objective as a market-positioning discussion. An evaluation-related conversation has different constraints again.

This is where experience across thousands of interactions matters. It helps distinguish normal friction from a genuine problem. It helps us recognise when an analyst is asking the kind of questions we would expect at that relationship stage, when a narrative problem is repeating across firms, when a spokesperson is consistently creating clarity and when the company is mistaking politeness for progress.

But it is important to keep the claim proportionate to the evidence. Kea’s 20,000+ interactions are operating experience, not one uniform research dataset. We should not pretend that every interaction can be collapsed into an industry average or that a single score predicts revenue. The value lies in accumulated context: comparable situations, recurring questions, repeated evidence gaps and a much larger reference frame than any one company can build from its own recent activity alone.

That context also helps avoid the opposite error: overreacting to a single difficult meeting. One analyst challenging a claim is information. Several relevant analysts independently challenging the same claim is a pattern worth investigating. A company should not rewrite its positioning every time somebody asks an uncomfortable question, but it should not dismiss the same question appearing repeatedly because leadership dislikes the answer.

Benchmarking therefore works best when it creates proportion. It tells you when something is merely different, when it is consistently weak and when it is improving in a way that deserves more investment.

A practical model for benchmarking analyst engagement

You do not need a complex scoring system to make AR measurement substantially better. For every priority interaction, six elements are enough to create a useful record.

1. Objective: What should be different after this interaction?
2. Starting point: What did the analyst already understand, question or misunderstand?
3. Evidence: Which important claims did the company substantiate?
4. Movement: What became clearer, remained unresolved or changed?
5. Follow-up: What commitments were created and who owns them?
6. Business use: What should AR, product, marketing, sales or leadership do differently as a result?

The strength of this model is that it keeps the meeting connected to the programme. Relevance matters because a perfectly delivered briefing to the wrong analyst is still a poor allocation of time. Evidence matters because a strong narrative unsupported by proof remains fragile. Movement matters because a pleasant discussion may leave the analyst’s understanding exactly where it started. Follow-up matters because a good interaction can still lose credibility if promised information never arrives. Business use matters because feedback that disappears into meeting notes is not yet insight.

This also gives us a better way to think about analyst briefing quality. Quality should be judged against the objective, the analyst’s starting point, the relevance and strength of the evidence, what changed during the discussion and whether the agreed follow-up happened. Briefing count and analyst friendliness are not substitutes for any of those things.

The same logic applies when people ask whether analyst relations benchmarking can prove ROI. It cannot do that by itself. Interaction quality and perception movement can show whether an AR programme is improving. Commercial contribution needs separate evidence from buyers, sales teams and opportunities. A strong briefing score is not proof that AR created a deal, just as a low score does not prove that an opportunity was lost because of Analyst Relations.

That distinction is central to our view of Analyst Relations ROI: contribution should be documented without pretending that one activity caused the full commercial outcome.

Follow-up belongs in the same framework. An analyst can ask for a customer example, technical clarification or additional evidence and receive nothing for three weeks. The call itself may have been strong; the programme performance was not. The separate guide on what happens after the briefing goes deeper into that ownership problem, but from a measurement perspective the principle is simple: promises made during the interaction are part of the interaction.

The patterns that matter more than one good or bad call

The most useful insight rarely sits in one meeting. It sits in repetition.

If analysts across different firms repeatedly struggle with the same category definition, the problem may not be the analysts. If customer evidence consistently changes the quality of the discussion, build more of it. If one spokesperson repeatedly creates clarity while another produces the same unanswered questions, preparation or role selection deserves attention. If the same competitor keeps appearing in analyst comparisons while the company insists that competitor is irrelevant, leadership may need to revisit its assumptions.

Patterns are also where benchmarking becomes operational. They help the programme decide whether to increase executive time with a relationship, change the briefing narrative, strengthen customer proof, rethink the analyst map or stop an activity that produces motion without movement.

This is why comparing like with like matters. A lower score can reflect poor preparation, but it can also reflect a harder assignment: a new analyst, a difficult market claim, a weak evidence base or a relationship that is only beginning. The purpose of benchmarking is not to remove judgement. It is to make the judgement visible enough that leadership can question it.

Set targets against your own objectives before reaching for generic industry averages. A benchmark tells you where you are. A target tells you where you intend to go. They are not the same thing, and pretending they are usually creates the kind of dashboard precision that looks impressive while directing attention to the wrong work.

What leadership should expect from AR measurement

Leadership does not need another activity report. It needs a view of the programme that supports decisions.

A useful AR scorecard should therefore separate at least three things: delivery, understanding and contribution. Delivery tells leadership whether commitments were completed. Understanding records what analysts demonstrably know, question or misunderstand. Contribution captures where AR or analyst research was used in a commercial or internal decision, with enough evidence to keep attribution proportionate.

These indicators should not be collapsed into one universal “AR effectiveness” percentage. A strong delivery score can hide a serious evidence gap. A busy quarter can coincide with no meaningful change in market understanding. Commercial support can be valuable without proving sole causation of revenue.

The more useful management question is always: what should we do differently because of what we now know?

That may mean putting product time into two weak proof points instead of booking five more briefings. It may mean moving an analyst up or down the priority list. It may mean changing the spokesperson, reworking the category explanation, preparing a stronger customer example or deciding that a recurring interaction no longer deserves the time it consumes.

If a metric cannot help somebody make a decision, question why you are collecting it.

For a more explicit leadership model, see An AR scorecard your leadership team can question. The scorecard and the benchmark serve the same purpose: make progress visible without pretending certainty the evidence cannot support.

Where Kea uses benchmarking

Kea uses experience across more than 20,000 analyst interactions as a reference point for preparation, programme reviews and interaction assessment. The value is not a league table of clients, and it is not a claim that all 20,000 interactions form one uniform research sample. They do not.

The value is accumulated context.

We can compare a current interaction with relevant patterns from other markets, objectives and relationship stages; challenge whether a “good call” actually achieved anything; identify when the same misunderstanding is repeating; and help the team decide what to change before the next conversation. That context improves the quality of judgement without turning judgement into fake mathematics.

It also creates a more useful conversation with leadership. Instead of reporting that ten briefings happened, the programme can explain what those interactions were meant to change, where understanding moved, which evidence gaps remain, which commitments are still open and what the business needs to do next.

That is the benchmark advantage: less dependence on gut feeling, more disciplined judgement and a clearer connection between analyst engagement and the decisions the programme needs to make.

If your AR reporting is still mainly a count of meetings, start with one question: what changed because those meetings happened?

Then build the measurement around the answer.

Explore Kea’s Analyst Relations as a Service if you want benchmarking, preparation, engagement and follow-through managed as one programme.

Put this into practice

Give analyst engagement an owner and a next step.

Build preparation, follow-through and measurement into a sustained programme around your business priorities.

Explore AR-as-a-Service ↗

Not sure where to start? Take the eight-question AR Readiness Check.

Share this article