Skip to content

Articles / Proving value

How to report AI visibility to leadership: the weekly brief, with a template

A leadership report on AI visibility has five parts: the rate with its range, the change, competitor gaps, top sources and next steps. Free template.

Published
Reading time
9 min read
By
AmpliRank
Tags
AI Visibility · Reporting · Weekly Brief · Template

Report AI visibility to leadership as a short weekly brief in five parts: the headline rate with its range and answer count, the change since the last run with its range, the gaps with competitors, the sources the engines cite most, and what the team will do next. Each figure says whether it is a clear change or within the noise, so leadership can act on the clear ones and ignore the rest.

In short

  • Lead with one number: the share of answers to unbranded questions that name you, with its 95% range and the number of answers behind it.
  • Report a change only as clear or not clear: clear when the range of the difference, on the same questions and engines, excludes zero.
  • Show competitors on the same answers, with the gap’s range and a verdict for each.
  • Name the sources the engines cite most, because that is where the next piece of work usually is.
  • End with actions tied to the evidence, an owner and a date. Report outcomes as observed, never as caused.

This guide covers:

What should a leadership report on AI visibility contain?

A leadership report on AI visibility should contain the five things a decision needs: where you stand, whether it moved, against whom, why, and what happens next. Anything more belongs in an appendix for the team that does the work.

Section What it says What makes it trustworthy
Headline How often the engines name you on unbranded questions The 95% range and the number of answers
Change Whether the rate moved since the last run A paired range on the same questions and engines; “clear” or “no clear change”
Competitor gaps Who is named more or less often than you Every brand counted in the same answers
Top sources Which sites the engines cite in your category Counts out of the same answers, with ranges
Next steps The pages to write and sources to earn a mention on Each step linked to the evidence behind it

The terms in this guide are defined in the AI visibility glossary, which is a useful link to put at the foot of the first brief you send.

How do you state the headline number?

State the headline as a rate, its range and its answer count in one line: “36.5% of answers named us, range 30.0% to 43.5%, from 192 answers to unbranded questions.” The examples in this guide use sample data for Taskmoor, a fictional brand; the brief would carry your figures.

Three rules keep the headline honest:

  1. Unbranded questions only. Questions that name you get answers that name you. Branded vs unbranded questions shows a blended rate rising 19 points with no change in discovery.
  2. The range travels with the rate. In charts, in slides and in the subject line of the email. A rate without its range invites decisions on noise; how many answers do you need? shows how the range narrows with more answers.
  3. The same definition every week. Same question set, same engines. When the set changes, start a new series and say so.

The pillar guide on measuring AI visibility explains the measurement behind the number.

How do you report a change?

Report a change as clear only when the 95% range of the difference between two runs, on the same questions and engines, excludes zero. Otherwise write “no clear change” and give the range anyway.

In the sample, last week named Taskmoor in 58 of 192 answers (30.2%) and this week in 70 of 192 (36.5%). Compared answer by answer, 21 answers gained Taskmoor and 9 lost it, so the change is +6.3 points, range 0.6 to 11.9: a clear change. The line in the brief reads: “Up 6.3 points since last week, range 0.6 to 11.9 points, on the same questions and engines.”

The same headline rise can be noise when answers flip in both directions; is that change real? shows both cases side by side. Most weeks, the honest line is “no clear change”, and a brief that says so builds the trust that makes the clear weeks count.

How do you show competitor gaps?

Show each competitor’s rate on the same answers as yours, the gap with its range, and a verdict. In the sample run, every brand is counted in the same 192 answers:

Brand Named in Rate Gap with Taskmoor 95% range of the gap Verdict
Workloft 84 of 192 43.8% Workloft leads by 7.3 points −1.5 to 16.1 points no clear gap
Taskmoor 70 of 192 36.5%
Stackwell 60 of 192 31.3% Taskmoor leads by 5.2 points −3.4 to 13.8 points no clear gap
Planhaven 42 of 192 21.9% Taskmoor leads by 14.6 points 7.1 to 22.1 points clear gap

Read the table from the verdict column. Taskmoor is clearly ahead of Planhaven; against Workloft and Stackwell, the sample can’t separate the brands yet, even though Workloft’s rate is seven points higher. A brief that ranks the four by headline rate alone would overstate both gaps.

How to compare your AI visibility with competitors fairly explains why the gap must be measured on the same answers, and how share of voice can mislead.

Which sources belong in the report?

The sources that belong in the report are the sites the engines cite most often in your category’s answers, especially in answers that name a competitor and not you. They show where the decision about who gets named is made.

Source (sample) Cited in Share of answers 95% range
pmreviews.example (a “best tools” roundup) 38 of 192 19.8% 14.8% to 26.0%
workloft.example (a competitor’s site) 22 of 192 11.5% 7.7% to 16.7%
agencyforum.example (a discussion thread) 17 of 192 8.9% 5.6% to 13.7%
taskmoor.example (your own site) 9 of 192 4.7% 2.5% to 8.7%

Keep the list to the top five, and mark what kind of site each is: a roundup, a competitor, a forum, a publication, your own site. Each kind calls for different work. The searches the engines ran before citing them, a step Google calls query fan-out, show which sub-questions those sources answer.

How do you turn the figures into next steps?

Turn the figures into next steps by linking each action to the answers and sources that justify it, with an owner and a date. Leadership approves actions more readily when the evidence comes with them.

From the sample, three actions follow:

  1. Earn a place on pmreviews.example. It is cited in about a fifth of answers and lists Workloft. Owner: content lead. Date: end of month.
  2. Answer the sub-question the engines search and Taskmoor’s pages don’t. For example, a page on agency resource planning, if that is what the engines searched. Owner: SEO lead. Date: in two weeks.
  3. Join the agencyforum.example thread honestly, as the company, where Taskmoor can genuinely help. Owner: product marketing.

Then report each action’s status in the next brief, and any change in the rate as observed. The sample report shows how AmpliRank lays out actions with their evidence.

The weekly brief template

The template below fits on one page and works without any tool: fill in the brackets from your own counts. Print this page to keep a copy, or download it as Markdown or as a CSV for a spreadsheet.

Download the template as Markdown or as a CSV.

To fill in the ranges, use the free AI visibility calculator: its range tab gives the headline range, and its gap tab gives the range of a change or a competitor gap from four counts.

How often should you report?

Report as often as your runs can show a change, and no more often. A weekly brief suits a weekly run large enough to see the changes you care about; with small runs, a weekly brief will mostly say “no clear change”, which is still useful as a heartbeat. How often should you measure AI visibility? gives the smallest clear change for each run size.

A common rhythm is a weekly brief for the team doing the work, and a monthly summary for leadership that compares month against month on the same questions and lists the actions completed.

What can the report not claim?

The report can’t claim that your work caused a change, that you will be named, or that visibility produced revenue. It can claim what was observed, with its range.

  • Observed, not caused. A clear rise after you published a page is reported as a rise since that date. Engines, competitors and third-party sites change in the same weeks.
  • These questions only. A rate describes the questions asked, on the engines asked, on the dates asked. It is not a market share.
  • Not pipeline. AI referral visits and conversions can sit next to the visibility rate, but the link is observed, not proven. Can AI visibility be tied to pipeline? explains what can be shown.

Where to start

Download the template and fill in this week’s brief from your last run, even if it is small: one headline with its range, one change line, your closest competitor and your top three sources. Send it to one person in leadership and ask which line they would act on. AmpliRank writes this brief from each run’s figures and shares it by link; the sample report shows one.

Sources

■ Ask · Measure · Compare · Trace · Fix · Share

Measure what AI says about you.

Cookies