Get it on the agenda in one page.
You know what AI tools like ChatGPT say about your company. Your exec doesn’t. This one-page memo says what changed, what you measured, what you recommend and what waiting costs, in the terms a planning meeting runs on.
Memo
- To
- [Name, title]
- From
- [Your name, role]
- Date
- [Date]
- Subject
- Where we stand in AI answers, and what I recommend
What changed
Buyers of [our category] are starting to ask ChatGPT before they search. ChatGPT gives one answer that names a few products, and a product it leaves out may never be considered. Our search rankings can look fine while this happens.
What I measured
On [date] I asked [n] questions our buyers ask before choosing [a product like ours], across [which AI tools]. That gave [n] answers.
- We were named in [n of n] answers and recommended in [n].
- Named most often: [competitors, with counts].
- Where we appeared: [how we were described; flag anything wrong or out of date].
- [Questions no page of ours answers, or pages AI tools can’t read].
What it means for us
[One paragraph in your own numbers. Label every figure measured or estimated, and say where estimates come from]
What I recommend
- [The first fix]. Owner: [name].
- [The page to write]. Owner: [name].
- Ask the same questions monthly, in the same way, and report the count. Owner: [name].
What it costs
[Team time and any tools, for one quarter]
What waiting costs
[The competitor already named] is named in [n of n] answers today. Every quarter it holds that place, it’s the option our sales team has to argue against.
One caveat
Nobody controls these AI tools, and nobody can promise a mention or a position. Treat any vendor who offers one as disqualified. What we can do is measure, fix what makes us hard to quote, and measure again. I’m proposing one quarter of that, with a number to judge it by.
Method: [n] questions across [AI tools] on [date], [settings]. Answers change with wording and over time; this is a dated snapshot, not a score.
Template: praised.co/templates/exec-memo
Memo
- To
- Dana Okonkwo, VP Marketing
- From
- Sam Reyes, Growth
- Date
- 8 October 2026
- Subject
- Where we stand in AI answers, and what I recommend
What changed
Buyers of contract software are starting to ask ChatGPT before they search. ChatGPT gives one answer that names a few products, and a product it leaves out may never be considered. Our search rankings can look fine while this happens.
What I measured
On 6 October I asked 12 questions our buyers ask before choosing contract review software, across ChatGPT, Google AI Overviews, Gemini, Claude and Perplexity. That gave 16 answers.
- We were named in 10 of 16 answers and recommended in 2.
- Named most often: Redlinery (8) and Clausewell (6).
- Where we appeared: Gemini said our team pricing isn’t published, which is true, and ChatGPT repeated Clausewell’s claim about clause libraries.
- Three questions have no page of ours to answer them, including both comparisons with a named competitor.
What it means for us
Procurement teams are about a third of our new pipeline (measured, CRM, last two quarters). AI tools answered all three of their renewal questions without naming us. If a quarter of those buyers start with an AI tool (estimated, onboarding survey), some of them are building a shortlist we’re not on before sales hears from them.
What I recommend
- Fill the gaps on pages we have. State team sizes and seat pricing on the legal-teams page. Owner: Priya.
- Write the three missing pages: renewals without a legal team, and a fair comparison with each competitor. Owner: Sam.
- Ask the same questions monthly, in the same way, and report the count. Owner: Sam.
What it costs
Team time only: about four days across Priya and me this month, then a few hours a month for the checks. No new tools this quarter.
What waiting costs
Redlinery is named in 8 of 16 answers today, including two of the three about renewals. Every quarter it holds that place, it’s the option our sales team has to argue against.
One caveat
Nobody controls these AI tools, and nobody can promise a mention or a position. Treat any vendor who offers one as disqualified. What we can do is measure, fix what makes us hard to quote, and measure again. I’m proposing one quarter of that, with a number to judge it by.
Method: 12 questions across five AI tools on 6 October, signed out, UK, default mode. Answers change with wording and over time; this is a dated snapshot, not a score.
Seven parts, one page.
Each part answers a question your exec will ask. Here is what to write in each, beside the worked example.
What changed
A few sentences on why this is on the table now: buyers ask an AI tool, it names a few products, and the rest may never be considered. Most of it is written for you.
In the exampleBuyers of contract software are starting to ask ChatGPT before they search.
What I measured
Which questions you asked, where and when, and how often you were named. Quote how you were described, and flag anything wrong or out of date.
In the exampleNamed in 10 of 16 answers and recommended in 2. Redlinery was named in 8.
What it means for us
One paragraph in your own numbers. Label each figure measured or estimated, and say where each estimate came from.
In the exampleAI tools answered all three renewal questions without naming us.
What I recommend
Three actions at most, each with one owner. Lead with the gap that matters most and costs least to close, and end with how you will measure it.
In the exampleWrite the three missing pages: renewals without a legal team, and a fair comparison with each competitor.
What it costs
Team time and any tools, for one quarter. A small, specific number is easier to approve than an open-ended one.
In the exampleTeam time only: about four days this month, then a few hours a month for the checks.
What waiting costs
Name the competitor already in the answers, and how often it appears. A competitor’s name makes the cost concrete in a way a trend never does.
In the exampleRedlinery is named in 8 of 16 answers, including two of the three about renewals.
One caveat
Say plainly that nobody can promise a mention or a position. It’s true, and saying it first is what makes the rest of the memo believable.
In the exampleWhat we can do is measure, fix what makes us hard to quote, and measure again.
Label every number.
An exec will test the first figure they doubt. Say which numbers you measured and which you estimated, and the memo survives the question.
“AI search is costing us about 100 deals a quarter.”
One number and no source. The first question in the room is where it came from, and the memo has no answer.
“We were named in 10 of 16 answers. If a quarter of procurement buyers start with an AI tool (our onboarding survey), some build a shortlist without us.”
What you checked is measured. What you assumed is marked as assumed, with where it came from. Nobody has to take either on trust.
Get the memo.
Every part of the memo is on this page. The file is the Word template with the filled example beside it, ready to edit for your own exec, with a PDF and a Markdown copy.
One email with your copy, then a note only if the memo changes. One-click unsubscribe on every email. How we use your address.
Keep going.
Questions, answered.
Something else on your mind? Contact us.
For the file, yes. Everything on this page is free to read without one: every part, with the worked example. Your address gets you the memo here straight away and a copy in your inbox, and we write again only if it changes.
Yes. Upload the Word file to Google Drive and open it with Google Docs. There’s also a PDF to read or print, and a Markdown version to paste into a wiki.
No. Every figure in the example is illustrative, written to show the shape. Page one of the file is the empty template.
A memo that pitches a vendor rarely gets forwarded. Our name appears once, in small print at the foot of the template, and you can delete it.
Start with the buyer-question workbook. A small set of questions asked by hand is enough for a first memo, as long as you date it and say how you asked.
It checks a public page’s readiness to be found, understood and cited. Ongoing tracking and work belong in the platform.
Start with your website.
Check how ready your page is to be found, understood and cited by AI.