The work your answers point to, ranked.
It starts where the engines started.
Before an engine names anyone, it searches, looks up a shortlist and reads pages. The Work Plan keeps that trail as it happened.
They searched.
The searches each engine ran, word for word.
They looked up a shortlist.
The companies each engine looked up by name.
They read pages.
The pages behind the answers. CitePack reads them too and shows where you and the company listed first sit on each.
Two counts, kept apart.
For each question and each engine: how many answers named you, and how many named you first. Nothing is folded into a single verdict.
Every item carries the same seven fields. An item with no answers behind it doesn’t go in.
- Question
- The buying question it comes from
- Answers
- The engine and repeat of each answer behind it
- Evidence
- The search, page or quote that shows it, word for word
- Kind
- Which of the nine kinds of work it is
- Effort
- How big the job is
- Owner
- Who on your team does it
- Check
- What to look at when you Measure Again
Each item is one of nine kinds.
Get onto a list AI reads
A list the engines used names other companies and leaves you out.
Move up a list you’re on
You’re on a list the engines used, below the companies they name first.
Make your own page the one AI uses
The engines read other pages for a question your own page should answer.
A page that doesn’t exist yet
An engine searched for something no page on your site answers.
Build the source
Another company comes first and no page of yours was used. You get the outline of the page to build.
Correct what AI gets wrong
An answer says something about you that your own site says differently.
Your words beside the leader’s
The words the answers use about you, next to the words they use about the company listed first. What to do about it is your call.
Technical faults
A fault on a page tied to a question where another company comes first, such as a page an AI crawler can’t read.
Communities
Forum threads the engines used, where your own people can reply under their own names.
No page of yours was used? Here’s the outline.
When another company comes first and the engines used none of your pages, the Work Plan sets out the page to build, in the shape of the ones they did use.
- 01A title with the year in it
- 02Written for the engines’ exact searches
- 03The columns the lists they used have in common
- 04The entries, in the order the engines named them
- 05Your row: quotes from your own site, or a blank your team fills
- 06An FAQ drawn from those searches
- 07A visible date line
The Action Pack carries the page draft. Facts about you are quoted from your site, and nothing about a rival is stated for you.
Beside every item, the material to do it.
Each Work Plan item comes with its piece, grouped by kind and ready to use or send. Do the work and mark the item done.
Email drafts
The pitch or the updated facts for each page the engines used.
Page drafts
The outline of each page to build, from the engines’ own searches.
Rewrites
The one claim you want the answers to repeat, in every place it belongs.
Correction notes
One line for each fact the answers state differently from your site.
Technical fixes
What stops the AI crawlers reading a page, and what to change.
Every check, answered on the new map.
Measure the same decision again and the Work Plan’s checks are answered beside what you did.
The same decision
The same sealed questions, four engines and buyer market, on Day 28 or X, the day you choose. We recommend Day 28.
Each item’s check
Answered on the new map, beside what you marked done. An item you did not do shows as not done.
The Before/After Snapshot
What moved, or that nothing did, engine by engine.
What makes it checkable.
- Every item names the answers behind it
- Every search and quote is shown word for word
- Every page is one the engines used
- Every item carries its own check
- A score
- A guess at why an engine answered as it did
- A promise that the answers will change, or that a change caused a move
- Facts about another company stated for you
- A generic list of fixes
Your map, Work Plan and Action Pack reach your account 30 minutes after you approve your questions, on average. We email you when they’re ready. If they aren’t delivered within 24 hours, you get a full refund.
Be first to run
your Decision Loop.
Join the waitlist. The first ten companies get their own sign-in and the founding price. How it works
Or talk to us about your offering.
$99$149
₹4,999₹7,499
One Decision Loop for one buying decision
A third off for the first ten companies. After those ten, the standard price is $149, or ₹7,499 if you pay from India, for one loop.
- One offering in one buying category and one buyer market
- Five buying questions you approve, across the five buying situations
- 120 answers60 on Day 0 and 60 when you Measure Again, from ChatGPT, Claude, Gemini and Perplexity
- Two Decision MapsThe Overview and a page per question, with where you were put forward and your place in each answer
- Who gets pickedThe other names in the answers, and who comes first
- What AI says and AI enginesThe words used about you, the facts stated and the pages behind the answers
- Evidence & TrustEvery answer in full
- Two Work PlansThe work each map’s answers point to, ranked, each item with the answers behind it and its check
- Two Action PacksThe material for each item, ready to use
- One Measure AgainThe same decision measured again on Day 28 or X, with its Before/After Snapshot
- Ready in 30 minutes on averageAfter you approve your questions. If it isn’t delivered within 24 hours, you get a full refund. We email you when it’s ready.