One Brand OS. One loop.
Measure your odds of being selected, fix the biggest gap, and measure again. Every product reads the same BrandGraph.
How RocMode works →Know where you stand
SelectabilityYour odds of being picked, moment by moment BrandYour BrandGraph and strategy, every fact sourcedProtect what’s true
BrandGuardEvery claim checked before it shipsDo the work
Website EngineYour site, rebuilt from the truth Email StudioBroadcasts and sequences, checked before they send Pitch EngineA credentials deck for each buyer, tracked Document StudioProposals and PDFs that stay in syncPlan and follow up
RFP EngineA brief agencies can answer, written from your record CampaignsA year of campaigns, judged on what you signed Journeys and SequencesFollow-up that never forgets
Get your Selectability Score
Who gets picked instead of you, and the three gaps to fix first. Free, no login.
Get your scoreSeen isn't selected.
Selectability is the odds that a buyer, at a given moment, picks you over the alternatives, doing nothing included.
What is Selectability? →The model
The seven stagesKnown to Kept. Being seen covers the first two Moments and buying modesOne kind of buyer, at one step, buying one way The loopMeasure, fix the biggest gap, ask again each quarterThe Score
How the Score worksv1.0The formula, the weights and the honesty rules, published Sample reportA full read for an invented brand, page by page GlossaryCompare, fairly
vs AI-visibility tools vs an agency vs doing it yourself vs doing nothingEach page says when the other option is the better choice.
Gate × quality, moment by moment, with a range. Questions that name you only test belief.
Read the method →One argument, three sizes.
The book tells you why, the playbooks tell you how, and your Score tells you where. All three are free.
The books
The manifestoFreeTwelve beliefs in ten minutes. Short enough to send your CEO Seen Isn't SelectedFreeMarketing's one job in the AI era. The whole book, free to read here The Selectability PlaybooksFreeEleven steps, fourteen builds, every templateResearch
Customer ZeroRocMode's own Score, published monthly, whether or not it improves Share your scoresHelp test the Score against real sales. Findings published, whatever they show Guides and findingsEvery finding with its source, and how strong it isEvery second Tuesday
The SelectionOne finding, one fix, one receipt
The whole book, free, right here.
All 206 pages, chapter by chapter. No email, no paywall.
Start readingWe make brands alive.
Brands are selected by buyers, and now by the machines that advise them. We hold ourselves to it first.
RocMode
AboutWhy we built it, and who built it Customer ZeroOur own numbers, published monthly TrustWhat we do with your brand's dataEvery buyer, every quarter.
The beta is limited to 30 brands. Free while it runs, and your price is held for 12 months after.
Request Beta access →Know where you stand
Selectability BrandProtect what’s true
BrandGuardDo the work
Website Engine Email Studio Pitch Engine Document StudioPlan and follow up
RFP Engine Campaigns Journeys and SequencesFor you
Marketing leaders Founders and CEOs Sales leaders Agencies and consultants Smaller businesses How RocMode works →Problems we solve · AI doesn’t name us
AI doesn’t name us.
Your buyers ask an assistant who to shortlist, and your name isn’t in the answer. Here’s why that happens, what to fix first, and why being named still isn’t being picked.
It happened to us first. On 29 August 2026, AI assistants named RocMode in none of the 30 questions our buyers ask. We publish every month what we change, and what it does.
Four reasons AI leaves you out.
Crawlers blocked, no sitemap, prices and details hidden behind a call. The assistant has nothing to quote.
Buyers ask specific questions. A page about how great you are doesn’t answer any of them.
Assistants lean on what others say about you. A claim with no outside source is easy to leave out.
For every question you leave open, a rival’s page closes it, and gets named instead.
Asked the way buyers ask.
Your main buyer’s questions, put to five assistants three times each, without your name in them. Every answer read, every rival counted.
Selectability ScoreMonday by Nine · read 7 Oct 2026 · method sel-score@0.1Download the PDFYour buyer's questions, put to the machines.
“Are hospitality-specialist bookkeepers worth paying more for?”
This answer named others.“What reports should a café owner get from their bookkeeper?”
This answer named you.“Can a bookkeeper send me wage cost against takings every Monday?”
This answer recommended you.“Best bookkeeper for cafés in Brisbane that reports wages weekly”
This answer named others.Being named is one stage
Found is a gate: if you’re not found, the score is zero, because zero times anything is zero. But a brand can be named in every answer and still lose at Answered, Believed, Preferred or Bought. Fix Found first, then keep going.
Ask the questions yourself.
Questions a buyer asks before they know your name. Try a few in any assistant, about your category.
“Which agency should run our rebrand?”
Named others“How do I know if a supplier like this is any good?”
Named others“Is it worth paying more for a specialist?”
Named no oneExample questions. Your free score uses your main buyer’s own.
Get named, then get picked.
Start free, by hand. Bring RocMode in when you want it measured every quarter and the work done.
- Check an assistant can read your site: crawler access, a sitemap, an llms.txt file.
- Publish what buyers ask about: prices, how it works, who it’s for, who it isn’t for.
- Put a source next to every claim, ideally someone else’s.
- Ask the same questions again in a quarter, without your name in them.
With RocMode
Named, or not.
Five assistants, three times each, with who gets named instead. Then the Website Engine and the Brand Protocol close the gaps.
Questions, answered.
Can you get us into ChatGPT’s answers?
Nobody can guarantee that, and we won’t pretend to. We measure how often you’re named and picked, fix the biggest gaps, and measure again.
Is this the same as AEO or GEO?
Those practices raise how often AI mentions a brand. That’s the Found stage. RocMode measures Found and the five stages after it, because being named isn’t being picked.
Why don’t questions with our name count towards being found?
A buyer who already names you has already found you, so those questions count only towards Believed: they test whether what you claim holds up. Found is measured on questions without your name.
You were at zero. Why trust you?
Because we said so in public, and we publish what changes every month. Customer Zero shows every scoreboard, up, down or flat.
Find out if AI names you.
Your main buyer’s questions, without your name in them, asked of five assistants. Who gets named instead, and what to fix first. Free.