Direct Traffic Is Changing: What 12 Businesses Reveal About AI-Era Search Discovery
When AI recommends a business and someone types the URL directly, analytics tends to call it direct traffic, mostly.
Read more
Context Map
What does AI know about your brand?
Context Map is an AI SEO audit. We ask the questions your customers ask across AI assistants, Google and social search, record what comes back, trace where those answers came from, and compare all of it with what is true about your business.
From $1,500 · Delivered in 10 business days
Deliverables
What you receive
The Context Map.
One row per finding: the buyer question, what AI and search concluded, what is true, the gap type, the source behind it, the competitor involved, and the recommended move.
The evidence sheet.
Every question, every answer, both runs, dated, with the sources cited. It is yours to keep and re-run.
Where to start.
The 3 to 5 moves most worth making, ranked.
A 60-minute readout call.
We walk through the findings with you.
- 40
- buyer questions
- 9
- AI, search and social surfaces
- 2
- runs, a week apart
- 3
- competitors compared
- 10
- business days to delivery
The idea
The Context Gap
What is true about your business
What exists online
What AI and search can find
What they conclude
What your customer hears
Every break in that chain is a Context Gap. Software can tell you how often your name appears. A Context Map shows why AI and search formed the picture they have of you, and where that picture differs from what is true. Visibility metrics are the evidence. The Context Map is the diagnosis.
Gap types
What a Context Map finds
GAP 01
Wrong
AI states something that isn't true.
GAP 02
Stale
AI repeats something that used to be true.
GAP 03
Invisible
The information exists publicly but isn't being picked up.
GAP 04
Unpublished
It's true and it matters to buyers, but there is no public evidence for it.
GAP 05
Conflicted
Credible sources say different things about you.
GAP 06
Misclassified
AI has the facts but files you in the wrong category.
GAP 07
Unsupported
You make a claim with too little independent evidence behind it.
GAP 08
Competitor-owned
A competitor has stronger evidence on a topic you should own.
Where findings go
Each gap type points at a service.
Wrong, Stale, Conflicted and Misclassified gaps, where the record itself is the problem. Entity SEO →
Unpublished, Unsupported and Competitor-owned gaps, where the evidence has to be written. Citation-Ready Content →
Invisible gaps, where the information exists and the AI assistants are not picking it up. AI SEO / GEO →
Gaps in Google's answer features, AI Overviews and AI Mode. Answer Engine Optimization →
Gaps on Reddit, YouTube, TikTok and Instagram search. Social Search Optimization →
Surfaces
Nine surfaces, two runs
ChatGPTGeminiPerplexityGoogle AI OverviewsGoogle AI ModeRedditYouTubeTikTokInstagram search
Process
How it runs
Interview. A 60-minute session with someone who knows the business: what you sell, who you serve, what makes you different, what has changed, and which facts matter when a buyer decides.
Build the question set. 40 questions across discovery, comparison, recommendation and purchase, branded and non-branded.
Run one. Every question across ChatGPT, Gemini, Perplexity, Google AI Overviews, Google AI Mode, Reddit, YouTube, TikTok and Instagram search.
Run two. The same questions a week later, so you see what held and what changed.
Trace the sources. We follow the claims, omissions and competitor recommendations back to where they came from.
Classify and rank. Every gap gets a type, and the moves get ranked.
Readout. 60 minutes with you, ten business days after the interview.
Next step
Ten business days from the interview to the readout.
Example
What a finding looks like
Illustrative example. Not a client.
- Buyer question
- Who manufactures titanium components for medical devices in the Midwest?
- What AI said
- Brand absent. Three competitors recommended.
- What is true
- The company has made titanium medical components since 2019, and medical work is a significant share of current production.
- Gap
- Unpublished
- Why
- The capability appears in sales materials but not on the public website or in any outside source.
- Competitor evidence
- Competitor A has a dedicated capability page, two distributor references and an industry association profile.
- Move
- Publish the capability on the site, connect it to the company's product information, add outside corroboration, and re-test.
Case Study
A Hawaiian brand older than the internet. AI had one frame for them: tourist convenience.
Elvis was rumored to have worn their stuff. The brand predated Google by decades, predated the web entirely, and had been part of Hawaiian culture long enough that locals remembered specific locations the way people remember places that shaped them. When they encountered the brand today the reaction was always the same: "Oh, they're still around?" Fond memory. Mild surprise. No purchase intent.
The digital footprint told a different story. Cruise ship shuttle stops. TripAdvisor listings filed under shopping near the port. A brand with cultural provenance had been indexed, cited, and recommended as a tourist convenience. Nationally it still moved volume on nostalgia, surfacing in niche searches for Hawaiian clothing. The context AI systems were working from had collapsed decades of real history into a single frame: souvenir shop.
The Context Map found the problem and the asset. Buried in the entity record were citation trails worth killing for. Vintage press, cultural references, decades of organic mentions from sources with authority. None of it was connected, none of it was current, and none of it was pointing anywhere useful. The first cycle wrote itself: stop starting from scratch and start activating what already existed. Legacy authority redirected at a modern brand presence.
- AI frame
- Tourist souvenir shop near the cruise port
- What is true
- A long-established Hawaiian brand with real cultural history
- Gap
- Misclassified, Stale
- Why
- The findable evidence was cruise and tourism listings. The historical authority existed but was not connected to the current brand.
- Move
- Reconnect the existing authority to the current brand instead of building authority from scratch.
Sample map
What a Context Map looks like
Illustrative rows. Not a client.
| Buyer question | What AI concluded | What is true | Gap | Source | Competitor | Move |
|---|---|---|---|---|---|---|
| Can they do X? | No evidence | Yes | Unpublished | None found | Acme | Publish the capability with evidence |
| Are they local? | Unclear | Yes | Misclassified | Directory A | Beta | Fix the business and location record |
| Do they serve Y? | No | Yes | Stale | 2019 article | Gamma | Correct it and add current corroboration |
Next step
Every row is a decision you can make with evidence.
Pricing
Context Map
From $1,500
- 40 buyer questions, branded and non-branded
- 9 surfaces: ChatGPT, Gemini, Perplexity, Google AI Overviews, Google AI Mode, Reddit, YouTube, TikTok, Instagram search
- Two runs, a week apart
- Up to 3 competitors
- 60-minute interview
- The Context Map, the evidence sheet, and where to start
- 60-minute readout
- Delivered in 10 business days from the interview
The final price depends on how many locations, products, markets or competitors you want covered. We confirm scope and price on a short call before anything starts.
Not a visibility score. Not monthly reporting. Not a 60-page list of website fixes.
90-Day Recheck, $750
Same 40 questions, same surfaces. We document what changed, what held, and any new gaps.
FAQ
Questions buyers ask about the Context Map
What is a Context Map?
An AI SEO audit. It shows what AI assistants, Google and social search tell your customers about you, where those answers come from, and where they differ from what is true about your business.
How much does a Context Map cost?
From $1,500. The price rises with the number of locations, products, markets or competitors covered. We confirm scope and price on a call before work starts.
How long does it take?
Ten business days from the interview to the readout.
Which platforms do you test?
ChatGPT, Gemini, Perplexity, Google AI Overviews, Google AI Mode, and search on Reddit, YouTube, TikTok and Instagram.
Why run every question twice?
AI answers change from one run to the next. Two runs a week apart separate a pattern from a one-off.
What do I need to provide?
An hour for the interview, your website, up to three competitors, and the products, services and markets that matter most.
Is this an AI visibility tool or score?
No. We use visibility data as evidence. The finding is why AI and search describe you the way they do and what would change it.
Do you fix what you find?
The Context Map ends with where to start. Fixing is separate work, done by your team or by us.
What is the 90-Day Recheck?
The same 40 questions on the same surfaces three months later, $750. It shows what changed, what held and any new gaps.
What happens after the Context Map?
Clients use it to decide what to fix first. Our measurement framework covers what we track over time.
From the Field
Find out what AI tells your customers about you.
From $1,500. Ten business days. You keep everything.