See what the machines see.
Search engines and AI assistants build a picture of your business from scattered, inconsistent sources — and answer from that picture, not from your website. We audit what they currently hold, repair it at the source, and instrument the whole thing so you can watch it move.
Nobody wrote it. It was assembled.
No model reads your homepage and forms an opinion. It assembles a description from whatever it can find and reconcile — and the failure modes are boringly consistent.
The legal entity, the trading name, the LinkedIn page and the invoice footer disagree. A model that cannot resolve those into one entity hedges, or picks a competitor it can identify cleanly.
What you do, where you operate, who you serve and what it costs live in prose and images. Without structured data they are guesses, and guesses get dropped from answers in favour of stated facts.
Directories, review sites and old press still describe a version of you from three years ago. Models weight those sources heavily precisely because they are third-party — which is exactly why the wrong version persists.
Most teams have no baseline for what assistants say about them, so there is no way to tell whether last quarter's work helped, hurt, or did nothing at all. Unmeasured work becomes opinion.
Four capabilities, one datasheet.
Scoped plainly, so you can tell what you would actually receive. Take one, or take the stack.
Six questions. Then a number.
Tick what is true today. Nothing is sent anywhere — the result is computed in your browser, and it will tell you which capability to start with.
Against a standard retainer.
Not a claim that agencies are bad — a description of what we do differently, so you can check whether it matters to you.
| Dimension | Typical retainer | Techrayon |
|---|---|---|
| Baseline | Set after onboarding, often re-set when results disappoint | Fixed before work starts, never re-set |
| Reporting | Dashboard of platform metrics | Raw assistant transcripts you read yourself |
| Scope | Hours per month across whatever comes up | Named deliverables with acceptance criteria |
| Ownership | Agency-held accounts and tooling | Everything in your name from day one |
| Exit | Notice period, then knowledge leaves | Documented system; we hand over and stop |
| Honesty | Upsell the next service | We say when a capability won't move your numbers |
The terms, stated up front.
Read this before you talk to us. If any line is a problem, it is cheaper for both of us to find out now.
- Entry point
- Every engagement opens with the audit. Two weeks, fixed scope, and the findings are yours whether or not you continue.
- Cadence
- Weekly written progress against a shared board. One monthly review call. No status meetings for their own sake.
- Deliverables
- Named artefacts with acceptance criteria — not hours. You can tell whether a thing was delivered.
- Reporting
- The same prompt set, re-run monthly against the original baseline, with the transcripts attached.
- Ownership
- Accounts, content, schema, automations and documentation are created in your name and stay there.
- Exit
- Thirty days. You keep the documentation and the system runs without us. We would rather be re-hired than retained by inertia.
Reasonable objections.
How can you influence what a model says? You don't control it.
We don't, and anyone claiming otherwise is selling something. What we control is the input: how consistently your entity is described across the web, whether your facts are machine-readable, and whether the sources models lean on carry a current version of you. Change the inputs and the output shifts — measurably, and not instantly.
Isn't this just SEO with a new label?
The foundation overlaps — crawlability, structure, authority. Above that it diverges. Ranking puts a page in a list. What we work on is whether a model can identify you at all, and whether your facts are clean enough to be lifted into a composed answer. A site can rank well and still never be named.
What if the audit says we don't need you?
Then it says that, and you keep the findings. It happens most often with businesses whose visibility is fine and whose real problem is response time or offer — in which case we would point at capability 04, or at nothing at all.
How long before anything changes?
Entity and schema repair often shows up in assistant descriptions within weeks. Citation footprint and authority are slower — three to six months for a durable shift. We would rather under-promise that than lose the engagement in month four.
Do you work with businesses outside your time zone?
Yes. Most clients are US-based; we overlap US business hours and work asynchronously outside them. Schema, content, citations and automations are reviewed on a schedule rather than in real time, so the overlap matters less than it sounds.
Can we start with one capability and add later?
That is the normal path. The audit will usually recommend one starting point, and adding a second capability is a scope change rather than a new relationship.
Find out what's on record.
Send your domain and the buyer you want to reach. We run your prompt set through the major assistants, check your entity and structured data, and send back the readout with the raw transcripts.
- Written by a person, not generated
- Back within two business days
- Yours to keep, with no obligation