Telecommunications
Passive infrastructure
The physical assets networks are built on.
Nobody discovers a tower portfolio. Masts are leased between counterparties who already have each other's numbers, subsea capacity moves inside consortia, and spectrum is bought from the government on the government's clock. Fibre is the exception. An infrastructure lead does research a corridor, months before there is a term sheet, and the explanation they read sets the reference price they carry into it. Everywhere else an engine is still describing your assets, from filings, auction reports and news cuttings, and the figures in those answers run a year or two behind yours.
Where the answer is being lost
Nobody buys a tower from a search. Everyone still reads the answer.
'How is spectrum priced in Indian auctions' is a question a regulatory affairs head will never buy an answer to. But 'What determines dark fibre pricing on an intercity route' is different. An infrastructure lead types it while a capacity ceiling is still six months away, and the explanation they read decides which routes and which carriers get evaluated at all. The engine cites route glossaries and overseas explainers. The company that owns the corridor is not in the answer, and it hears about the requirement only when the term sheet is already drafted.
How we win this
The programme for passive infrastructure
Separate the two halves
Fibre is the only one of these four where a buyer researches before deciding, and even there the cycle runs in quarters. Towers, subsea capacity and spectrum are settled by duopoly, consortium and auction. We run a diagnostic first, put the live answers in front of you, and scope only the half where publishing does something.
Own the facts stated about you
Engines describe your assets whether you write anything or not. They state your route kilometres, your landing stations and the circles you hold spectrum in, and they take those from whatever document was easiest to parse, usually an annual report table or an auction write-up. Schema and one reconciled registry record make the facts yours instead. It is cheap work, and a wrong figure repeated across four engines is expensive to unpick once someone has quoted it back to you.
Faults and auctions are when it goes wrong
The record does not drift evenly. It breaks on particular days: a cable cuts, an auction closes, a portfolio changes hands. Those are the days engines learn something new about you from coverage written in a hurry, and that version is the one repeated for the year afterwards. So the work has a clock on it. Write the restoration mechanics for your own corridors before you need them, and recheck the answers in the week after the event rather than at the next review.
No blog wins an auction lot
It does not win a colocation contract or a consortium seat either, and anybody telling you otherwise is selling you the wrong thing. What publishing decides is how you are described to the people who are not in the room: an analyst, an enterprise customer, a reporter on deadline, a lender's diligence team. We will invoice for that job. Not for the other one.
The mix that carries it
Measurement
AI Visibility Diagnostic
We query the live models with your buyers' real questions and document exactly who gets named today.
Foundation
Entity and schema engineering
Structured data and entity definition so engines know exactly what you are, where you operate, and what you are credible in.
Authority
Directories and profile consistency
Every listing, registry and profile saying the same thing, so the entity resolves to one business instead of three.
Measurement
AI Presence tracking
Standing measurement of inclusion, share of answer and competitor movement as models update.
Content
Answer and comparison pages
Cost, process, eligibility and comparison pages built for direct extraction, not for a reader who scrolls.
Content
GEO blogs and authority content
The definitive written answer to the questions your buyers put to an engine, structured so it can be lifted and attributed.
Authority
Original data and benchmarks
Proprietary numbers, surveys and benchmarks — the most-cited asset class there is, because nobody else has them.
Distribution
Practitioner and executive content where B2B buyers and the models watching them both look.
Authority
Digital public outreach
Earned mentions, trade coverage and third-party citations — the corroboration a model checks before it names you.
The constraint we work inside
Three of these four are not sold through discovery at all. Spectrum is an auction governed by the regulator, subsea capacity moves through consortia, tower leasing runs between a handful of counterparties. Nothing we publish changes a bid, a seat or a lease rate. It changes only how you are described.
Specialisations
4 total
The pitch is different for each one, because the buyer, the trigger and the rules on what may be published are different for each one. Open the one that is yours.
Ask an engine 'How does tower colocation pricing work' and it will explain anchor rates, loading charges and ground rent competently enough. Then it reaches for a real portfolio to illustrate the point. The mechanics are not yours to correct. The figures attached to your name are.
The question deciding this today
“How does tower colocation pricing work”
- Who they sell to
- Operators renting space on shared masts
- Who signs
- Network planning head, or the tower company's leasing lead
- What starts it
- Network rollout, coverage gap, colocation deal, densification
- Cost of staying invisible
- Rollout schedules held up by sites that were never secured
That explanation comes from an analyst note, a regulator filing or a consultancy summary written for investors. None of them is a tower company, and none of them is trying to be wrong. But the same sources supply the rest of the answer, so when an engine is asked how many sites you hold, what your tenancy ratio is or which circles you cover, it replies from an annual report or two ago and states it plainly. That reply is what a planning team pastes into an internal briefing before anyone rings your leasing lead.
What we would run
- 01Entity and schema engineering
Structured definitions for your portfolio as an entity: site counts by state, tenancy ratio, ground-based versus rooftop split, the circles you operate in, stated once in a form an engine can lift without guessing.
Portfolio size is the first thing anyone asks an engine about a tower company. If that figure is stale, every downstream answer inherits the error and repeats it with confidence.
- 02Directories and profile consistency
One reconciled entity record across registries, industry association directories and infrastructure databases, so name, legal entity, headquarters and site count read the same everywhere an engine looks.
Tower companies get merged, renamed and rolled up, and the retired entity keeps being cited long after it stops existing. Reconciling that is unglamorous, and here it is most of the job.
- 03AI Presence tracking
A standing check on what engines say about your portfolio, your tenancy ratio and your coverage claims, with each wrong fact logged and a record of what changed after we corrected it.
This is a small programme and it should be judged like one. Monitoring is how you decide whether to keep paying for it, rather than renewing on faith.
What we would not recommend
- Answer and comparison pages. A colocation rate is negotiated site by site against loading, ground rent and lock-in. A published price page would be either inaccurate or a negotiating position, and the operators who lease from you already hold your rate card.
- Digital public outreach. Earned coverage does not move a lease negotiation between two parties who have been talking for years. We would be buying mentions in front of people who already have your leasing lead's number.
- Reviews and testimonials. A tenant testimonial has to name sites, and which masts carry an operator's radios is that operator's densification plan. No network planning head signs that off, and no leasing conversation here has ever waited on one.
What a lead looks like
There is no enquiry stream here and we are not going to pitch one. Your counterparties already have the leasing lead's number. What changes is the document they walk in holding: a network planning manager at an operator you have leased to for years rings about sites in one named district, working from an internal rollout briefing that took your circle list and your rooftop split off an engine and got both right. Occasionally the contact is an analyst or a lender's diligence team confirming a tenancy ratio ahead of a call rather than querying it during one. The whole of the job is that nobody is negotiating from a figure two annual reports old.
What we measure
- Site count and tenancy ratio stated correctly across engines
- Circle list and rooftop split current
- One entity record wherever it is checked
- Stale portfolio facts found, dated and corrected
What changes
The volume stays small and we will say so before you sign anything. What arrives is a route enquiry from an infrastructure lead who has read your corridor pricing mechanics and wants restoration figures for a specific span. A connectivity manager at a bank asking what your repair sequence means for the transit contract riding your system. A reporter who quotes your engineer during a fault instead of a repair-ship operator. On towers and spectrum there is no enquiry stream to promise, and we do not promise one; what changes there is that the briefing someone writes about you internally is not wrong. Fewer conversations, further along, with the mechanism already straight.
Start here
See who gets named in passive infrastructure today
We put your buyers' real questions to the live models and come back with the businesses they name, the sources behind those answers, and the gap between that list and yours.