Borderly
0 → 10K+
From zero to 10,000+ monthly visitors — 45% conversion rate from ChatGPT traffic
SEO & GEO agency
Be the answer,
not a result.
Buyers stopped scrolling through ten blue links. They ask, and something answers. We do the technical, content and authority work that decides whether your brand is the name in that answer — and whether you still rank when they go looking.
- Borderly
- 0 → 10K+Borderly
- Fluency
- 40KFluency
- United Travels
- $40K/moUnited Travels
Surfaces we optimise for
ChatGPTGoogle AI OverviewsGeminiPerplexityClaudeCopilotGoogle SearchAI ModeChatGPTGoogle AI OverviewsGeminiPerplexityClaudeCopilotGoogle SearchAI Mode
The shift
Search is no
longer just
a search box.
The channel did not disappear. The interface changed, and the work required to win it changed with it.
01
An answer replaces the list
A generated answer names one or two options. There is no page two to be on, no position four to climb to. You are in the answer or you are not in the consideration set at all.
02
A recommendation is not an ad
Being named by an assistant reads as advice, not as a placement someone paid for. That is a different kind of trust, and it arrives before your marketing gets a say in the framing.
03
The sources are already being set
Assistants ground answers in pages, profiles and roundups that exist today. Every month those references harden. Competitors establishing themselves in that source set now are laying groundwork that gets more expensive to displace.
The distinction
“Isn’t GEO just
SEO with a new name?”
Partly, and anyone claiming otherwise is selling something. The technical foundations are shared. What differs is the shape of the win — and that changes what you build.
A results page
10 slots1
2
3
4
5
6
Rank fourth and you still get seen. There is room to be a contender.
An answer
1 slot“For what you’ve described I’d go with one name. It covers the requirement you mentioned, and the setup is straightforward.”
Nobody else is shown
No position two to climb to
You are the answer or you are not in the conversation.
| Dimension | Classic SEO | GEO |
|---|---|---|
| What you win | A position in a list | A mention inside the answer |
| How many can win | Ten on page one | Usually one or two |
| What the system does | Ranks documents | Extracts and synthesises passages |
| What content must survive | Being scanned | Being quoted out of context |
| What settles ties | Authority and relevance | Corroboration across independent sources |
| How you measure | Rankings, sessions, conversions | Citation share, plus the same conversions |
This is why we do not sell them separately. Ranking well is increasingly a prerequisite for being cited — assistants that retrieve live results ground their answers in pages that already rank. Running the two as separate projects means paying twice for overlapping work.
What we do
GEO
Getting named inside the answers ChatGPT, Gemini, Claude and Perplexity generate — not ranked underneath them.
- Citation share
- Retrievability
- Entity work
What we do
Eight ways in.
One programme.
Most engagements start with two or three of these. They are listed separately because they are scoped separately, not because they run separately.
Where we work
Search behaves differently by sector.
What counts as a good query, a good page and a good result changes a lot between a catalogue of forty thousand products and a nine-month enterprise sale.
The through-line
Many options on the wheel. One at the marker.
How we work
Diagnose first.
Then build.
- 01
Baseline and diagnosis
What you rank for, what gets cited, what is technically holding you back, and which sources are feeding the answers your buyers see. Nothing gets recommended before this is on the table.
- 02
Strategy and sequencing
A prioritised plan with the reasoning attached: what we do first, what it should move, and what we are deliberately not doing yet. Sequencing is most of the value — order changes outcomes.
- 03
Build and ship
Technical fixes specced and reviewed through release, content briefed and produced, authority work run as campaigns. We stay involved through implementation, which is where most strategies quietly die.
- 04
Measure and compound
Rankings, organic performance and citation share reported against the baseline on a set cadence, with the next cycle prioritised from what actually moved rather than from the original plan.
Why sequencing matters
Organic returns are back-loaded. The first two months look like nothing is happening, because the work that pays later has to happen first.
Illustrative. Shows the shape of a compounding programme, not any client’s results.
Measurement
You should be able
to check our work.
Citation share, not vibes
A fixed prompt set built from your real buying questions, re-run on a cadence, so "are we showing up in AI" becomes a number that moves rather than an anecdote someone screenshotted.
Reconciled to pipeline
Rankings and citations are instruments. We tie them to trials, opportunities and revenue — and where attribution genuinely cannot be traced, we say so rather than modelling something reassuring.
Against a baseline you keep
Everything is measured against the state we found on day one. It stays visible for the life of the engagement, which also makes it obvious when we have not earned the retainer.
Reporting cycle
vs. baselineCitation shareacross 5 assistants
Organic rankingstracked query set
Technical healthcrawl + render + index
Authorityearned placements
Pipelinereconciled to revenue
Illustrative. Shows what a cycle covers, not any client’s numbers.
The work
Case studies
Fluency
40K
Peak monthly traffic driven entirely through LLM-optimized content
United Travels
$40K/mo
10 high-ticket bookings ($4,000 each) per month from day one
Fit
We are not right
for everyone.
Good fit
- You sell something with a considered evaluation, where being recommended matters
- You have a site with enough substance to work with, or the appetite to build one
- Someone internally can make decisions and unblock developers
- You can commit to two or three quarters — organic does not pay out faster than that
Probably not
- You need traffic this month — that is a paid problem, not an organic one
- You want volume publishing with no strategic layer
- Rankings are the goal in themselves, disconnected from revenue
- The plan is already decided and you want someone to execute it unquestioned
How we engage
Stay because
it works.
Retention should be a verdict on the work, not a clause in an agreement.
- No long-term contract
- Month to month. If the work stops being worth it, you stop paying for it. Lock-ins mostly protect agencies from the consequences of underperforming.
- You own everything
- Audits, briefs, maps, data, documentation — all of it is yours and leaves with you. Nothing is held in a system you lose access to when the engagement ends.
- You talk to the people doing the work
- No layer of account management relaying questions to whoever is actually on your site.
Plans
Three ways to
work together.
They differ on one thing: who implements. The diagnosis and the strategy are the same in all three.
Foundation
Strategy package
For teams who need to know what is wrong and what to do about it, and have the capacity to act on it themselves.
One-off engagement
Consultancy
No implementation
For teams with in-house developers and writers who can execute, but want the strategy, briefs and measurement owned by specialists.
Ongoing retainer
Full Hands-On
We do the work
For teams who want the whole programme run — strategy, technical implementation, content production and authority work.
Ongoing retainer
Before you commit
Take the tools. Read the writing. Decide later.
Free tools
Use these without talking to us.
Sixteen of them, no signup, no email wall. If they are all you ever take from us, that is a fine outcome.
All sixteen tools
Writing
How this actually works.
FAQ
Straight
answers.
01What is GEO, and is it different from SEO?
GEO is Generative Engine Optimization — the work of getting a brand named inside answers from ChatGPT, Gemini, Claude, Perplexity and AI Overviews. It overlaps heavily with SEO, because both reward authority, clarity and technical health. What differs is the target: assistants extract and synthesise rather than rank, so content has to survive being quoted out of context, and independent corroboration matters more than it does for a blue link.
02Can you guarantee we get cited by ChatGPT?
No. Nobody controls model output, and any agency guaranteeing a citation is either misunderstanding the mechanism or misrepresenting it. What we control is whether your brand is retrievable, correctly described and corroborated by the sources these systems draw on. We report citation share against a baseline so you can see whether that work is moving.
03How long before we see results?
It depends almost entirely on where you start. A site with existing authority and a technical problem holding it back can move quickly. A new domain with no third-party presence is a multi-quarter build. We set expectations against your baseline in the first few weeks rather than quoting a generic timeline up front — and if the honest answer is longer than you can wait, we will say so.
04Do we have to sign a long contract?
No. Engagements are month to month. We ask for two or three quarters of intent because organic returns are back-loaded and stopping at month two wastes the work, but that is a recommendation and not a clause.
05Do you work with our in-house team or replace it?
Either. We frequently run the strategy, technical and AI-visibility work while an in-house team produces content, because your team has the product knowledge and we have the search reasoning. We can also run the whole programme where there is no internal resource.
06Do you still do paid ads?
As a supporting channel for existing clients, yes — covering the gaps while organic builds. Organic is the focus and where the compounding value is. If you want a paid-first agency, we are not it, and we will say so on the call rather than after you have signed.