Business owners do not ask us what generative engine optimisation is. They ask a harder question: why this, why now, and what do I get for it. This piece answers those three in order, and it ends with the reasons you might reasonably decide to wait.
What has already changed
The behaviour that matters is small and specific. A person with a problem opens an assistant instead of a search engine, describes the situation in their own words, and reads a written answer. If that answer names a business, that business is on a shortlist of one or two before anybody has compared websites, read reviews or made a call.
This is not most of your enquiries. For almost every service business the search results page is still the larger channel by a wide margin, and anyone telling you to abandon it is wrong about the numbers. What makes this channel worth attention is not its size today. It is three properties it has that search does not.
It names one or two, not ten. There is no page two to be on. The gap between being named and being third is the whole result.
It reaches people earlier. Somebody asking an assistant is usually still working out what their problem is and who handles it, which is before they would have typed a service and a city into a search box.
It carries more weight than a listing. A recommendation inside a sentence of prose reads as advice. A position in a list reads as advertising, and people have thirty years of practice discounting that.
Why now rather than next year
One. The ground is still uncontested in most markets
Open an assistant and ask who to call for your service in your city. In most markets the answer is a directory, a trade or professional association, or general advice with no business named at all. That is not a competitive field. It is an empty one, and it is empty because the work is unglamorous and nobody has done it. It will not stay empty, in the same way that nobody could buy their way cheaply to the front of a search results page after about 2012.
Two. The slow part is calendar, not money
The highest-impact piece of this work is reconciling your public records so every independent source describes the same organisation. Licensing boards, professional registers and major directories update on their own schedule, not yours. Six to ten weeks is normal. Starting in January rather than October does not cost you three months of fees. It costs three months of elapsed time that no budget can buy back.
The part of this that takes longest is the part money cannot accelerate. That is the whole argument for starting early.
Three. Early citations compound
Assistants reuse sources that worked. A page cited once for a question tends to be reached for again on similar questions, and corroboration accumulates: once three independent records agree about your business, the fourth is easier to get and the whole set is harder to dislodge. Absence compounds in the same direction, which is the uncomfortable half of the same sentence.
Four. Your competitors’ agencies are already renaming the retainer
Every search agency now has a page about this. Some are doing real work. Many have relabelled the content line item and changed nothing underneath. The practical consequence is that within a year or two every business in your market will be paying for something called GEO, and the ones that took a real baseline early will be the only ones able to prove what it did.
What it actually returns
Three returns you can measure, and one you cannot. We will be precise about which is which, because the gap between them is where most of this industry’s overselling lives.
Measurable: your citation share
A fixed panel of questions, run on a schedule, produces one number: the share of answers that named your business. It moves, it is comparable month to month, and you can hand it to a committee without explaining a methodology.
Measurable: competitor share on the same panel
The same run tells you how often each competitor was named. This is the figure owners find most useful, because it converts an abstract channel into a position relative to the three competitors you already think about.
Measurable: whether a page you control was the cited source
Being named because a directory listed you is worth less than being named because your own page was quoted. The second is durable; the first is borrowed. Tracking the split tells you whether the content work is paying or whether you are riding someone else’s listing.
Not measurable: signed clients
Nobody can currently trace an assistant mention to a signed client. There is no referrer, no click, and often no visit at all. Somebody hears your name, searches for it directly, and phones. Your analytics will book that as direct or branded traffic and give the credit to the wrong column.
We will not pretend otherwise, and you should treat a confident conversion figure from any agency in this field as a reason to ask how it was derived. What can be said honestly is narrower: a business that appears in none of these answers is absent from a channel a measurable share of its future clients now use before picking up the phone.
What it costs, honestly
Our retainer is $2,500 or $6,500 a month on a ninety-day initial term, depending on how many records describe you and how contested your market is. Where the hours go matters more than the figure. The largest block by some distance is record reconciliation, which is forty small chases rather than one project and produces nothing you can look at. Next is rewriting pages. Then technical implementation, then measurement, then the time it takes to explain all of it to you, which we bill and most agencies hide.
| Month | What happens | What you see |
|---|---|---|
| One | Baseline, audit, corrections filed | A dated baseline. Nothing moving. |
| Two | Rewrites, structured data, corroboration | Pages changing. Numbers flat. |
| Three | Records propagate, re-measurement | The first honest comparison. |
Because registries take six to ten weeks to propagate a correction, the end of month three is the earliest point at which a comparison means anything. Any agency promising movement in thirty days is either not doing the entity work or not telling you when it will land.
How to judge it at ninety days
Ask for three numbers against the week-one baseline, on the same panel, run the same way.
- Named share. How many of the forty answers named your business, then and now.
- Competitor share. The same figure for each benchmarked competitor.
- Owned-source share. How often the cited source was a page you control.
If the first two moved and the third did not, the gain is borrowed from directories and will not hold. If the third moved and the first did not, the content is landing but the records have not propagated yet, which is a timing problem rather than a strategy problem. Both are honest answers, and each tells you what the next quarter should contain.
The honest reasons to wait
Your website cannot be edited quickly. If every page change queues behind a six-week vendor slot, the content half of this will not happen and you will be paying for half an engagement. Fix that bottleneck first.
You have no capacity to take the work. Being named raises enquiries at the research stage, which are earlier and less qualified than a referral. A business already turning work away does not need more enquiries this quarter.
Your profession has advertising rules you have not settled. A results claim that is compliant on your page can be restated by an engine in a form that is not. If whoever signs off your advertising has not looked at that, look at that first.
What to do this week
None of this requires an agency to begin. Write down the forty questions a real client would ask before calling you. Put them to each assistant in a fresh session. Record who was named, which sources were cited, and the date. That document is the baseline, it costs an afternoon, and whether you hire us or not, nobody can sell you movement afterwards without it.