You Are Not Buying Visibility. You Are Renting It.
· AI Visibility · By Chris Latham, Founder of Optimus Consulting
PPC holds a position for exactly as long as you keep paying. Owned and AI-citable visibility compounds. Here is what that means for brokers, accident management, leasing and fleet, including the trade-off most people selling the idea leave out.
A piece in Legal Futures last week made a good argument about paid search and owned visibility. It was written for law firms. The same structure sits under brokers, accident management, leasing and fleet, so here is the wider version, including the part that usually gets left out.
Worth saying up front. Optimus sells AI visibility audits, so I have a commercial interest in where this argument lands. Read it with that in mind and check the reasoning rather than the conclusion.
The budget line that only ever goes one way
Most owner-managed service businesses have one marketing line that behaves differently from the rest. It goes up every year, enquiry volume stays roughly flat, and the explanation is always the same. The market got more expensive.
That explanation puts the cause outside the business, which makes it easy to accept. It is also incomplete.
John Anderson, founder of Somuna, made the point in Legal Futures on 31 July 2026. Every pound spent defending a paid position, he argues, is a pound that builds nothing which outlasts the campaign. Somuna is a paid Legal Futures Associate and sells the alternative to pay per click, so the article is content marketing. The argument still holds up, which is why it is worth taking apart properly rather than dismissing on provenance.
Why the price is rising, and who else is in the auction
Anderson puts cost per click on injury-solicitor keywords at £15 to £50. That range comes from Somuna's own UK Legal Services AI Impact Report 2026, published by a firm that sells the alternative to paid search, so treat it as a vendor estimate rather than neutral market data. Other published UK estimates for legal keywords sit across a much wider spread, and that variation is itself the useful signal. No headline range tells you what your own account is doing.
The mechanism behind the rise travels better than the number. In personal injury, the bidders pushing the price up are not only other law firms. Claims management companies and aggregators bid on the same terms with a different economic model. They do not need to win the case. They need to win the click.
That same shape appears across the sectors Optimus works in.
- Insurance brokers share their highest-intent keywords with comparison sites and direct writers whose acquisition maths is set at a completely different scale.
- Accident management companies bid against claims aggregators and lead resellers who monetise the enquiry itself rather than the case behind it.
- Leasing and fleet brokers compete with marketplace platforms funded to buy market share rather than margin.
Add consolidation on top. When private equity money lands in a sector and starts acquiring, the resulting group can outbid almost everyone in the auction and treat that as a strategy rather than a cost problem.
What your spend is not buying you
Paid search buys a position for exactly as long as the card keeps working. Stop, and the visibility stops with it. Nothing carries forward into next month.
Anderson also has a view on account management. He describes Google's Optimisation Score as a measure of how many of Google's recommendations you have accepted rather than a measure of how well the account is run. That is his characterisation and Google would dispute it, since Google presents the score as an estimate of how well an account is set up to perform.
The practical point survives the disagreement. A score calculated by the platform that earns your spend is one input, not a verdict. Two businesses bidding on identical keywords can end up with very different costs per lead depending on whether someone treats those recommendations as instructions or as suggestions worth arguing with.
The two curves: rented visibility versus owned visibility
This is really a finance question rather than a marketing one.
Rented visibility is an operating cost that resets to zero every month. Owned visibility behaves more like an asset. An article that earns a citation keeps earning it without a daily budget behind it.
So the two cost curves move in opposite directions. Each year a competitor builds owned, citable visibility, their cost of winning the next client through search drifts down. Each year a business relies purely on paid, its cost stays pinned to whatever the market's rising floor happens to be. Delay is not neutral, because the gap is cumulative.
Where AI search changes the maths again
AI assistants answer questions rather than return a list of links. When ChatGPT, Perplexity or a Google AI Overview names three providers, it is drawing on sources it can read, parse and attribute. That is a different qualifying process from an ad auction, and most businesses have never checked how they perform in it. We covered the mechanics in why your business is invisible to AI search .
One caveat, because the optimistic version of this argument overstates it. Owned visibility in AI surfaces is not permanently free of paid competition. Hiscox is already paying to appear inside ChatGPT, the same coin's other side. A paid layer is arriving in AI answers just as it arrived in search results twenty years ago.
The honest framing is not that paid dies and owned wins. It is that a business with owned visibility has something to fall back on when the paid layer arrives and gets expensive.
The honest trade-off nobody selling this mentions
Owned visibility is slow. It does not produce next week's enquiries, and it will not produce next month's either.
"Stop renting and start building" is easy advice to give when you are not the one meeting payroll. A business with a cash-flow problem cannot switch off the channel that produces this quarter's enquiries in exchange for something that might work by spring. That is not caution, it is arithmetic.
The sequencing that actually works is overlap rather than switch.
- Keep the paid spend that demonstrably produces enquiries you can trace.
- Cut the worst-performing slice of it, not the channel. Most accounts have campaigns nobody has questioned in a year.
- Fund the build from that trimmed slice, or from margin, rather than from the budget that keeps the lights on.
- Give it two to three quarters before judging it, and measure it against the enquiry record rather than a ranking report.
Anyone promising faster than that is selling something. So am I, which is why the number is a range rather than a promise.
The one question to put to whoever runs your marketing
Anderson closes on a good question. What percentage of the enquiry pipeline comes from paid versus organic and referral, and is that ratio moving in the right direction.
Most businesses genuinely cannot answer it, usually because paid and organic get reported as separate line items by different parties and nobody puts them in the same table.
Finding out takes an afternoon.
- Pull the last 90 days of genuine enquiries from your CRM or inbox, not from the ad platform. Ad platforms report the conversions they can claim.
- Tag each one with its first known source: paid, organic search, AI assistant, referral, direct, or unknown.
- Count the unknowns honestly and leave them as unknowns.
- Repeat the exercise for the same 90 days last year.
Two numbers come out. The current split, and the direction of travel. If the unknown pile is the largest, that is the finding, and it is a measurement problem to fix before any budget decision gets made. Our AI ROI calculator is a reasonable place to sanity-check the numbers once you have them.
What "owned" actually looks like in practice
The ingredients are not exotic. Named authorship with real credentials behind it. Structured data that states plainly what the business does and who works there. Content that answers the questions clients actually ask.
Two additions from what we see in audits.
The questions should come from your own phone log and inbox rather than a keyword tool. What people put to an AI assistant looks far more like the questions they ask a receptionist than the two-word phrases that fill keyword reports.
And the facts about the business need to match everywhere they appear. Site, Companies House, LinkedIn, sector directories. AI systems cross-check, and inconsistency reads as uncertainty. The overlap between this and conventional SEO came up in Google just said AEO is just SEO.
None of this replaces paid search. It gives a business something that is still there when the auction gets expensive. If you want to know where you currently stand before deciding anything, book a discovery call and we will look at it together.
Frequently Asked Questions
Is PPC still worth it for a service business in 2026?
For many, yes. Where buying intent is immediate and the sales cycle is short, paid search remains one of the fastest ways to generate enquiries, and switching it off abruptly carries its own risk. The problem is not using it. The problem is treating it as the whole strategy, so that the cost of winning a client is permanently tied to whatever the auction charges this year.
What is the difference between paid visibility and AI visibility?
Paid visibility is a position bought at auction which disappears when the budget stops. AI visibility is whether assistants such as ChatGPT, Perplexity and Google AI Overviews can find, understand and cite your business when someone asks a relevant question. It is earned through content, structured data and consistent facts about the business. Paid options are starting to appear inside AI answers too, so the two are not permanently separate.
How do I find out what share of my enquiries comes from paid?
Start from your own enquiry record rather than the ad platform. Take 90 days of genuine enquiries from your CRM or inbox, tag each with its first known source, and count the unknowns honestly. Then repeat for the same period last year. You get the current split and the direction it is moving. If most enquiries are untagged, fix that before making a budget decision.
How long does owned visibility take to build?
Longer than paid and shorter than people fear. Structured data and entity consistency can be sorted in weeks and often move things sooner than expected. Content that earns citations tends to need two to three quarters before the pattern is clear. It varies by sector and by how much authority the site already has, so treat any specific promise with caution.