Google Ads growth, measured by what the next pound buys.

The number most accounts never check before they scale

Growth starts with a question most reports never answer: what does the next £1,000 of spend actually buy? Not what the account has averaged so far, but what the last slice, right at the edge of the budget, costs to fill.

Take an account that reads £25 a lead across the month. It can still be buying its last leads at £45, because the cheap early clicks pull the average down while the marginal pound goes further down the demand curve — into weaker searches, wider match types, quieter hours of the day. That is an illustration of the shape, not a claim about any specific account; the real split only comes out of the account's own numbers. Read the average alone and this account looks ready to scale. Read the marginal cost and it is telling you the next £1,000 buys leads at nearly double the price of the ones already sitting in it.

We work this out before recommending an increase. It comes from the account's own auction and spend data, not a rule of thumb: the search terms report split by spend band, and what the last campaigns, match types and hours in the budget are actually converting at. If the marginal number still holds against your target cost per lead, growth is safe. If it does not, more budget just buys more of the expensive stuff.

None of this works on broken measurement. If a call goes uncounted or a form fires twice, the marginal number is built on the wrong inputs and scaling optimises towards noise rather than real leads. Tracking is verified first, as part of every audit — see conversion tracking.

Where growth actually comes from, in the order we look

Before budget goes up, three other places usually pay for a chunk of the increase themselves. We work through them in this order on every account.

1. Reclaim what is already being spent. The cheapest growth often sits inside the current budget already: search terms, placements and locations that were never going to convert, still quietly burning spend every week.
2. Geography next. Budget follows the areas already converting, reviewed weekly rather than set once and left. Only once that split is settled does expansion into an adjacent area get its own deliberate campaign, rather than a few extra postcodes bolted onto the one that exists.
3. New service lines, as their own campaigns. A genuinely new line of work earns its own campaign and its own landing page, matched to that specific search, kept separate from an account built to sell something else.
4. Raising the budget, last. Only once the first three are done and the marginal cost per lead (the number above, not the average) still holds against target.

Reclaiming spend is usually the fastest of the four. It needs no new budget and no waiting on fresh conversion data — the search terms and location reports already hold the answer. Geography is the one with a worked example, below.

Geography, worked through a real account

EngiClean's DPF cleaning network is the book's clearest worked example of geography as a growth lever: across the rebuild, budget followed the locations producing enquiries, reviewed weekly, rather than a bigger cheque leading the way.

Some markets run out of searches. We say so.

More budget cannot buy demand that does not exist. A local market for a trade only has so many people searching for it each month, and once an account is already capturing most of that volume, the next £1,000 does not buy more leads at the same price — it buys the same leads bid up against themselves, or clicks from further away that were never going to book.

Where that is the case, we say so, with the search-volume numbers behind it rather than a vague line about competition. The benchmarks calculator shows the realistic ceiling for a given trade and area before any spend goes up — worth checking ahead of a scaling conversation, not after it: see the benchmarks calculator →.

A ceiling is not a reason to stop growing. It is a reason to grow in a different direction: a new area, a new service line, a wider match on searches not yet being caught, rather than pouring more money into demand that has already been used up.

What changes as the budget scales from £1,000 to £7,500

Two mechanics govern that whole range: how evenly the money actually gets spent through the month, and how long the account structure stays simple.

Google does not spend a monthly budget in thirty equal instalments, and seasonal trades peak harder than their average suggests — a roofing account chasing demand after autumn gales is the clearest case, where a flat daily budget starves the surge and overspends the lull (see Google Ads for roofers). Growth nobody watches day to day looks fine on the monthly total and terrible on the Tuesday it actually happened, so pacing gets checked against the plan through the month rather than totted up at the end of it.

The account structure should change more slowly than most agencies claim. A single campaign, well fed, holds together considerably further up the budget than the usual advice to split it as soon as it grows, because every extra campaign is also a split in the conversion data that Smart Bidding learns from — two campaigns each getting half the enquiries teach the algorithm half as fast as one campaign getting all of them. Splitting comes later, for a reason rather than a milestone: a genuinely new service line, a new area already proven by the geography work above, or a budget large enough that one campaign is structurally capped rather than merely comfortable.

Raising the budget is the last step in every case above — after the waste is out, the geography is settled and the marginal cost per lead still holds. Get there in that order and the increase tends to look, in hindsight, like the obvious next move rather than a gamble. The fee framework for accounts across that whole range is on the pricing page.