The budget question is the first one almost every prospective PPC client asks, and the honest answer is that there is no useful number that applies across businesses, markets, and objectives. A budget that is grossly inadequate for a law firm trying to generate conveyancing enquiries in London might be entirely sufficient for a specialist B2B software company targeting a narrow buyer profile in a low-competition market.
What exists is a logical framework for working backwards from your business objectives to a budget that makes economic sense. Here is how that reasoning works.
Start with your target cost per acquisition
The foundation of any rational PPC budget discussion is your target cost per acquisition (CPA) — the maximum you are willing to pay for a conversion, given the value that conversion generates for your business. For a business selling a £40,000 consulting contract with a 60% gross margin, a CPA of £800 is very different commercially from the same £800 CPA for a business selling a £200 software subscription.
If you don’t yet have a clear view of customer lifetime value and gross margin by channel, establishing those numbers is the prerequisite for any rational budget discussion. An arbitrary budget disconnected from conversion economics will produce an arbitrary result.
Understand the market’s cost per click
Google Keyword Planner provides estimated CPC ranges for keywords in your target market. These estimates are approximations — actual CPCs depend heavily on Quality Score, competition, and bid strategy — but they give you a starting point for modelling. If keywords in your sector have CPCs of £5–15 and your target CPA is £150, you need your campaign to convert at 1–3% or better to hit that target. If conversion rates in your category typically run lower than that, you either need to improve your landing page or revise your CPA target.
Model your required monthly volume
Working backwards: if you need 10 qualified leads per month from PPC, and your target CPA is £200, you need a monthly budget of approximately £2,000 for the leads alone — plus management cost. If your target lead volume is 50 per month at the same CPA, the budget is £10,000.
This modelling is more useful than benchmark percentages (like ‘spend 10% of revenue on marketing’) because it connects budget to specific outcomes rather than to a percentage that is untethered from the economics of your specific market and offer.
Account for the learning period
Google’s Smart Bidding algorithms require a data learning period to optimise effectively. The general guidance is that campaigns need approximately 50 conversions in a 30-day period for Smart Bidding to perform well. During the learning period — typically the first one to three months of a new campaign — performance will be less efficient than it will become as the algorithm accumulates data.
This means your initial budget should account for a learning phase where CPA may be higher than your long-run target. Building that reality into the business case for PPC — rather than expecting target CPA from day one — produces more accurate ROI projections and avoids the common mistake of pulling the plug on a campaign that was just starting to perform.
The minimum viable budget concept
There is a minimum viable budget below which PPC campaigns produce so little data that optimisation is effectively impossible. If you are targeting competitive B2B keywords at £8–15 CPC and your daily budget is £10, you may get one or two clicks per day — not enough to draw any conclusions about performance, test ad copy variations, or generate the conversion volume that Smart Bidding needs. In high-competition markets, an underfunded campaign is often worse than no campaign at all, because it generates enough spend to feel significant but not enough data to learn from.
ThynkrSystems models PPC budgets from first principles — starting with your conversion economics and working backwards to the budget that gives your campaigns a genuine chance of hitting target CPA. We don’t recommend budgets that aren’t justified by the numbers.
