A paid search budget allocation guide should not begin with a percentage split between Google Search, Shopping and Performance Max. It should begin with the question that actually determines whether growth is profitable: how much can you afford to pay for a new customer or qualified lead, after margin, fulfilment, sales costs and repeat purchase behaviour are accounted for?

For established eCommerce and lead generation teams, the challenge is rarely finding another place to spend. Google Ads will happily take more budget. The real work is deciding where the next pound produces incremental revenue, where it protects demand you have already created, and where it is merely claiming credit for conversions that would have happened anyway.

Start with business economics, not platform recommendations

Set a clear acquisition target before opening the campaign dashboard. For eCommerce, this may be a blended customer acquisition cost, a contribution-margin ROAS target, or a new-customer ROAS target. For lead generation, it is usually a cost per qualified lead, cost per opportunity, or cost per sale rather than a headline cost per lead.

A £20 lead is not automatically better than a £40 lead if the cheaper source produces low-intent enquiries that the sales team cannot convert. Equally, a campaign that appears to hit a 4x ROAS target may be unprofitable once returns, discounting, agency fees and fulfilment are included. The budget has to be governed by the metric that reflects commercial reality.

This is also where targets need context. A brand with strong repeat purchase rates may reasonably accept a lower first-order return while it acquires high-value customers. A business with tight cash flow or low margins needs a faster payback period. There is no universally correct ROAS threshold, only a threshold aligned to the business model.

Build a paid search budget allocation around intent

Search spend has different jobs. Branded terms capture existing demand. Non-brand search acquires customers who are actively comparing solutions. Shopping campaigns put products in front of high-intent shoppers. Remarketing helps bring back visitors who did not convert first time. Treating all of these as one pool obscures what is driving growth.

Start by separating campaigns according to their role, then give each role a budget expectation. Branded search should have sufficient budget to maintain strong impression share for commercially valuable terms. It is usually efficient, but it is not a reliable measure of new demand generation on its own. If brand spend is growing because other channels are creating awareness, taking credit for all branded revenue can create a false sense of paid search performance.

Non-brand search and Shopping commonly deserve the largest share of acquisition budget because they reach people with a defined need. However, their capacity is constrained by search volume, competitiveness, product price and feed quality. Increasing spend without improving ad relevance, product coverage, landing pages or offers can quickly push campaigns into less profitable auctions.

Remarketing should be funded deliberately but capped sensibly. It often produces attractive reported ROAS because the audience already knows the brand. That makes it useful for conversion efficiency, not a licence to over-invest in a small, frequently reached audience.

Use a three-part budget model

A practical way to avoid overreacting to daily platform data is to divide the paid search budget into three buckets: protection, proven scale and controlled testing.

Protection covers campaigns that defend revenue already at risk, such as brand terms, top-selling products and high-performing lead generation keywords. These campaigns should not be allowed to run out of budget in peak trading periods simply because another campaign has a lower reported cost per conversion.

Proven scale receives the largest share. This is where performance is repeatable across a meaningful period, tracking is trustworthy, and the campaign has room to capture more demand without a sharp decline in efficiency. Look beyond a single seven-day window. Review conversion quality, auction trends, search terms, stock availability and performance by device, geography and audience where relevant.

Controlled testing is the budget that creates future growth. Reserve a defined amount for new keyword themes, category expansion, feed improvements, landing page tests, seasonal offers or alternative bidding approaches. The percentage depends on account maturity and risk tolerance, but teams that allocate nothing to testing often plateau. They keep concentrating money in the campaigns that worked yesterday while competitors find the next profitable pocket of demand.

For a mature account, 70 to 80 per cent on proven scale, 10 to 20 per cent on protection and 10 per cent on testing can be a sensible starting point. It is not a fixed rule. A highly branded business may need more protection; a business entering a new market may invest more heavily in tests for a defined period.

Do not let automated bidding set the strategy

Automated bidding is useful when conversion tracking is accurate, volume is sufficient and targets reflect the right outcome. It is not a substitute for budget allocation. Smart Bidding will optimise towards the data you feed it, including poor-quality leads, duplicated purchase events or revenue that ignores cancelled orders.

Before increasing budgets, audit the signals that steer bidding. Check that primary conversions represent genuine commercial value, enhanced conversions are working where appropriate, and consent settings have not created unexplained gaps in measurement. For eCommerce, ensure transaction values, refunds and new versus returning customer behaviour are visible somewhere in your reporting. For lead generation, connect offline outcomes back to the source wherever possible.

Budget decisions made on incomplete attribution are still decisions. They are simply less reliable ones. This is why tracking infrastructure is not a technical side project. It determines whether the media team can scale with confidence or has to guess.

Scale in increments that preserve learning

Large budget jumps can destabilise a campaign, particularly where conversion volume is limited. Increase spend in measured increments, then allow enough time to assess performance against the sales cycle and conversion lag. A lead generation campaign with a 30-day sales cycle cannot be judged properly after 48 hours, no matter how convincing the dashboard looks.

Watch marginal performance rather than blended averages alone. If a Shopping campaign has delivered a 5x ROAS at £500 per day, the key question is what happened when it moved from £400 to £500, and what is likely to happen at £600. The next pound is more valuable than the historical average when you are deciding whether to scale.

Impression share, lost impression share due to budget, search volume and auction insight data can indicate headroom, but they do not guarantee profitability. A campaign may have plenty of available impressions in low-quality queries. Review search terms, product-level performance and conversion quality before treating lost impression share as a reason to spend more.

Reallocate budget with a set decision rhythm

Constantly moving budget between campaigns creates noise and makes learning harder. A better approach is to set a rhythm: monitor daily for tracking failures, overspend, stock issues and major anomalies; review weekly for pacing and efficiency; and make more material allocation decisions monthly or around a clear promotional event.

When a campaign underperforms, diagnose the cause before cutting spend. Has demand softened? Has a competitor changed its offer? Are products out of stock? Has the landing page slowed down? Did the conversion action change? Budget is often blamed for problems that begin elsewhere in the acquisition system.

Likewise, when performance improves, ask whether the improvement is incremental. A promotion, email campaign or organic trend may have lifted branded searches and remarketing conversions. That does not mean paid search caused all the additional demand. A collaborative review across media, merchandising, creative and conversion rate optimisation gives the account a much clearer picture.

Treat seasonality and stock as budget inputs

Your ideal allocation in January may be wrong in November. Plan budget around promotional calendars, product launches, stock depth, average order value and known demand peaks. There is little value in aggressively scaling ads for a hero product that is about to sell out, unless the business is prepared to redirect demand to alternatives.

For eCommerce teams, product feeds deserve particular attention. Excluding low-margin products, separating bestsellers, fixing disapprovals and improving titles can make existing budget work harder than a broad spend increase. For lead generation businesses, the equivalent may be excluding poor-performing locations, refining qualification questions or routing high-intent leads faster.

The best allocation is not the one that makes a platform report look strongest. It is the one that gives your business a repeatable path to profitable growth, with enough protection for existing demand and enough room to discover what works next. Keep the numbers honest, make changes with intent, and let commercial outcomes – not vanity metrics – decide where the next pound goes.