Budget allocation, pacing and bidding are three different decisions

When a month goes off plan, the usual response is to change several settings at once: move budget between campaigns, raise or lower daily budgets and adjust bid targets. If results then change, nobody can say which move caused it, and the next decision is made on guesswork.

We find it more useful to separate three decisions. They are connected, but each answers a different question and needs different evidence.

• Allocation: how much of this month's money should each channel and campaign have, given the commercial plan?
• Pacing: within the month, is actual spend tracking that plan closely enough, in both directions?
• Bidding: within each campaign, what should the budget buy, and at what target?

Allocation should start from the business: the monthly total, the products or services that matter most this period, seasonality and what each channel is accountable for. A brand search campaign that protects demand you already have and a prospecting campaign that sets out to create new demand should not be judged on the same target, so they should not compete for budget on the same ratio.

Last month's split is a weak default because it encodes last month's constraints. A better starting point is evidence of headroom and ceilings. Is a campaign held back because its budget runs out, or because its bids, ads and quality do not win enough auctions? Only the first is a budget problem. Impression share data can indicate whether a campaign might reach more people with more budget or a higher bid, and that distinction decides which lever to pull. Is efficiency holding as spend rises, or falling away? That tells you whether the next pound is likely to earn more or less than the last one.

Write the plan down: the total, the split by channel, how spend should be shaped across the month, the tolerance you consider on track, and the target each allocation is accountable to on a stated basis. A plan that lives in someone's head cannot be checked, by you or by anyone else.

Pacing is the in-month question. It needs a reference line, which is the plan shaped by the spend curve you agreed, and a view of how actual spend compares so far and where it is heading by the end of the month.

It helps to know how the platforms treat daily budgets. Google Ads describes an average daily budget: for most campaigns, spend on a single day can reach twice that amount, while charges in a month are capped at 30.4 times the average daily budget. Meta describes a daily budget as a target across the week: spend on one day can reach 175% of it, and spend across the week will not exceed seven times the daily budget. Both behaviours are sensible, but together they mean the sum of your daily budgets is an envelope around the plan rather than the plan itself. If budgets are raised on several campaigns during a restructure, that envelope can grow quietly while the plan stays the same.

Pacing has a second failure that is easier to miss. A campaign that under-spends while beating its efficiency target is not being prudent; it may be leaving profitable demand unmet. Separate cannot spend, where there is not enough eligible demand, inventory or approved creative, from will not spend, where a budget is set too low. They need different fixes.

Bidding is a campaign-level calibration. The target should come from the business's own economics rather than from a platform recommendation: a revenue target that reflects margin for ecommerce, or a cost per qualified lead that reflects what a lead is worth for lead generation. Where margin data exists, the break-even return is the floor below which a target should not knowingly sit.

Move targets in steps rather than leaps, and give each change long enough to be judged before making the next one. Automated bidding needs time and data to respond, and a large jump makes it hard to tell whether the strategy or the market moved.

The rule that ties the three together is to change one lever at a time on any given campaign. If a campaign is limited by budget, give it budget before touching its target. If it is limited by competitiveness, look at bids, ads and landing pages rather than budget. Record what you changed, why, and what you expect to happen, then check it at an agreed point. That record is what makes the next month's plan better than this one.

A hypothetical example

The figures in this example are hypothetical. A retailer plans £30,000 for the month across search and social. Halfway through the month, total spend is close to plan, which looks fine. Split by campaign, a prospecting campaign is running out of budget each afternoon while returning well above its target, and a brand campaign is spending freely on searches that would largely have arrived anyway.

The allocation answer is to move budget towards the constrained prospecting campaign. The pacing answer is to check that the account's combined daily budgets still sit sensibly within what remains of the plan after that move. The bidding answer, for now, is to leave the prospecting target alone, so that the effect of the budget change can be read cleanly next week.

• Is the monthly plan written down, with a stated basis for each target?
• For each campaign that looks limited, have you established whether budget or competitiveness is the constraint?
• Do the combined daily budgets, multiplied by the days remaining, sit sensibly against the budget remaining?
• Are campaigns that under-spend while beating their targets flagged, not just the ones that over-spend?
• Is each change a single lever, with a stated expectation and a review date?