What a restore trigger returns, and what has changed since the cut
The restore trigger is the last decision of a well-made cut. The scenario planning framework gives it a shape: a signpost the business already measures, a level at which it acts, and a response agreed in advance, such as a named sum returning to a named line. The marketing budget cut checklist asks for the trigger and for a record of the starting settings before the first change is made, because, without a restore trigger, a cut tends to become the new budget. By the time the signpost reaches its level, whether money comes back has already been settled. What remains is how it goes back in, and that is a different job from the cut.
Three things make it different. The accounts are no longer as they were left: some campaigns were paused, some targets tightened, some lists left without a refresh and some tests stopped halfway. The money returns into a market that has moved, with a different season, different competitors and different prices for the same attention. And the old figures are only a starting guess. The marginal returns and forecasting guide treats every spend forecast as a hypothesis to be tested against what happens, and it explains why the final slice of a budget earns less than any before it. The money that comes back is that final slice: the spend that was earning least when it left. It buys results at its marginal cost, never at the line's average.
This page does not decide whether a cut was right or how deep it should have been, and it does not restart marketing after a business crisis, which the guide to leading marketing through a business crisis covers with its own pause list and restart order. Nor does it restart activity paused for capacity or for one market. Nothing here is accounting advice.
Before you start
- The restore trigger as agreed, with its signpost, level, response and approver, and the budget reallocation memo that recorded the cut: what moved, the marginal evidence, the expected loss and what would reverse it.
- The record of starting settings for every line: budgets, targets and bid strategies, exclusions added, ad sets consolidated, campaigns paused, tests stopped and flights shortened, each with its date. The Google Ads budget-cutting guide asks for this record at the cut; where it was never written, the account's change history, where the platform keeps one, and the memo are the next best source.
- The order in which the cut was made, and the result each step was judged on.
- The business's own results since the cut: orders, qualified leads or new customers, by line where the business can attribute them.
- The current target, and, where the business runs one, the agreement on a target-led paid media budget, whose ceiling, floor and stopping rule the restore works inside.
- The trading calendar for the coming months, with peaks, sales and holidays marked.
- The marketing budget calculator or a spreadsheet, to work from the results each restored line should buy, at the marginal cost per result it showed when it was cut, to the sum those results imply, and compare that with what the trigger returns.
The steps
Confirm the trigger has been met and what it returns
Read the signpost in the business's own records, over the period the trigger names, rather than in a platform report. A signpost that crosses its level in one strong week may be ordinary variation rather than a real change, and the guide to reading small numbers helps tell the two apart when counts are low. Then read the response exactly as written: how much returns, to which lines, from when, and whether it is the whole cut or part of it. If the trigger would send money to a line the business no longer wants to fund, or the reason for the cut has changed in a way nobody foresaw, take it back to finance instead of spending to the letter of the agreement. Write down the date the trigger was met, because the reports will need it as the start of a new period. Keep the wording of the restore triggerThe condition agreed with finance when a budget is cut that says when money comes back and to which line: a measured signal, a threshold and an agreed response. beside that date, so the final report can show what was agreed as well as what was done.
Read the record of starting settings and the order of the cut
List every change the cut made, in the order it was made, beside the setting it replaced and the result it was judged on. Two kinds of order sit in that list. The cross-channel order the cutting guide recommends is an order of value: capture spending at the margin goes first, tests next, then tools, then commitments at their break points, with the slow-to-rebuild work trimmed but kept alive and the team itself left until the end. Reversing it brings back first what the business least wanted to lose. The order inside a platform was chosen mainly to limit relearning: the Google cutting guide starts with exclusions and ends with pauses of spend that was poor all the way through. There, follow the value logic rather than running the sequence backwards. As general practice, bring back first the slices that earned best at the margin and leave the exclusions and the pauses of poor spend until last, or for good. Note which lines are still running at a lower budget: the guide to reducing spend on Meta advises lowering rather than pausing anything that will return, and, as general practice, a lowered ad set is the easiest line to restore. For booked media, the replanning tutorial for booked plans has the planner note, line by line, whether a cut line can still come back inside its campaign.
Check what has decayed while the money was out
Before any money moves, check the state of what it will buy. The learning phase guide records that Meta counts resuming an ad set after an extended pause among the significant edits that can restart the Learning PhaseThe period after launch or a significant edit when an ad set's delivery is still being optimised., and the Google cutting guide that a paused campaign brings its own learning period when it restarts. The guide to customer lists across ad platforms notes that a list stays exactly as uploaded until a new file replaces it, so a refresh routine that lapsed during the cut leaves exclusions missing recent customers and lookalike seeds describing the old base. The cutting guide's table of costs names a thinner pipeline of creative as one price of cutting production, and the ads that kept running have run for longer. As general practice, check the conversion tracking as well, since a gap that opened while spend was low is easy to miss. Write the findings beside each line in the record.
Decide the order of return from the marginal evidence
If the trigger names the first line, start there. Otherwise rank the returning slices by the marginal cost per result each showed at the cut, adjusted for anything the decay check found. Then look at timing. The cross-channel planning guide explains that demand creation has to lead demand, with the reach that persuades people at a peak bought in the weeks before it, while capture channels follow demand on their own, provided no budget cap holds them below it; a reach line with a peak approaching therefore moves up the order. The same guide warns that core channels short of money tend to eat the test reserve before anything else, so restoring the reserve is part of restoring the plan. Write the order down with an amount and an expected number of results for each line, worked out from the marginal cost rather than the average.
Change targets and budgets as separate steps
A cut that tightened a bid target cannot be reversed with budget alone. The Smart Bidding strategies guide explains that when the target is what holds a campaign back, a bigger budget does not make it spend more, and it records Google's documentation that large changes to targets or budgets can put Smart BiddingGoogle Ads' automated bid strategies that set each auction's bid from conversion signals. back into a learning period. Loosen such targets in small moves, as the Google Ads scaling guide sets out, keeping each move apart from any budget change and, as the Google cutting guide did when tightening, taking one campaign after another. For budgets, the budgets and pacing guide treats each campaign's budget as a daily average held inside a monthly limit, so set the daily figure that delivers the month's restored position and judge Budget PacingHow fast a campaign spends its budget across the flight compared with the plan. on the running total for the month, not on any one day. On Meta, a large budget change is another of the significant edits that the learning phase guide says can restart learning, so lift lowered ad sets a step at a time.
Bring paused campaigns back in turn
Resume a paused campaign where it still fits the plan rather than rebuilding it. The Google Ads account structure guide advises keeping old campaigns paused rather than deleted so that their history stays visible, and that history is part of what the restore reads. Resume one campaign or ad set at a time, outside sales and holidays, allow for the learning period that can follow, and leave its efficiency judgement until any such period has passed. Spend that the cut paused because it was poor all the way through stays paused unless new evidence says the line now pays. Where the plan has changed, with a new market, a new margin split or a new structure, build the campaign new and treat it as a launch with its own checks, not as part of the restore.
Restore prospecting before remarketing, on fresh lists
The prospecting and remarketing comparison sets out that prospecting and every other source of visitors fill the remarketing pool, that starving prospecting shrinks the pool and raises frequency on the people left in it, and that remarketing spend should follow the size of the pool rather than its reported return. A cut that took money out of prospecting therefore left a smaller pool, and money returned to remarketing first would reach fewer people, each of them more times. Restore prospecting first, let the pool refill, and bring remarketing back to the level the pool now supports. Replace the customer list files before the prospecting money arrives, so that the exclusions keep recent buyers out and the seeds describe the customers the business wants more of.
Return the test reserve and the slow-to-rebuild lines
The rule that splits a budget between proven, promising and experimental work keeps a share for experiments, and a cut that stopped tests left questions without answers. As general practice, restart a stopped test from its written design and read it as a new run, rather than joining results from either side of the gap. Restart creative production before the money reaches the ads that need it: restoring budget onto ads that have run through the cut raises frequency among people who have already seen them, which is how Creative FatigueThe decline in an ad's response as the same people see it again and again. sets in, and the creative production playbook sets the rhythm from brief to launch. Reach flights and booked media come back where the note made at the replan says each line still can, and anything newly signed becomes Committed SpendMoney already promised to suppliers through a signed order, booking or statement of work, tracked apart from what has been spent or invoiced., which the cutting guide counts against the remaining headroom. The slow-to-rebuild work the cutting guide protects, such as brand reach and the email programme, is restored as a written decision with its own timing, not as a side effect of the paid media steps.
Re-read the marginal cost at each step
After each step has settled and its late conversions have arrived, work out what the step added: its extra spend over the extra results the business recorded. Compare that figure with the target, using a Marginal ROASExtra revenue returned by the last increment of ad spend. where the target is a return. Where a step costs more than expected, the CPA calculator splits a cost per result into the CPM, the click-through rate and the conversion rate, to show which of them moved. Restored spend sits on the same curve of Diminishing ReturnsEach extra unit of spend or effort buys less extra result than the one before it. as before the cut, but the curve may have moved, so the recorded marginal cost is a prediction to check, not a result to assume. If a step's marginal cost passes the target, hold the line at its previous level and move to the next line in the order; the Meta scaling guide reads its budget steps the same way. Where the business runs a target-led agreement, its stopping rule decides how far each flexible line goes, whatever the trigger returned.
Report the restore against the trigger and the forecast
Report what the trigger returned, what was spent, the results each line produced in the business's records, and the marginal cost at each step against the forecast in the memo. Money that could not be spent at the target is reported as not spent, with the reason, rather than pushed into lines that no longer pay. Record each step in the marketing decision log with its date, its expected effect and the date it will be checked. At the company's next formal ReforecastA rebuild of the full-year view of the budget from actual spend and what marketing now knows, done at the points finance sets; not a fresh budget. point, the reforecasting tutorial carries the restored lines into the full-year estimate. From then on, the monthly paid media review checklist reads them in the normal rhythm. Keep the new starting settings: the next cut will need them.
What each kind of cut needs on the way back
The table sets each change a cut can make beside what it left in the account and how it is reversed. The rows follow the guides named in the steps; where a row says "general practice", it is this page's advice rather than a platform statement.
| What the cut did | What it left behind | On the way back | What to watch |
|---|---|---|---|
| Lowered the budget of a running campaign or ad set | History and learning intact at a lower level | Raise in steps, reading each one | Marginal cost per result at each step |
| Tightened a bid target | A campaign that will not spend more on budget alone | Loosen in small moves, one campaign at a time, apart from budget changes | Spend and results once each move has settled |
| Paused a campaign or ad set that was meant to return | A learning period that can follow when it resumes | Resume one at a time, outside peaks and sales | Efficiency only after learning has passed |
| Paused spend that was poor all the way through | Nothing worth restoring | Leave paused unless new evidence says it now pays | Whether the evidence is new, not remembered |
| Added exclusions | The cheapest-to-lose spend removed | Keep them; lift one only on evidence that the traffic pays at the margin (general practice) | The excluded traffic's results, where measured |
| Consolidated overlapping ad sets | Fewer ad sets, each with more data | Keep the consolidation; add new ad sets only with a reason (general practice) | Learning on any ad set that is new |
| Stopped tests | Questions without answers | Restart from the written design as a new run (general practice) | Whether the restored budget can reach a readable answer |
| Shortened or postponed a reach flight | Effects that may appear later in branded search, or not at all | Restore ahead of the next peak | Branded search, read on a lag |
| Thinned the creative pipeline | Older ads carrying the remaining spend | Restart production before the money arrives | Frequency and signs of fatigue |
| Let list refreshes lapse | Exclusions missing recent customers; seeds describing the old base | Replace the files before prospecting returns | The share of new customers in the business's records |
A beard care shop restores part of a spring cut (illustrative)
Kwibbolark Grooming restores £3,600 of a returned £6,000 a month, in order
Kwibbolark Grooming, an invented online seller of beard oils, balms and grooming kits, cut its paid media from £24,000 to £15,200 a month in spring to protect cash. Its target is £22 per order in the shop's own records. The cut trimmed category search from £8,000 to £6,500, tightened the Shopping return target so that spend fell from £6,000 to £4,500, paused one of two Meta prospecting ad sets and lowered the other so that the line fell from £6,000 to £3,500, reduced remarketing from £1,500 to £700, postponed a £1,500 video flight and stopped a £1,000 creative test. Finance agreed that £6,000 a month would return once the shop recorded more than 1,900 orders in a calendar month, first to Meta prospecting and the test.
The trigger is met in August. In the first week the owner replaces the customer list files and briefs two new creative concepts. The restore then runs over the following three months, one change at a time, each read once its late orders are in.
| Step | Change | Extra spend a month | Extra orders a month | Marginal cost | Decision |
|---|---|---|---|---|---|
| One | Lowered Meta ad set raised from £3,500 to £4,500 | £1,000 | 50 | £20 | Kept |
| Two | Paused Meta ad set resumed at £1,000, read after learning | £1,000 | 37 | About £27 | Paused again |
| Three | Creative test restarted from its design | £1,000 | Not judged on orders | Not applicable | Kept |
| Four | Shopping target loosened in two moves | £1,200 | 60 | £20 | Kept |
| Five | Category search raised from £6,500 to £7,500 | £1,000 | 35 | About £29 | Returned to £6,500 |
| Six | Remarketing raised from £700 to £1,100 once the visitor pool regrew | £400 | 19 | About £21 | Kept |
Monthly spend settles at £18,800, so £3,600 of the £6,000 is used. The owner reports the other £2,400 to finance as not spent at the target, and the reforecast takes it out of the quarter. The video flight is rebooked for the weeks before the Christmas peak from the next quarter's budget, and the new settings are recorded as the starting point for any later cut.
Common mistakes
- Switching everything back on the day the trigger is met. Learning periods can stack, and nobody can tell which line earned the money back.
- Restoring the settings instead of the results. The old budget and target are a record, not a promise; the marginal cost at each step decides how far a line goes.
- Adding budget where the target is the limit. Spend will not move until the target is loosened.
- Bringing back spend that was poor all the way through. Pauses and exclusions that removed it stay in place unless new evidence says otherwise.
- Putting the money into remarketing first. The pool shrank with prospecting, so the spend lands on fewer people, each of them more times.
- Restoring onto stale lists and tired ads. Replace the list files and restart production before the money arrives.
- Spending the whole sum because it was returned. Money that cannot be spent at the target goes back to finance at the reforecast.
- Not recording the restore. The next cut needs starting settings as much as this one did.
The optimisation hub sets restoring spend beside testing, scaling, reporting and the target-led budget.
