A sales decline is a symptom, not a marketing diagnosis
The finance view is immediate: marketing is a discretionary line and reducing it preserves cash this month. The commercial consequence is delayed. Fewer prospects may enter the market now, but the lost enquiries, opportunities and customers often appear later.
That delay creates two common mistakes. One business keeps every campaign because “marketing is an investment,” even when the offer, measurement or economics are broken. Another cuts everything because sales fell, removing productive demand before it has diagnosed pricing, sales follow-up, stock, service capacity or a wider market slowdown.
My verdict: protect the demand that works, remove spend that cannot earn the right to continue, and fund repairs before buying more volume. A cut should be a commercial decision about customer creation and cash, not a blanket percentage handed to every channel.
The most recent IPA Bellwether summary, published 16 July 2026, found UK companies still revised total marketing budgets upward in the second quarter despite difficult conditions, while becoming more selective about where investment could create business impact. That is useful context, not a rule for your company. Your liquidity and unit economics remain the first constraint.
The Revenue Defence Decision: five tests before any cut
I use five tests to separate a cash decision from a marketing opinion. Leadership should answer them with one shared view from finance, sales and marketing.
Liquidity
How much cash must be preserved, by when, and what contribution can each customer source support?
Demand
Are fewer suitable buyers entering the market, or is the business losing attention and preference?
Conversion
Are relevant prospects arriving but failing because of the offer, price, proof or buying experience?
Sales capacity
Are response time, qualification, follow-up, delivery or inventory limiting revenue after marketing?
Future demand
What customer and market-position cost will appear in 30, 90 or 180 days if discovery stops now?
Read cohorts, not only the current month
Compare customers and qualified opportunities by the period in which marketing first created or captured the demand. A slow month may contain revenue from earlier investment, while this month's spend may not mature until the next quarter. Google Ads explicitly documents conversion lag: the delay between an ad interaction and the conversion can make current cost per acquisition look worse and return on ad spend look lower before later outcomes arrive.
This does not excuse weak performance. It means the review window must match the buying cycle. Reconcile the CRM, transaction records and marketing data before concluding that the newest spend caused the newest sales result.
Protect, repair, pause or stop: the decision matrix
The framework is deliberately more demanding than “keep what has the best ROAS.” Reported return can favour brand search, existing customers and activity that captures demand created elsewhere. Use commercial evidence and the role of the investment together.
Protect
Keep the smallest effective level of reliable customer acquisition, measurement, sales handoff and future-demand activity.
Repair
Fix message, offer, proof, landing experience, follow-up or CRM feedback before paying for more reach.
Pause
Freeze an experiment with a named restart condition when cash is tight or the decision window is not complete.
Stop
End duplicated, ownerless or commercially weak activity and record why so it is not restarted by habit.
| Business symptom | Evidence to inspect | Likely decision |
|---|---|---|
| Sales down, qualified pipeline stable | Deal age, win rate, pricing, sales capacity and delivery constraints | Protect demand; repair the revenue step |
| Traffic stable, qualified enquiries down | Buyer mix, message, offer, page conversion and tracking changes | Repair before adding spend |
| Demand and conversion both down | Category demand, competitor movement, share of search and customer interviews | Protect core discovery; change proposition or market |
| Customers won but contribution is weak | Gross profit, discounting, fulfilment, returns, servicing and payback | Reprice, narrow or stop the source |
| Evidence is unreliable | CRM definitions, duplicates, missing offline outcomes and conversion delay | Hold scale; protect measurement |
If acquisition cost is the visible pressure, use the CAC Pressure Map. If activity cannot be connected to a business result, use the Marketing Proof Stack. The guide to setting a paid advertising budget and the B2B Ads Budget Calculator help rebuild the budget from economics rather than history.
A 30-day response when cash and sales are under pressure
Reconcile
Define the cash requirement, buying cycle and commercial stages. Match spend to qualified demand, customers and contribution.
Repair and reallocate
Close obvious leakage, protect reliable demand and move budget only where evidence and capacity support it.
Decide
Approve each line as protect, repair, pause or stop, with an owner, threshold and next review date.
Google's budget-agility research argues for collaboration between marketing and finance, shared business outcomes and the ability to shift resources towards better opportunities. The useful principle is flexibility, not constant intervention. Reallocating every few days without enough evidence can make the cause of improvement impossible to learn.
Keep one written decision record. For every retained line, state its business job, evidence, monthly limit and failure condition. For every pause, state what will trigger restart. For every stop, record which assumption proved wrong.
My growth partnership services connect paid acquisition, conversion, measurement and CRM feedback. Review documented case studies, compare paid ads and SEO by commercial constraint, see how I work directly, or request a 48-hour diagnostic.
When a deeper marketing cut is the right decision
A business should cut hard when survival requires immediate cash, when the product cannot be delivered profitably, when the sales team cannot accept more demand, or when sufficient evidence shows the market, proposition or channel is not viable. Marketing cannot rescue negative contribution or missing product-market fit by spending faster.
However, distinguish a temporary liquidity action from a claim that marketing was ineffective. State the consequence honestly: the business is choosing cash now and accepting a possible demand cost later.
Ehrenberg-Bass Institute research tracked 70 competitive consumer brands over more than 20 years and identified 57 cases of stopping mass-media advertising for at least a year. Sales fell by an average 16% after one year and 25% after two, with wide variation and faster decline among smaller or already declining brands. The authors explicitly note limits to generalising from one study and do not claim the stop was the only cause. I use this evidence as a warning against going completely dark—not as a forecast for an individual business.
Practitioner note: when sales slow, I first look for the break between demand, conversion and revenue. Cutting the channel before locating that break may preserve spend while leaving the actual constraint untouched.
Sources and evidence notes
Sources were checked on 12 August 2026. The Revenue Defence Decision, five tests and Protect / Repair / Pause / Stop matrix are original ThomPerformance analysis. Search prioritisation is qualitative; no search volume, universal budget percentage or client result is claimed.
Frequently asked questions
Should a small business cut marketing when sales are down?
Do not make an automatic percentage cut. First identify whether the sales decline comes from weaker demand, poor conversion, sales capacity, pricing, delivery constraints or unreliable measurement. Protect activity that repeatedly creates profitable customers or qualified pipeline, repair fixable leakage, pause unproven tests and stop spend that has sufficient evidence of weak commercial value.
Which marketing costs should be cut first?
Start with activity that has no defined business job, no owner, no usable measurement and no credible path to customers. Then review duplicated tools, unused retainers, overlapping channels and experiments that passed their decision date without evidence. Do not remove measurement, customer research or productive demand simply because those lines are easy to cancel.
How long should marketing continue before it is judged?
Use a window that matches the buying cycle and conversion delay. Inspect tracking, audience and response early, then allow enough time for qualified demand and revenue to appear. Set the evidence threshold and review date in advance. Waiting indefinitely is not discipline, but judging before customers could reasonably convert is not evidence either.
Should I keep brand marketing during a downturn?
Maintain a defensible level of future-demand activity when cash and economics allow, especially if the business is small or already losing demand. The right level depends on market position, buying frequency and liquidity. Evidence from consumer brands warns against a complete long-term advertising stop, but it should not be treated as a universal budget formula for every B2B or service business.
What should a CEO ask before approving a marketing cut?
Ask how much cash the cut preserves, which customer or pipeline source it removes, when the sales effect is likely to appear, what evidence supports the decision, who owns the repair and what would trigger restart. Require marketing, sales and finance to use the same stage definitions and review window.
Cut waste, not the business's route back to demand
Start with cash and customer economics. Then separate demand, conversion, sales capacity and timing. Protect what repeatedly creates viable customers, repair the system around relevant demand, pause incomplete bets with a restart rule and stop activity that has enough evidence of failure. That is a defensible budget decision; an equal percentage cut is not.
Which budget line can your team prove should be protected—and which one is surviving only because nobody has defined its job?