The email arrives on a Tuesday. The board wants 30 percent out of marketing by the start of next quarter.
What happens next is almost always the same. The marketing lead opens the budget spreadsheet, goes line by line, and takes roughly a third off each row. Ad spend down. Content down. Events down. Tools down. Everything trimmed, nothing killed, and the number lands where finance asked.
That is the worst available answer, and it is the most common one.
Why the across-the-board cut is the worst option
An even shave takes every channel below the level where it works, while removing none of them from the ledger.
Your paid campaigns now have too little budget to gather useful data, so they keep spending and stop teaching you anything. Content drops from weekly to whenever, which reads to both readers and search engines as a site that lost interest. The events you still attend are the cheap ones nobody good goes to. You end up paying for six broken things where you could have run two good ones.
A budget cut is a prioritisation decision. The even shave exists specifically to avoid making it, because making it means telling someone their channel is being switched off.
There is a second reason it happens. An even cut is defensible in a meeting. Nobody can accuse you of playing favourites, and every channel owner suffers equally, which feels fair. Fairness across internal teams is not the objective. Keeping the business generating pipeline on less money is.
Sort every line into two buckets
Before you cut anything, split the budget into two groups.
Switchable spend stops producing the day you stop paying and starts producing again the day you resume. Paid search. Paid social. Sponsorships. Most software. Contractor hours. Turn it off Friday, turn it on in six months, and you are roughly where you left off.
Compounding assets keep producing after the spend stops, and they decay slowly rather than instantly. Content that already ranks. An email list of people who open. Clean CRM data. Reviews and case studies. Measurement that lets you tell what worked. These take months or years to build and months or years to rebuild.
The rule is simple. Cut switchable spend hard. Protect compounding assets even when protecting them looks expensive this quarter.
Most marketing budgets are cut in exactly the opposite order, because the compounding assets have the softest short-term numbers. Nobody notices when you stop publishing. Everybody notices when the lead form goes quiet.
Do the sorting on paper before anyone starts negotiating. Open the budget, add one column, and write "switchable" or "compounding" next to every line. It takes twenty minutes and it changes the conversation, because the argument stops being about whose channel matters and starts being about what you can put back later.
A few lines will not sort cleanly, and that is useful information too. A retained agency that both runs your ads and writes your content is two things billed as one. Split it on the page so you can cut half of it.
What to protect
Four things, in this order.
Your existing content library and the site it sits on. Pages that rank today keep bringing in traffic next month whether or not you publish anything new. That library is an asset you already paid for. Reduce publishing frequency if you must, from four posts a month to one, but do not go to zero and do not let the site sit untouched for a year. Also keep a small budget for fixing broken things, because a site with slow pages and broken redirects loses ground quietly.
The email list and whoever writes to it. Email is the only channel you own outright. No algorithm change, no auction, no platform decision. A list that goes silent for eight months does not come back at the same open rate, and warming it up again takes longer than most people expect. Cut the design budget for the newsletter before you cut the newsletter.
Customer data and CRM hygiene. This is the least glamorous line in any marketing budget and the most expensive to restore. When nobody maintains contact records for a year, you get duplicates, dead addresses, missing lifecycle stages, and reporting that nobody trusts. A CRM that becomes untrustworthy takes a project to fix, not a cleanup. Keep whoever does this work, even at reduced hours.
Measurement. Your analytics setup, your dashboard, and the person who reads it. This one is counterintuitive because reporting produces no leads. But a cut budget makes every remaining dollar a decision, and you cannot make those decisions without knowing what is working. If you have already built the dashboard that shows where pipeline actually comes from, that is the tool you use to defend everything else on this list.
What to cut first
Start with the software. Pull login data for every tool in the stack over the last 90 days. In almost every audit I run, there is a tool nobody has opened since a project ended, a second tool that duplicates something HubSpot already does, and a set of per-seat licences for people who left. Total that up before you touch anything else. It is usually more than the team guesses.
Then paid media in channels you cannot connect to pipeline. Not paid media generally, only the parts where the data does not show a line to closed revenue. Every account has some of this: a campaign that has been running so long nobody questions it, a retargeting audience that mostly re-serves existing customers, an awareness campaign bought on impressions.
Then events and sponsorships booked out of habit. Ask what each one produced last year by name, in deals. If nobody can answer, that is the answer.
Then agency or contractor work that duplicates something internal. Two people writing similar content, or an agency running reports your own dashboard already produces.
Every one of these can be restarted. That is the point. You are choosing the cuts that are reversible over the cuts that are not.
Two things people forget while cutting. Check contract end dates before you plan anything, because an annual licence that renews in three weeks is a decision you are making this month whether you meant to or not. And when you cut a programme, say clearly what happens to the person who ran it. If the paid budget goes to zero, the person managing paid needs a different job description on Monday, and telling them beats leaving them to work it out from the spreadsheet.
How to defend the list to a CEO
Do not argue about marketing philosophy. Price each item by what it would cost to rebuild it.
A paid search campaign costs nothing to restart. You turn it on and it runs. A content library that has ranked for two years, if abandoned and then restarted, costs you writer time plus six to twelve months of waiting before traffic returns to where it was. An email list left cold for a year comes back at a lower open rate and takes months to warm. CRM data left unmaintained becomes a project.
Now present the same cut number two ways. Option A hits the target and keeps the compounding assets. Option B hits the target by cutting the assets and keeping more of the switchable spend, and next to it, the rebuild cost when the budget returns. One is arithmetic. The other is an opinion. Give them the arithmetic.
One more thing to put in front of the CEO before the cut lands: the delay. B2B sales cycles run 60 to 120 days in most of the businesses I work with, so cutting demand generation in October shows up in closed revenue in February. Say that out loud in the meeting where the cut is made. Otherwise the January review shows revenue holding steady, someone concludes the cut was free, and a second cut follows.
What to do with the money you keep
A smaller budget removes the option of doing six things adequately. Take it.
Pick the two channels with the clearest line to closed revenue in your own data, fund them at full strength, and stop the rest on purpose. Not starved. Stopped. A channel running at 30 percent of what it needs produces almost nothing and still costs 30 percent.
Most teams find this clarifying once the initial panic passes. The budget cut forces a decision they should have made two years ago and kept deferring because there was enough money to avoid it.
The takeaway
Sort the budget into what stops when you stop paying and what keeps working after. Cut the first hard. Protect the second even when it is uncomfortable this quarter.
Then take the cut list to the CEO priced by rebuild cost, not by run cost. Anyone can cut 30 percent. Cutting 30 percent that you can put back is the actual job.