Q4 is when agencies stop marketing. That's why Q1 is quiet.

It's the second week of September, which means somewhere right now an agency founder is typing this into Slack:

"Right. Heads down for peak. Let's pick marketing back up in January."

The newsletter gets paused, social media budgets cut and events budget rolls into next year. And the reasoning sounds fine on the surface: our clients are flat out, nobody's hiring an agency in November, so why shout into an empty room?

Because the room isn't empty. It's the fullest it's been all year.

You've confused "not buying" with "not deciding".

Brands don't leave in January. They leave in November.

Setup runs an annual survey of brand and agency people. In 2023, 55% of brands said they were likely to switch their primary agency within six months. In 2024 it was 40%. That's a big slice of your market sat with one hand on the door.

More interesting is why they say they leave. In their latest data the top reason clients gave for firing an agency was delivery, at 48%. Agencies, asked the same question, ranked delivery seventh. Agencies think clients leave because of budget cuts and new CMOs. Clients say they leave because of what the agency did, or didn't do.

Now ask yourself when delivery gets tested. Not in a March workshop, it gets tested on Cyber Monday when the checkout is throwing errors and the account manager's phone is going to voicemail. It gets tested when the flows don't fire, when the site falls over, when the ad account gets cooked and the answer is "we're looking into it".

That's the moment the relationship gets marked. The brand doesn't fire anyone in December. They're too busy. But they've decided.

The search starts in January. The shortlist was written in November.

Anyone who's done agency new business for a while will tell you the same thing: Q1 is when the RFPs land.

So the lazy conclusion is: be visible in January. Be there when they search.

Except they don't really search. Bain and Google surveyed 1,200 B2B buyers and found that 80 to 90% of them had a shortlist of vendors in their head before they did any research at all. And 90% of the time, the winner came off that list. 6sense ran their own version and got 86%. The January "search" is mostly a formality. The buyer is looking for reasons to back a decision they've already made.

And when was that list written? During the eight weeks the brand was under the most pressure of the year, thinking about which agencies they wished they had.

The Ehrenberg-Bass lot call this the 95:5 rule. At any moment about 5% of your market is actually buying. The other 95% are getting on with their lives, forming opinions about you that they'll never share with you. January is the one time of year the 5% swells, because peak broke things. But the swell is made of memories from the 95% window. If you went dark for three months, you're not in anyone's memory. So you're not on the list, and you're not in the room.

Going dark is expensive.

Peter Field's work for the IPA on the 2008 recession looked at what happened to brands that cut their share of voice when things got hard. Short version: they lost share, became more price sensitive, and some took up to five years to recover. The brands that held their nerve got attention cheaper than at any other time, because everyone else had gone quiet.

That's a study about consumer brands in a downturn. Same mechanism for an agency in Q4, though. Everyone else in your category is heads down. Your competitor's newsletter is on pause. Their founder hasn't posted since October. The attention you can get in November is the cheapest attention of the year, and it lands on people who are, at that exact moment, forming a view about who they trust.

So who do they actually go to?

Not the ones who disappeared. Not the ones who resurface on the 8th of January with "Hope you had a great peak! Would love to chat about your 2027 roadmap."

That email gets deleted with something close to contempt, because where were you in November?

They go to the agency that was in their inbox on a Tuesday in November saying "here's what's breaking on Shopify this week and here's what we're doing about it". The one who told them not to ship that change on the Thursday before Black Friday. The one whose founder wrote up a client outage and what they learned while it was still raw, rather than a case study in March with the bruises airbrushed out.

The Bain research is useful here too. The main reasons a vendor makes the day one list are a previous positive experience and a recommendation from a colleague. Neither of those happens by accident. They happen because someone was useful, visibly, at a moment when it counted.

In a crisis, the most valuable thing you can sell is the absence of panic. Everyone else in the brand's inbox during peak is either selling something or apologising for something. The calm, useful voice stands out because it's the only one.

What to actually do

Don't stop. Drop the volume if you must, but don't go dark. A fortnightly newsletter that keeps going through November beats a weekly one that vanishes.

Change what you're saying. Nobody wants "Five ecommerce trends for 2027" in the middle of peak. They want a war reporter. What are you seeing across your clients this week? What's breaking? What's the fix? What would you do if you were them, right now?

Be the calm one. You'll be tempted to be loud because everyone's stressed. Don't. Be the person whose emails get opened because they're never a waste of time.

Publish your own post-mortem before they do theirs. A December piece on what went wrong on your accounts and what you'd do differently is the most credible thing you'll write all year. Honesty in December is worth ten case studies in March.

Book January now. The advice every agency gives brands about peak is "book the real engagement for January". It works in reverse. The January meetings you want are set in November, while you're still the one being useful.

The agencies with a full Q1 pipeline aren't the ones who worked hardest in January.

Sources

Setup, Marketing Relationship Survey: eight years of data on why clients leave agencies. setup.us
Setup 2023 survey (55% likely to switch within six months). setup.us
Search Engine Journal on Setup 2024 data (40%, value as top reason). searchenginejournal.com
Everything PR synthesis of Setup's 2025 survey (delivery 48% vs agencies ranking it seventh). everything-pr.com
Bain and Google, "What B2Bs Need to Know About Their Buyers", HBR. hbr.org
6sense replication of the day one list finding (86%). revcity.6sense.com
LinkedIn B2B Institute / Ehrenberg-Bass, the 95:5 rule. business.linkedin.com
Peter Field on advertising in a downturn (Campaign Canada). campaigncanada.ca
Association of Canadian Advertisers on Field's "Long, Short or Dark" (five year recovery). acaweb.ca
Agency search volume seasonality (Google Keyword Planner analysis). jheenan.com
Needle, January 2026 DTC report (30 to 40% paused ads, CPMs down 40 to 60%). askneedle.com