Scale Ad Spend Without Blowing Up Your CAC (Part 3 of 3)
How to respond when scaling spend goes badly. Which it will, because this is marketing and nothing works the way you thought it would.
In partnership with Ballpoint
I’m writing this series with Josh Lachkovic, managing director at Ballpoint, a performance marketing agency I worked with at Yonder. If you’re thinking about scaling your performance spend and want expert help, they’re worth a conversation.
Congratulations. You made it to part three of three. You’re evidence that TikTok-brain hasn’t fully set in for the best of us. There’s still just enough room up there to learn something useful (hopefully) before some genius in Austria builds a Claude Skill that replaces you entirely, and you can finally join that free love nudist colony you read about in Nat Geo once.
Anyway. Ads.
So far we’ve covered a lot of ground. You now know when to scale. You’re an expert in setting up the kind of tracking that gives your CFO butterflies. And you known that creative is fundamental to the whole thing.
Sounds like you’re doing well. That’s great. Let me ruin that for you.
Parts one and two were the easy bits. Processes, at the end of the day. Glorified to-do lists your office dog could get through with enough treats on offer.
Part three is different. It’s all about how you respond when things go badly. Which they will, because this is marketing and nothing works the way you think it will. What you do when that happens is ultimately the measure of your marketing ability. It’s why they pay you the medium bucks.
I am, as it turns out, exceptionally qualified to write about things going badly.
Here’s what we’ll cover:
What to do when CAC balloons and you don’t know why
When to hire in-house versus work with an agency
Let’s get into it.
What to do when CAC goes the wrong way
“Every avenue of growth (markets, products, channels) suffers diminishing returns. Our job as growth people is to slow that down. With ad channels, the biggest levers we have to slow those down are creative and the audience we’re speaking to. There are always external factors at play, but these are the ones you control.”
– Josh
CAC is going to go up at some point. Better to accept it now than act surprised when it happens. See this as an ongoing dance with your ad platforms as part of your job and just get on with it.
Here are a few reasons why and what to do about it:
Creative fatigue
The most common culprit. You’ll see frequency climbing (Meta being forced to spend your budget by showing your same creative more), CTR declining and engagement dropping. Your ads stop winning because everyone’s seen them. The underlying mistake is almost always that you scaled spend before your creative engine could keep up (you can’t increase one without the other), or you kept running under-performers longer than you should have because you were hoping they’d turn around.
How to fix it → more creative, pls.
Optimising for Click Through Rate (CTR)
A trickier problem to solve because it doesn’t look like a problem. Click-through rate looks good in a dashboard. Some product manager out there will think it’s great. But you and I both know that clicks don’t pay the bills, baby. CTR is a hint at acquisition cost, not a measure of it. If the goal of your ad creative is acquisition (and not say, brand building), then measure them on acquisition and not clicks.
How to fix it → measure your ad performance through full funnel acquisition
Assuming all conversions are incremental.
This doesn’t necessarily increase your CAC, but it hides a bad one by taking credit for conversions that would have happened anyway. It’s a slippery slope because it convinces people in the business that Meta is far more effective than it really is. Running holdout experiments where a group sees no ads is the only way to understand what you’re actually buying.
How to fix it → run holdout experiments to measure the real channel CAC
Audience saturation
Looks different from creative fatigue. If narrower targeting is suddenly outperforming your broad campaigns, or repeat exposure keeps climbing even on fresh creative, you’ve probably exhausted a meaningful chunk of your reachable audience. New ads can fix this but at some point you’ll need to look at adding in new channels. Josh reckons you can build a £25M revenue business on Meta alone, so I wouldn’t rush to launch a new channel if you’re still small.
How to fix it → launch new channels, new products but honestly more creative will help too
Seasonal effects
Not a lot you can do about these, but worth paying attention to and getting ahead of. Unless Black Friday is an important period for your brand, I’d look at switching off ads altogether. Christmas could be the same. Use that money more effectively elsewhere.
If CAC went up at the same time last year and there’s an industry-wide pattern (at Yonder we saw new credit card applications spike in the new year) it’s probably not your fault and it’s probably not fixable. Worth communicating this broadly so it doesn’t look like an excuse when CAC spikes or you pull back spend. If your CEO is worried, ask them: “We can keep spending in November if you want, but are you okay with doubling our CAC for that time?”
How to fix it → fluctuate spend throughout the year to avoid expensive periods
Platform change
Again, not much you can do here. If you’re working with an agency like Ballpoint, they can help flag these ahead of time as they’ll have the added benefit of seeing multiple ad accounts at once. Most ad platforms like Meta, Google and TikTok change their ad algorithm often.
How to fix it → keep an eye on them and adjust accordingly
Competitive pressure
This can happen either on product-market fit, or in ad spend, but both hurt you slowly. If you’re finding CAC going up, it may just be that your product isn’t as competitive in the market as it was a few months ago. Or that your competitors have been flooding that ad platform with ads and you can’t show up as much. You’re fighting harder for the same impressions. Make your ads look different enough from the category to try and win on distinctiveness, however there’s no replacement for improving your product.
How to fix it → improve your products and launch new creative. Creative really does solve a lot of your problems, if you haven’t noticed.
Poor documentation and experimentation
This will slowly catch up with you. Make sure you’re diligent in writing down all your learnings. Maintain a strict experimentation database in Notion or similar and record how each experiment works. Test one influencer against another? Write up your hypothesis, attach the creative and then record your learnings. These will compound over time and help when you grow your team so they can see what you’ve done before.
How to fix it → build an experimentation table and update it weekly
Finally, not managing internal expectations
Send this to your boss, and their boss. Everyone in your organisation needs to understand how this stuff works. Your CAC will go up and it can’t be a surprise to anyone when it happens. If the first time leadership hears that performance doesn’t scale linearly is the week CAC goes up 30%, you’re off to a bad start.
How to fix it → never, ever, ever shut up about how marketing works to anyone ever
When to hire someone full-time, and when to use an agency
“A performance marketing team today is really a marketing strategist or CMO, an insights manager or planner, a copywriter, a creative strategist, a creative designer or editor, a creative operations manager, and a data analyst. The role that was one person at £10k/month is now effectively ten people.”
— Josh, in Part 1, in case you skipped it.
All going well and your spend continues to go up, you’ll need to grow your team. And you’ll be faced with the age old question of whether to hire into the team or work with experts from outside your business. Or the third option, which is to use AI to do it all, before it capitulates and you’re back to options one and two.
The annoying answer is, it depends, but here’s a way to think about it.
Most agencies will charge on a percentage of spend. I don’t know what the going rate is now, but assuming it’s 10-15%, then it generally makes sense to work with an agency until you get to around £80-100k a month which would be enough to hire a senior performance marketing lead.
There are nuances to this obviously so here’s some of the trade offs you can expect.
The benefits of hiring in-house
Things like product knowledge, brand context, faster day-to-day iteration, and someone who’s always around and accountable when things go wrong. These are real advantages. Your in-house person will understand your product and your customers in a way no external team ever fully will, and that knowledge eventually starts showing up in better creative and smarter targeting decisions. You’ll also just get the added benefit of them caring more.
The benefits of hiring an agency or freelancer
You’ll get out-of-the-box platform expertise, creative production volume, and institutional knowledge built from working across dozens of accounts. A good agency has seen your problem before, on several other clients, probably in the last six months. That pattern recognition has genuine value and it takes time to build in-house.
If you do go in-house, hire for creative taste and analytical rigour before you hire for channel credentials. Meta certification matters less than someone who can look at a week’s worth of creative data and know intuitively what to test next. And don’t hire too junior expecting them to figure it out — you need someone who’s done this before at meaningful scale, or you’re adding a learning curve to an already expensive problem.
The benefits of doing a little of both
You have your performance marketer in-house who owns strategy, planning, oversight and depending on their seniority, is ultimately accountable. A specialist agency and some freelancers manage the creative production and channel expertise. This is what we eventually got to at Yonder and it worked well.
Be careful to hire an expert media buyer (someone who can work the ad platforms) before having the creative infrastructure to support them. A brilliant media buyer with bad creative is still going to get bad results. Creative solves all your problems, remember?
How to work with an agency or freelancer
I know, I know. Ballpoint are sponsoring this series and they’re an agency. How could you possibly be impartial here, Tom? They’re paying your rent. First off, I wish Josh was paying my rent. And second, I asked them to sponsor this after I decided to reflect on how things went with them. The order is important.
Agencies aren’t always the answer, and Ballpoint are great but they may not be the right one for you. But I have two years of direct experience working with them, and I have some strong opinions on what makes an agency relationship actually work versus quietly not work.
I’ve already written about this here:
What agencies are good at
Creative testing velocity, platform expertise that’s current because they’re living in Ads Manager for multiple clients simultaneously, and a structured approach to experimentation that most in-house teams don’t build until they’re much further along.
What agencies are not good at
Knowing your product and brand as well as you do. Understanding the nuance of your customer, the politics of your brand, the context behind a brief. That’s your job. The way you close that gap is by writing genuinely good briefs, showing up to the weekly sync, and not sitting on feedback for two weeks. I was bad at this at first. Very bad. Josh was too polite to say it directly, but I was the bottleneck. Don’t be the bottleneck.
Ultimately, agencies just aren’t incentivised like in-house people are. Their businesses rely on bringing on as many clients as possible. So assume that the best you’ll ever get out of an agency is 60-70% of what a full-time employee will do. It’s just the nature of how their businesses work.
Naturally, Josh does disagree her.
"While you may get less time than a FT employee, we've often found we outperform in-house team because of the collective experience you get. You basically get access to £1m of salary working on your business when working with us. It's only a % of their time, but our goal is to make sure our fee provides better return than the in-house option."
– Josh
What to look for in a performance agency
Look for a track record of success with similar brands. Perhaps similar business models or growth stages. They should be able to tell you specifically what they did and what changed, not just show you a slide with big percentage improvements and no context.
Make sure you maintain full platform access. Any agency that won’t give you full visibility into your own accounts is not an agency worth working with. You should be able to see everything, any time.
Ask whether they do creative strategy and production or just media buying. At this scale, the creative and the targeting are not separate problems. An agency that thinks in both is far more valuable than one that’s just good at bid optimisation.
Josh’s approach when we started working together was to spend time auditing our setup before suggesting a single change. He wanted to understand what we had before telling us what to do differently. That rigour, the willingness to look before touching, is something I’d look for in any agency conversation.
“A lot of agencies come in and want to restructure things because it looks like progress. But you risk throwing away things that were actually working. A good agency will spend time listening before changing anything.”
— Josh
Red flags
Agencies that won’t share access, don’t send regular reporting, overpromise on specific outcome numbers, can’t explain their methodology clearly, or disappear between monthly calls. And if they’re presenting stuff in a deck each week, they’re wasting their time and yours.
Wrap it up Tommy, I’ve got work to do
Thanks for sticking with me through three parts. I’ve never written a series before, and the honest reason it became one is that I tried to write it as a single post and it was absolutely enormous. I hope the extra weeks were worth it.
Here’s where we’ve landed
In Part 1, we covered what has to be true before you scale. Your unit economics need to be solid, your product-market fit signals must be strong, your budget needs to be able to absorb a CAC spike, and your leadership team needs to understand what they’ve signed up for.
In Part 2, we went deep on the machine itself. Server-side tracking and CAPI, the metrics that actually matter, building a creative production system that can keep pace with your spend, and when to kill an ad.
In this final part, we’ve covered what goes wrong, how to diagnose it, and how to think about building the team around it.
Ultimately, this whole thing is one big system you can optimise. The measurement, the creative engine, the team structure, the expectation management - none of it works in isolation. Get them working together and it gets a lot less stressful. Not stress-free. Less stressful.
If you want to talk through your specific situation with Josh Lachkovic and the team at Ballpoint, then I’d encourage you to reach out or check out his Substack where he writes about all of this in more depth.
In partnership with Ballpoint
Hi, I’m Tom.
I’ve launched products at Monzo, Wise, and built the brand and marketing team at Yonder from scratch. I write about what actually works in startup marketing for marketers on the verge of breakdown.
If you want help understanding why your brilliant product isn’t selling itself, find me on LinkedIn.