There’s a scene that plays out in marketing departments everywhere. A marketer pulls up their dashboard, sees a ROAS of 8:1, and feels momentarily victorious. The CFO nods approvingly. The spreadsheet glows green. Everyone’s happy.
Except the business isn’t actually growing.
This is the paradox Andy Griffiths, Chief Growth Officer at Climb Agency, sees repeatedly: companies drowning in data while dying of thirst for actual insight. In our latest Browse Basket Buy conversation, Andy unpacked why so many eCommerce brands are optimizing themselves into a corner—and what they can do about it.
The comfortable prison of brand terms
Andy talked about “the Google tax”—the portion of your budget spent bidding on your own brand name. It’s defensible, sometimes necessary, and utterly seductive because it delivers spectacular numbers.
“No one’s saying [brands] shouldn’t be bidding on their own terms,” Andy explained, using Fender guitars as an example. “However, a complete religious pursuit to bid as much as they possibly can on that kind of term would potentially be at the expense of going more aggressively after someone bidding for ‘electric guitar’ or ‘best electric guitar for [insert genre here].’”
The problem isn’t the tactic—it’s the addiction. Branded search converts beautifully because you’re capturing people already looking for you. But you’re not creating new demand. You’re just paying to intercept customers who were coming anyway.
The mathematics are cruel: spend 80% of your budget protecting existing demand, and you’ve only got 20% left to actually grow. Yet this is precisely what happens when CFOs with spreadsheets set marketing budgets based purely on ROAS metrics.
The education gap
Andy was diplomatic but direct about the gulf between those who allocate budgets and those who spend them. “The skill or understanding gap between those who are stipulating budget and those asked to carry it out is quite often a big gulf.”
It’s not about CFOs speaking the language of CMOs (or vice versa)—it’s about having the analytics infrastructure to demonstrate what’s actually driving incremental revenue versus what’s just mopping up demand that already exists.
This requires tools that can track true lifetime value, understand customer journeys beyond last-click attribution, and model what would happen if you reallocated spend.
The technology exists. The willingness to implement it, apparently, is harder to find.
Be more like a bee
When I asked Andy about encouraging teams to take more risks, he invoked one of Rory Sutherland‘s best observations: the waggle dance of bees.
When scout bees return to the hive, they perform a dance indicating where to find pollen. But only about 50-60% of the other bees follow that advice. The rest go exploring elsewhere.
“If they didn’t go off elsewhere,” Andy noted, “they wouldn’t find new sources of pollen.”
It’s a deliberate strategy for survival, not a failure of communication.
Yet most marketing teams operate with the opposite philosophy: 100% of the budget chasing the known, proven tactics. The problem isn’t that these tactics don’t work—it’s that they stop working eventually, and you’ve built no capacity for discovery.
Andy’s recommendation is stark in its simplicity: designate 15-20% of your budget specifically for test and learn. Not “we might try something different at Black Friday” but a systematic, protected allocation for experiments that might fail.
“There has never been more opportunities to test something new that could give you a competitive advantage through early adoption,” he argued.
The AI tools, the attribution models, the optimization platforms—they’re all there. What’s missing is the deliberate intent to use them.
The operational revolution
When we discussed AI, Andy’s perspective was refreshingly grounded. Forget the hype about AI replacing marketers—the real value is in removing the tedium that prevents good marketing.
He pointed to brands like Sweaty Betty generating thousands of ad variants, letting AI handle the creative combinations while humans focus on strategy. “Thousands of variants, probably tiny little differences between them, but they’re really finding out what drives performance.”
This isn’t about AI being creative. It’s about AI being tireless. The holding companies can now profitably serve mid-market clients because they’re not paying humans to manually build campaign after campaign. Agencies can pivot budgets between channels in days, not weeks, because the operational overhead has collapsed.
The winners, Andy suggests, will be those who use AI to eliminate the administrative burden and reinvest that capacity into strategic thinking.
The losers will be those who keep paying for legwork in a world where legwork is free.
One thing you can actually do
I always end these conversations asking for something practical, something listeners can implement immediately. Andy’s advice surprised me in its specificity:
“Give GA4 another crack.”
Two years after Universal Analytics died, he’s still encountering marketers who’ve retreated to using their CRM or Shopify as their “single source of truth” because they found GA4 too frustrating to master.
“Yes, it’s a pig. Yes, it’s less user friendly. Yes, it needs more customization. But I promise you it’s not as hard as you think it is and if you adopt it, there’s massive benefits.”
The discrepancies between your ad accounts, your eCommerce platform, and what you’re reporting to finance will undermine every attempt at sophisticated testing or reallocation. GA4 isn’t perfect, but it’s the connective tissue that lets everything else work.
Returns, not metrics
The conversation circled back repeatedly to Andy’s core philosophy: focus on returns, not metrics. This doesn’t mean metrics don’t matter—it means they’re servants, not masters.
Your ROAS can be spectacular while your business slowly suffocates. Your CTR can be enviable while your customer acquisition cost spirals. Your conversion rate can trend upward while your market share evaporates.
The hard work isn’t generating numbers. It’s building the analytical capability to understand which numbers actually indicate growth, and then having the organizational courage to optimize for those instead of for what makes the spreadsheet look good.
It’s the difference between feeling successful and being successful. And in eCommerce, that gap can be expensive.
This article was written with the assistance of AI.
