The Scaling Paradox

Here's the pattern almost every ecommerce brand follows. You find a channel that works, usually Meta or Google. You scale spend. Revenue grows. Life is good.
Then at some point, the economics start to shift. You're spending more but the return feels softer. ROAS drifts downward. CAC creeps up. The marketing team says the campaigns are performing well. Finance says margins are shrinking. Both are looking at different numbers, and both are right from their perspective.
This is the scaling paradox, and it hits almost every brand somewhere between £5M and £15M. The issue isn't that marketing has stopped working. It's that the way you're measuring it has stopped being adequate.
At £2M in spend, you could get away with trusting platform reported ROAS. Meta says it delivered a 4x return, Google says 5x, and you believe both because the bank account is growing.
At £10M+ in revenue, with multiple channels, campaigns, and customer segments, this breaks down for three reasons:
Platform over-reporting. Every ad platform takes credit for more conversions than it actually drove. Meta counts view-through conversions that might have happened anyway. Google claims brand search conversions that your SEO would have captured. When you add up the revenue each platform claims, it often exceeds your actual total revenue by 30-50%. Someone is lying or more precisely, everyone is double counting.
Blended metrics hide channel level truth. Many brands fall back on blended ROAS or blended CAC as their north star. This is comfortable but dangerous. A blended 3x ROAS might mean one channel is delivering 6x and another is delivering 1.2x. The average looks fine, but you're cross-subsidising a losing channel with a winning one — and you can't see it.
Revenue is not profit. A campaign might deliver strong revenue returns but attract customers who buy low-margin products, return at high rates, or never purchase again. ROAS measures revenue efficiency. It says nothing about whether you actually made money.
The shift that needs to happen at this stage is from "is our marketing working?" to a much more specific set of questions:
What does each channel actually contribute after all costs? Not revenue attributed by the platform, but genuine incremental profit after cost of goods, after shipping, after returns, after the marketing spend itself. This is the difference between knowing your ROAS and knowing your true marketing profitability.
Where is the next pound best spent? Every channel has a point of diminishing returns. Facebook might deliver a 5x return on the first £20K per month and a 1.5x return on the next £20K. Understanding where each channel sits on its efficiency curve is how you allocate budget intelligently rather than just spending more on whatever worked last month.
What's the ceiling? Some channels are inherently limited. Branded search captures existing demand it doesn't create it. A high-performing niche audience on Meta might saturate quickly. Understanding the natural limits of each channel prevents you from pouring money into diminishing returns.
Are we acquiring the right customers? Not all new customers are equal. Some buy once and disappear. Others become loyal, high LTV customers who buy repeatedly at full price. If your acquisition strategy optimises for volume rather than quality, you might be growing your customer base while shrinking your customer value.
Most brands at £10M+ are running across multiple channels like Meta, Google (Search, Shopping, Performance Max), TikTok, email, affiliates, possibly influencer and organic social. Each channel plays a different role:
Top of funnel channels (Meta prospecting, TikTok, influencer) create demand. They introduce your brand to people who weren't looking for you. These channels typically show lower platform reported ROAS because they're doing the hard work of generating awareness.
Bottom of funnel channels (branded search, retargeting, email) capture demand. They convert people who already know you. These channels typically show high ROAS, but much of that revenue would have happened without the spend.
The strategic challenge is balancing investment between demand creation and demand capture. Many brands over-invest in bottom-of-funnel because the reported numbers look better but they're essentially paying to collect revenue they would have earned anyway, while starving the top of funnel that creates future customers.
The attribution black hole. You know you're spending £X per month on marketing. You know you're generating £Y in revenue. But the connection between specific spend and specific outcomes is murky. Platform numbers don't match Google Analytics, which doesn't match your actual order data. Nobody trusts the numbers fully, so budget decisions become part data, part politics, part gut feel.
The efficiency vs. growth tension. The CFO wants higher ROAS. The marketing team wants more budget. Both are right. The resolution lies in understanding exactly where spend is efficient and where it's not, but without granular profitability data, this conversation goes in circles.
The creative bottleneck. Even when you know where to spend, scaling requires a constant pipeline of creative assets. The best media buying in the world can't overcome ad fatigue. But creative output is often constrained by resources, process, or a disconnect between the brand team and the performance team.
Dema.AI was purpose-built for this problem, connecting marketing spend to real commercial outcomes.
Multi-touch attribution that reflects reality. Rather than trusting each platform's self-reported numbers, Dema uses multi-touch attribution to distribute credit across the actual customer journey. This means you see a more honest picture of what each channel truly contributes, rather than the inflated view each platform wants you to believe. When every channel's contribution is measured consistently, budget allocation decisions stop being arguments and start being arithmetic.
Profit-based marketing measurement. This is where Dema fundamentally changes the game. Instead of measuring marketing by ROAS, which only tells you about revenue, Dema measures it by epROAS (efficiency profit ROAS) and Net Gross Profit 3. These metrics tell you what each campaign, channel, or ad actually earns after cost of goods, operational costs, returns, and the marketing spend itself. A campaign with a 3x ROAS might be losing money once you factor in a 40% return rate and thin product margins. Dema makes this visible.
Channel-level and campaign-level profitability. Dema doesn't just give you a blended view. It breaks profitability down by channel, by campaign, by funnel stage, and by market. You can see exactly which campaigns are generating real profit and which are burning cash behind a flattering ROAS number. This means you can reallocate spend with precision,scaling what works and cutting what doesn't.
Marketing Mix Modelling for budget optimisation. For brands ready to go deeper, Dema offers Marketing Mix Modelling that analyses the true incremental impact of each channel, accounting for diminishing returns, saturation effects, and the interaction between channels. This answers the hardest question in marketing: where should the next pound go?
Top-of-funnel vs. bottom-of-funnel visibility. Dema lets you analyse performance by funnel position, so you can see whether you're over-investing in demand capture at the expense of demand creation. This is the strategic view that prevents brands from optimising themselves into a corner chasing high ROAS on branded search while the prospecting that feeds the funnel slowly starves.
New customer vs. returning customer economics. Not all revenue is equal. Dema breaks down performance by customer type, so you can see the true cost of acquiring a new customer versus reactivating an existing one. This feeds directly into the growth vs. efficiency decision and helps you understand whether your marketing is building long-term customer value or just buying short-term transactions.
Marketing at £10M+ is no longer about finding one channel that works and scaling it. It's about orchestrating a portfolio of channels, each with different roles, different economics, and different ceilings.
The brands that win at this stage aren't necessarily spending more. They're spending with more clarity. They know what's truly profitable, they know where the diminishing returns begin, and they have the data to make reallocation decisions with confidence rather than gut feel.
The spend isn't usually the problem. The visibility is. Fix the visibility, and the spend starts working harder on its own.