The virtuous circle: steer your growth by real margin, not ROAS
Never so many levers, so much data, so many tools. And yet, few leaders truly know whether their growth is profitable. Here's how to orchestrate marketing around a single arbiter: real margin.
As a brand grows, it almost always ends up with the same setup: one provider for Paid, an agency for SEO, a tool for email, an Amazon account managed on the side.
Each one optimizes its own channel. Each one has its numbers, its wins, its dashboard.
And in the middle of all this, the only question that really matters stays without a clear answer: is all of this actually making money?
This is the paradox of modern marketing. We've never had so many levers, so much data, so many tools. And yet, ask a leader whether their growth is profitable, truly profitable, and you'll often see hesitation.
This article explains why, and above all how to get out of it.
The performance of a single channel says nothing about the health of your business. The only reliable arbiter is your real margin. Everything else is just an intermediate indicator.
The three blind spots of siloed marketing
Three mechanisms explain why so many brands fly blind. None is a moral failing. They are structural blind spots, shared across the whole market, including by the best agencies.
First blind spot: the levers don't talk to each other. Paid on one side, SEO on the other, CRM somewhere else again. The synergy between them: everyone talks about it, everyone looks for it, and almost no one truly delivers it. Yet that's where most of profitable growth is won.
Second blind spot: steering by vanity metrics. A brand can post a gorgeous ROAS while selling at a loss. Another can post a modest ROAS and generate solid EBITDA thanks to its cost structure. A ROAS that rises while margin falls is one of the most common traps, and one of the most expensive.
Third blind spot: inflated attribution. Every platform claims everything. Meta says it converted X, Google says it converted Y. Add X and Y together and you often exceed actual revenue. The same sale, claimed twice. Every decision made on these raw numbers is undermined.
One important clarification, because it changes everything. A falling margin is not an alarm signal in itself. Many brands deliberately lower their margin to buy volume, and that's a perfectly legitimate choice. The real danger isn't a margin that falls. It's a margin that falls without anyone knowing, and without a plan. A brand unaware that its margin is melting is a brand in danger.
Changing your compass: real margin
Escaping the blind spot fits in one sentence: stop judging performance by a channel's revenue, and judge it by the business's real margin.
In concrete terms, two numbers frame everything, and they're within reach of any leader.
The first: current revenue against the annual target. It sets the pace. Behind, you're in catch-up mode. Ahead, you keep scaling without braking, while watching the margin.
The second: margin. It's what tells you whether the machine is healthy.
But you have to look at the right margin. Most brands stop at revenue. The most serious ones go all the way down to net margin after acquisition, what's truly left once you've paid for producing the product AND the cost of going out to win the customer. That's the margin that should guide every acquisition decision.
And that's where the figure that sums up our approach comes in: LTM, for Long Term Margin.
Why does it change the reading? Take a customer whose twelve-month value is 500 euros. If they've consumed 400 euros of product cost, their real lifetime margin isn't 500 euros, but 100. Acquired at a cost of 80 euros, they only bring you 20 euros of net margin. The classic LTV reading would have made you believe in a far more generous return. LTV makes it invisible. LTM makes it legible.
Same logic on the advertising side. ROAS measures the revenue generated per euro spent. POAS measures the profit per euro spent. When you steer by POAS, you're already speaking the language of margin, not that of gross revenue. On levers like Google Shopping or Performance Max, this simple change of compass realigns every decision around what really matters.
The end of "LTV/CAC above 3 or you die"
The market loves universal thresholds. Minimum ROAS of 2. LTV/CAC above 3. Return on investment in under 6 months. These rules are reassuring. They're also often wrong, because they ignore each brand's cost structure.
A simple example. A solo entrepreneur, with no fixed costs, working from home, can be perfectly profitable with a margin-to-acquisition-cost ratio of 1.2. The margin generated drops almost entirely to the bottom line.
The same brand, scaled to thirty people, with offices, inventory and logistics, at that same ratio of 1.2, is a disaster. The margin will never cover the fixed costs.
Conclusion: there is no universal threshold. Each brand has a target that depends on its fixed costs, its stage of maturity and its ambitions. Not a rule. A bespoke figure. And that bespoke figure can only be defined starting from the P&L, never from a generic benchmark.
The virtuous circle: how each lever feeds the others
This is where it all connects. Profitable growth isn't won within a channel. It's won in their orchestration.
We start from the P&L. It dictates which products to push. From there comes the angle, meaning the need you're answering. The angle dictates the persona, the persona feeds the creative, the creative points to a coherent landing page, and attribution measures the real impact to update the P&L. The circle closes, and it starts again, better informed with each turn.
Around this circle, the levers amplify one another. A few concrete examples.
CRM tells Paid how much it's allowed to pay for a customer. The lifetime margin observed on the CRM side, through cohorts and retention, sets the maximum acceptable acquisition cost, which you can estimate in a few seconds. If lifetime margin rises, you can acquire at a higher cost and accelerate. If it falls, you tighten. It's the most powerful synergy, and the most rarely wired up.
SEO and Paid illuminate each other. The keywords that convert in advertising are high-potential intents: you hand them to SEO so it no longer works blind. Conversely, the pages that rank and convert organically reveal the angles that resonate with the audience, and feed the ad copy. And a page that already performs organically, boosted by a small advertising budget, becomes a minimal-risk, high-leverage test.
Amazon enriches the whole chain. The volume of reviews there often exceeds that of the direct site: it's a goldmine of verbatims for hooks. The best-performing keywords there carry pure purchase intent, usable in advertising. And price consistency between Amazon and the direct site avoids cannibalizing your own channel.
The mental model to remember: an isolated lever, however excellent, plateaus. Orchestrated around margin, the same levers produce value that none of them could generate alone.
Attributing to the truth
A circle is only worth something if it closes honestly. That's attribution's role.
The problem, as we've seen, is that each platform's raw numbers contradict and inflate one another. An agency that manages only one channel has a direct interest in believing the figure that suits it. A multi-lever approach doesn't have that bias: it has no reason to attribute a sale to Meta rather than Google, it just needs to know what actually created the value.
For that, we cross-reference several readings rather than relying on a single one. The customer's spontaneous declaration at the moment of purchase ("how did you discover the brand?") is part of it. So are incrementality measurements, which isolate a lever's real effect from what would have sold anyway. The goal isn't to reassure yourself with a nice number. It's to find the truth, and to update the P&L with each cycle.
The mindset that changes everything
Everything above stems from a mindset, more than from a toolbox.
We hold ourselves accountable for the client's margin, not just their ROAS. That's what makes the difference between a provider and a business partner.
This mindset has concrete consequences. On the creative side, we keep a human core, an authentic voice, real emotion, and then amplify with AI. AI is a tremendous accelerator, as long as it stays invisible. A 100% AI-generated creative becomes generic, recognizable, and loses its ability to move people.
It also has a consequence for the method. Our job isn't to predict what will work. It's to build the testing machine that reveals it, then scale what wins.
And it demands a dose of humility. The depth of orchestration is calibrated brand by brand. Not all will grant the same access, and that's normal, especially with large accounts where certain doors stay closed. We don't claim to see everything. We look at what we can, we tie it back to margin, and we steer.
That's what it means, for us, to grow a brand without sacrificing its profitability. Not stacking channels. Making them run together, around a single arbiter.
In practice
What is LTM (Long Term Margin) in e-commerce?
LTM (Long Term Margin) is the real margin generated by a customer over their entire lifetime: it's calculated by subtracting the cost of goods sold (COGS) from gross LTV, i.e. LTM = LTV − total cost of goods purchased. Unlike classic LTV, which thinks in revenue terms, LTM exposes real net profitability, accounting for variable margins by category, returns and discounts. It's the reference indicator for comparing the real value of two customer segments with different baskets and product mixes.
ROAS or POAS: what's the difference and which should you choose to steer your advertising campaigns?
ROAS (Return On Ad Spend) measures the revenue generated per euro invested in advertising, while POAS (Profit On Ad Spend) measures the gross margin generated by that same euro, i.e. POAS = gross margin generated / ad spend. Steering by ROAS amounts to optimizing sales volume without accounting for margins, which can make profitable campaigns that actually destroy value on low-margin products. POAS aligns the advertising algorithm with real profitability and avoids over-investing in SKUs with a high ROAS but near-zero margin.
Why is the LTV/CAC above 3 ratio a myth for e-commerce merchants?
The LTV/CAC > 3 dogma is a rule of thumb from 2010s SaaS, applied by analogy to e-commerce without structural justification: it ignores the real margin of products, the speed of CAC recovery and the channel mix. An e-commerce merchant with a 20% gross margin and an LTV/CAC of 4 can be less profitable than a competitor at an LTV/CAC of 2.5 operating at 55% margin. The right indicator isn't a universal ratio, but the payback period in real margin (LTM) and the incremental profitability threshold per acquisition channel.
How do you measure the incrementality of advertising campaigns in e-commerce?
Advertising incrementality measures the sales volume actually caused by a campaign, as opposed to sales that would have happened without it (organic or brand sales). It's measured via geo-lift tests (comparing exposed zones vs control zones) or conversion lift experiments on Meta and Google. Without an incrementality measurement, classic attribution models (last-click, data-driven) artificially inflate the ROAS of retargeting and brand campaigns, leading to overspending on conversions already won.
What is multi-lever orchestration in e-commerce and why does it replace channel-by-channel steering?
Multi-lever orchestration means steering paid media, CRM, SEO, pricing and merchandising simultaneously toward a shared real-margin objective (POAS or LTM), rather than optimizing each channel on its own KPIs in a silo. A channel may show a weak POAS in isolation yet trigger high-LTM repeat purchases via email: evaluating it alone leads to under-funding it. Orchestration allocates budget where the marginal contribution to total margin is highest, accounting for the interactions and halo effects between levers.

Sophian Petitprez