Ecommerce & Retail · Growth problem
Paid media dependency
A business that can only grow by increasing paid media spend each quarter has not necessarily built a demand problem — but it has built a fragile one, since any pause or platform disruption exposes how much of its apparent growth was really just harvesting demand that already existed. The useful diagnostic question is what share of current spend is creating new demand versus capturing demand the business would have received anyway.
Symptoms
What this usually looks like
- Revenue growth closely tracks media spend growth, quarter after quarter
- Pausing or reducing spend in a test period causes a rapid and roughly proportional revenue drop
- Meta and Google together account for the large majority of the acquisition budget
- Organic, direct and CRM-driven revenue are flat or declining as a share of total revenue
- Marginal ROAS on the largest channels has been declining for several quarters
Diagnostic questions
What we would test first
- Plot revenue growth against total media spend growth over the last 6–8 quarters
- Calculate contribution margin after media cost by channel, not just ROAS
- Estimate channel concentration by both spend share and revenue share
- Run or review a geo holdout or spend-reduction test to estimate marginal ROAS
- Track non-paid revenue (organic, direct, CRM, referral) as an absolute figure over time
- Split paid search and social spend into prospecting vs retargeting/branded cohorts
Root causes
Why it happens
- 01
Retargeting and branded search harvesting rather than creating demand
A meaningful share of typical paid media budgets is spent capturing customers who were already going to purchase — through retargeting existing site visitors or bidding on the brand's own name — which inflates apparent paid performance without genuinely growing the customer base.
- 02
Channel concentration risk
Reliance on one or two platforms for the majority of acquisition volume leaves the business exposed to any change in that platform's algorithm, cost structure, or policy, and removes the natural diversification that comes from a broader demand mix.
- 03
Organic, brand and CRM demand not being actively built
Where all growth investment goes into paid media, the channels that could reduce dependency over time — SEO, brand awareness, referral, CRM-driven repeat purchase — receive comparatively little investment and don't develop the compounding effect they otherwise would.
- 04
No incrementality testing to establish a genuine spend ceiling
Without deliberate testing — geo holdouts, spend-level experiments — it's difficult to know whether current spend levels are still profitable at the margin or whether the business has quietly passed the point of diminishing returns some time ago.
- 05
Contribution margin after media not tracked as a primary metric
Reporting that stops at revenue or ROAS, without going on to contribution margin after media cost, can make a channel look healthy even when its true profitability has been eroding.
Evidence
The numbers we would look at
These are the metrics that make the constraint visible, and the cuts that stop them being reassuring by accident.
| Metric | What it tells you | How we read it |
|---|---|---|
| Revenue growth vs media spend growth | The clearest indicator of how tightly growth is coupled to spend.A ratio close to 1:1 over several quarters is a strong sign of dependency. | A ratio close to 1:1 over several quarters is a strong sign of dependency. |
| Contribution margin after media cost, by channel | The true profitability measure once acquisition cost is accounted for, not just ROAS.Track trend over time, not a single-period snapshot. | Track trend over time, not a single-period snapshot. |
| Channel concentration (share of spend/revenue in top 1–2 platforms) | Quantifies platform risk exposure.Consider both spend share and revenue share, which can differ. | Consider both spend share and revenue share, which can differ. |
| Non-paid revenue share (organic, direct, CRM, referral) | Indicates whether durable, non-paid demand is growing or shrinking as a proportion of the business.Absolute non-paid revenue, not just its percentage share, which can mislead if total revenue is also shifting. | Absolute non-paid revenue, not just its percentage share, which can mislead if total revenue is also shifting. |
| Marginal ROAS / incremental return on additional spend | Shows whether the next pound of spend is still worthwhile, which average ROAS cannot tell you.Requires a spend-level or geo-holdout test to estimate reliably. | Requires a spend-level or geo-holdout test to estimate reliably. |
| Branded search and retargeting share of paid spend | Identifies how much of paid performance is demand harvesting rather than demand creation.Compare CAC and ROAS for these segments against genuine prospecting spend. | Compare CAC and ROAS for these segments against genuine prospecting spend. |
Measurement traps
What can mislead you
- Looks fineOverall ROAS is healthy, so paid media is performing well
- A blended ROAS figure can be flattered substantially by retargeting and branded search performance, while genuine prospecting spend on new audiences is performing considerably worse underneath it.
- Looks fineWe've diversified across several platforms, so we're not dependent on one
- Diversification across Meta, Google, and other paid platforms doesn't reduce dependency on paid media as a category if organic, CRM and brand-driven demand remain a small share of the total — the concentration risk has just moved from one platform to the channel type as a whole.
- Looks fineSpend cuts in a downturn will simply be reversed once conditions improve
- If a spend reduction is tested and revenue falls sharply and immediately, that is itself useful evidence of dependency — but it should be read as a diagnostic finding to act on, not treated as confirmation that nothing needs to change.
Outcome
What better looks like
Not a promised number. A clearer basis for the next investment decision.
- Contribution margin after media cost, not ROAS, is the primary measure of paid channel health
- Leadership knows, from testing, roughly where marginal returns start to decline on the largest channels
- Investment in organic, CRM and brand-building demand is tracked and resourced as a deliberate counterweight to paid dependency
- Channel concentration risk is understood and factored into planning, not treated as a secondary concern
Where a Growth Diagnostic would start
A three to four week senior review across demand, discovery, acquisition, conversion, retention, measurement and capability — sequenced so this problem is either confirmed as the constraint or ruled out early. Read alongside the ecommerce & retail model page for how we frame the wider system.
Related problems
Ecommerce & Retail
Rising customer acquisition costs
The cost of winning a new customer keeps climbing, even where individual platforms report healthy performance.
Ecommerce & Retail
Marketing attribution is unclear
Every platform claims the credit and the numbers don't add up. A pragmatic approach to measurement that decisions can actually be based on.
Ecommerce & Retail
SEO traffic isn't driving revenue
Organic visibility looks strong but it isn't translating into sales. A look at intent, landing pages and product availability.
Questions about this problem
- Will testing a spend reduction hurt our revenue in the short term?
- A controlled test, such as a geo holdout or a modest, time-limited reduction, is designed to be informative without materially damaging revenue — the point is to learn the marginal effect deliberately, rather than discover it by accident.
- Is this relevant for a business that isn't struggling right now?
- Yes — it's often more valuable to establish this before a downturn or a platform disruption forces the question, since the diagnostic work is the same either way but the options for responding are wider when done proactively.
- Do we need a media mix model or econometric analysis for this?
- Not necessarily at this stage — for most businesses in this range, a well-run incrementality test and a clear contribution-margin view answer the immediate question; full MMM tends to be worthwhile once media spend is large and complex enough to justify it.