The same number can look like a success in one room and a problem worth investigating in another. The number has not changed. The business question around it has.
That is one of the tensions behind ecommerce channel strategy: media performance can look strong while the broader economics of the channel tell a more complicated story.
At the agency, a large share of store GMV coming from media was a strong signal of impact. The focus was straightforward: make the media budget work harder, improve ROAS, allocate investment across channels and ad formats, find the right audiences, and balance traffic with sales.
On the client side, the question became broader.
What does that media performance actually mean for the business?
The number sits inside a bigger system
An e-commerce channel does not operate on media alone.
There is commercial investment — discounts, vouchers, free shipping and other platform activities. There are platform and eDistributor commissions. There are operating costs such as media, livestreaming, content production, KOL bookings and agency management fees.
All of these ultimately sit within the channel P&L, by brand and by platform. That changed the context in which media performance was evaluated.
Not because ROAS became less important.
ROAS still answers a critical question: is the media investment efficient?
Incrementality answers another: is that investment actually creating additional business? The distinction becomes important when attributed performance and incremental growth tell different stories.
And P&L answers a broader one: after considering the wider economics of the channel, does the business make sense?
These metrics are not substitutes.
A brand can invest heavily in media, drive significant GMV and show strong contribution — while still being inefficient because too much media investment was required to generate those sales.
So the shift was not from ROAS to contribution. It was from treating ROAS as the answer to understanding it as one input into a broader business decision.
The same number can tell a different story
At the agency, seeing media generate 70–80% of a store’s GMV — sometimes close to 100% — was a strong signal that the media strategy was working.
That work created another kind of value too: a framework that could continue improving without depending on the person who built it. Years later, media efficiency remained a strong performance signal — consistently above market benchmarks and, at times, nearly twice the benchmark.
That matters for a different reason. A framework creates more value when it can keep working after its creator leaves.
But from the client side, the same 70–100% media contribution naturally raised more questions.
How much investment was required to create it?
How much of the sales was incremental?
What happened to margin after commercial investment and platform costs?
And ultimately:
Is the business actually healthy, or does media simply make the channel look healthier than it is?
The number is not wrong. The question changed.
Sometimes media is not the problem
Media performance can be weaker than expected for reasons that have little to do with media optimization itself.
Limited brand awareness can restrict demand. A small assortment can limit visibility and consumer choice. Weak organic traffic can leave paid media carrying too much of the demand-generation burden.
And when investment is constrained by the size of the business, there is naturally less room to test and learn.
In that situation, asking media to simply perform better would not solve the underlying constraints.
Sometimes media is not the problem. It is simply where the problem becomes visible. The constraint may sit in demand, assortment, commercial strategy, investment level, store architecture — or several of them at once.
Growth creates another layer of trade-off
The tension becomes harder when growth itself is non-negotiable.
The business had already delivered roughly 100% growth for three consecutive years, with the expectation to sustain that trajectory into another year.
At the same time, growth still had to make economic sense.
There were situations where one platform could look healthier from a P&L perspective, while senior leadership still expected acceleration somewhere else for strategic reasons.
That is not necessarily a contradiction.
It is a trade-off.
The answer is rarely to choose the platform with the best economics or the platform with the highest growth potential in isolation.
It is to understand what each choice gives the business, what it costs, and which trade-off the business is willing to make. That is where business ownership starts to feel different from channel optimization.
The question becomes:
How much should this channel contribute, what should we invest behind it, and what trade-off are we willing to make to achieve the broader business ambition?
What changed after moving client-side
Moving client-side did not make media optimization less important. It changed where optimization sits within the decision.
Media efficiency still matters.
So do incrementality, commercial investment, margin, contribution, channel mix and broader growth ambition.
The difference is that none of these numbers can be interpreted in isolation.
A strong ROAS can be a genuine efficiency win and still be insufficient evidence of incremental growth. A high contribution can look attractive while being heavily supported by media investment. And a channel with weaker economics today may still deserve investment if there is a strategic reason to accelerate it.
The job is therefore not to find the single number that looks best. It is to understand what the number is telling you — and what it is not.
Because the same number can be a success at the channel level and a question at the business level.
The number did not change. The context did.
