A chunk of what counts as performance would likely have happened without any campaign running, and no number exists for how much. Every recommendation signed off rests on a figure that treats demand already headed toward the business as if spend created it, and the doubt persists every time it is signed. Without a fixed reference point for nothing turned on, a good result reads as skill and a bad one reads as failure, regardless of which is true. The name on the sign-off belongs to numbers nobody there built and nobody there can check.
A baseline built from the business's own spend history, not a guess, lets credit and blame finally land on the part of the result that is actually earned. Over-crediting demand that would have arrived anyway stops, changing what future spend decisions look like. And a fixed point exists to defend, replacing a private suspicion that used to go nowhere.
A large share of demand looks like it would arrive with or without marketing, driven by urgency that formed before anyone saw an ad, and influencing it earlier in that window is close to impossible. Every cost-per-acquisition and return figure is inflated by that same demand, credited to spend that was, at most, recapturing intent that already existed. The performance history that justified past scaling decisions, and the growth owner's own record, may have been measuring demand nobody created. For a founder accountable for the number, that is the entire question of whether growth came from marketing or from the market.
The baseline read shows what would have happened without that spend, isolated from a channel nobody fully controls anyway. Marketing's own contribution comes apart from demand it was never responsible for creating. An answer to the question replaces the guessing that came before it.
The category is led by reputation, but promo-to-sales patterns are only visible for roughly half of revenue, with no visibility into what the rest would have done without marketing. Without that reference point, every sales swing is claimable by trade terms, promotions, weather, or media, and nobody can say with confidence which one caused it. Credit and blame get assigned by whoever argues loudest, not by evidence, on a scale where being wrong is expensive either way. Leading the category by reputation while running budget on that basis leaves a gap nobody wants noticed.
A number for what survives with marketing turned off closes the visibility gap that promo-to-sales tracking alone cannot close. Every swing in the numbers gets assigned to an actual cause instead of the loudest argument in the room. Measurement maturity finally matches the market position already held.
In a market with strong brand recognition and major merchant lending partnerships already in place, it is genuinely hard to say how many of the conversions campaigns claim would have happened anyway on that existing awareness. Budget keeps growing on the strength of those numbers, almost never shrinking, while doubt persists over how much of the result the spend actually caused versus conditions inherited rather than built. If the pre-existing baseline is doing most of the work, the marketing function is being credited, and funded, for outcomes it never produced. Eventually someone has to say, out loud, whether a strong period was driven by marketing or simply coincided with it.
The baseline separated from campaign-attributed conversions means growth in reported performance is not silently borrowed from brand equity built in an earlier quarter. A number exists to defend when budget for next quarter is being set. The boom-or-baseline question gets answered with evidence instead of a guess.
A marketing baseline is the revenue that arrives without paid media behind it: existing demand, brand equity and repeat behaviour, plus trend, seasonality and non-media factors like price and promotions. It is the reference point every other performance number is measured against. Across 138 advertisers covering 871 models, the median advertiser had 56% of revenue arriving as baseline, and the middle half fell between 30% and 80%. The spread is the point: too wide to borrow, which is why each business measures its own.
By reading historical variation in the business's own spend, periods where a channel ran lighter, was paused, or scaled up, and separating its effect from every other channel and from trend, seasonality and price moving in the same weeks. Where that history does not exist yet, a controlled geo holdout gives a causal read on the tested channel, and that read anchors the baseline once enough history has accumulated to model.
By modeling the portion of conversions that tracks with brand-driven demand independent of campaign activity, and treating only the remainder as what the campaigns themselves caused. In markets with strong pre-existing awareness, this step matters more, since campaign-reported numbers are more likely to be counting demand that already existed.
When the account is too new or has never varied its spend enough to produce a readable baseline, when the decision needed is at ad-set level, or when outside factors like weather or a competitor promotion need to be judged case by case rather than folded into one baseline number. In those cases the honest output is a range, not a point estimate.
Credit and blame for a given month stop defaulting to whoever is in the room and start attaching to what marketing actually caused versus what the business would have done anyway. Budget conversations shift from arguing about a single blended number to arguing about the smaller, more honest number that is actually marketing's.