Why Incrementality Testing Alone Won’t Fix Your Paid Media Budget – The Missing Metric via @sejournal, @tonyadam

Incrementality testing has become the default answer to a problem most direct-to-consumer brands genuinely have. Platform attribution disagrees with itself; Meta and Google routinely both claim credit for the same conversion. Not to mention studies we have done reviewing one transaction at a time to find out organic search or Google Shopping transactions were being attributed to direct.

Somewhere in that noise sits the question of how to actually allocate paid media budget.

The standard pitch is that incrementality cuts through it. Run a lift study, find out which channels are creating demand versus harvesting it, and reallocate spend accordingly. Most of the content you’ll find on incrementality over the last couple of years lands somewhere in that neighborhood. That framing is incomplete, and acting on it has probably led some growth-stage brands into bad decisions. The most common one is cutting upper-funnel channels that fail standalone lift tests, only to watch total revenue drop because those channels were doing work no single-channel test could see.

The conversation needs a different anchor.

Why Incrementality Alone Doesn’t Answer The Allocation Question

Incrementality measures the causal impact of a specific channel or campaign. That is genuinely useful information, but it is not the same as understanding how marketing contributes to the business as a whole.

Consider a customer who sees a Meta ad on Monday, doesn’t click, then searches for the brand on Wednesday and converts through a paid brand search ad. Meta records a view-through. Google records a last-click conversion. A lift study on either channel in isolation might show a modest incremental contribution. The honest answer is that both ads did real work, just different work. The Meta impression created the brand consideration, while the branded search closed the demand. Cutting either one breaks the journey.

This is exactly the conclusion most brands reach when they read incrementality results without the right context. They see Meta’s lift study come in low, conclude the channel is taking credit for conversions that would have happened anyway, and reallocate the budget. Six weeks later, brand search volume drops, blended efficiency drops with it, and the team is trying to figure out what happened.

One lift study on one channel cannot tell you whether that channel deserves the budget; it can only tell you what happened inside the test, which is why allocation decisions need a metric that captures the whole business.

Marketing Efficiency Ratio (MER) Is The Metric The Conversation Is Missing

Marketing Efficiency Ratio, total revenue divided by total ad spend, is the only commonly available metric that doesn’t care which channel gets credit. It treats marketing as one investment producing one revenue stream. That is what marketing actually is at the business level, and that is the question chief financial officers and founders are actually asking when they look at performance.

MER on its own is not enough. It can’t tell you how to allocate within a budget, and it can be inflated by seasonality or organic demand growth. But it answers the question that should anchor every other measurement decision: Is the blended marketing investment producing acceptable returns at the business level? Once that anchor exists, the role of every other layer becomes clearer.

The Three-Layer Stack That Actually Works

A solid measurement stack has three layers, each answering a different question.

The mistake brands make is using any one layer to answer questions that require the others. Cutting Meta because brand search closed the sale reads attribution as causation. Trusting Meta’s reported return on ad spend does the same thing in reverse. Treating an isolated lift study as a verdict on whether a channel deserves spend ignores what that channel might be contributing to MER through its effect on other channels.

How To Actually Run Incrementality Inside This Stack

Incrementality testing was not as simple as it is now, and in some cases, the price tag was much higher than they would want to invest. The good news is that the cost of running incrementality tests has dropped meaningfully in 2025. Four testing methods, ranked by accessibility:

Platform-Native Lift Studies

Meta Conversion Lift and Google Conversion Lift run inside the existing ad platforms at no additional cost. Per Google’s official Conversion Lift documentation, the platform now reports directional lift results for studies with budgets above $5,000 USD and 1,000 conversions, supported by a transition to Bayesian statistical methodology that allows studies to run with lower budgets and fewer conversions than the older frequentist approach required. Google Ads Highlights of 2025 confirms Conversion Lift now works at lower spend levels and conversion volume than in prior years.

Meta’s Brand Lift studies sit at the other end of the spend spectrum. Per Meta’s minimum requirements documentation, Brand Lift in the United States requires a $120,000 minimum budget over the study duration. This is up from $30,000, which is a significant increase and puts Brand Lift out of reach for many brands. That said, Meta’s Conversion Lift studies have lower thresholds and remain a viable starting point. The two products measure different things and carry very different costs, which is worth understanding before designing a testing program.

Platform-native tests have a clear limit. They only measure incrementality inside the platform running the test, so they cannot account for cross-channel effects. Read the results as one input, not the verdict.

Geo Holdout Testing

If your sales are spread across enough markets to run a real holdout, geo testing produces cleaner results than user-level lift studies. Pause spend in matched markets while continuing it in others, then measure the revenue gap. Test and control markets need to be matched on baseline performance, seasonality patterns, and customer demographics, with several weeks of pre-test baseline data to confirm the markets behave similarly under normal conditions.

Spend-Down Testing

This is the most direct way to measure MER sensitivity. Cut a channel’s budget by 50 to 75% for a defined window and measure total business impact, not channel-level metrics. If you cut Meta by half and total revenue drops 40%, that channel is contributing more than its lift study suggested. If you cut it by half and revenue holds, the channel was likely harvesting demand that other channels were creating. Spend-down testing produces less statistically rigorous results than a properly structured geo holdout, but it is the only test that explicitly measures the channel’s contribution to MER rather than to its own attributed revenue.

Full Causal Inference Models

Synthetic controls, difference-in-differences analysis, and test-calibrated media mix modeling sit at the top of the methodology stack. Google’s open-source Meridian MMM, released in 2025, brought Bayesian causal inference modeling to advertisers without requiring proprietary vendor relationships, but the methodology still requires meaningful data science capability to implement well. Most brands do not need to operate at this layer to make defensible allocation decisions. The first three methods will answer the budget questions that matter day-to-day.

A Testing Cadence That Builds Real Signal

A practical cadence for a brand spending $100,000 to $1 million monthly across paid channels:

Brands that build a quarterly testing rhythm develop a defensible view of channel sensitivity that no platform dashboard can give them, and pairing that with a steady MER read sharpens every allocation conversation.

Reading Incrementality Results Without Overcorrecting

The hardest part of incrementality testing is interpreting results in context. A low lift study on Meta does not mean Meta should be cut. It means the channel is not creating standalone incremental volume during the test window, which is different from whether the channel is moving MER through its effect on brand search, direct traffic, or returning customers.

Read the lift study as one signal alongside MER. If lift comes in low and MER holds steady when you reduce spend, the channel may be replaceable. If lift comes in low but MER drops, the channel is doing work the test could not measure.

The Stack Most Brands Are Almost Building

Most brands at growth stage have all three layers available to them and are not using them as a stack. They are looking at platform ROAS, occasionally checking a lift study, and treating MER as a number that lives in a finance report rather than a measurement decision.

The incrementality conversation has spent two years arguing about whether attribution is broken. It is not the right argument. Attribution describes the journey. Incrementality measures sensitivity. MER is the metric the business runs on. The brands building all three into a single decision-making system will allocate paid media budget more confidently than the ones still arguing about which platform’s number to trust. If you are not anchoring on MER and using incrementality as the diagnostic that explains its movement, that is the gap to close first.

More Resources:

Featured Image: Igor Link/Shutterstock