Your tracking is working. Your attribution data looks great. Your ROAS is 4x. And yet, when you increase the budget, revenue doesn’t move. Attribution tells you who got credit. Incrementality tells you who deserved it. Most WooCommerce store owners have never run that second test — and the tools to do it are free.

The Uncomfortable Truth About ROAS

A purchase you would have gotten anyway costs you the same as one your ad genuinely drove. Attribution models don’t make that distinction. They assign credit based on exposure, windows, and statistical modelling — not on whether the ad actually changed a customer’s decision.

68% of multi-touch attribution models over-credited digital ad channels in 2025, with poor conversion data quality as a primary driver (MarTech Series, 2025). That’s not a minority problem. It’s the industry norm. And WooCommerce stores are not exempt.

The specific pattern looks like this: your ROAS target is consistently hit, but total revenue growth doesn’t match budget increases. You scale from £5K to £10K monthly ad spend and revenue goes up £3K. The platform still shows 4x ROAS. Something isn’t adding up — and it’s your attribution model.

What Incrementality Actually Means

Incrementality is the additional conversions that occur specifically because of ad exposure — purchases that would not have happened without seeing the ad. It’s the only measurement that answers the question every store owner actually needs answered: did this ad spend generate growth, or did it just claim credit for growth that was already coming?

The contrast with attributed conversions matters: attribution tells you which touchpoints a converted customer encountered. Incrementality tells you whether those touchpoints changed the outcome. A customer who was going to buy anyway will appear in your attribution report — and inflate your ROAS — whether your ad influenced them or not.

You may be interested in: Every Ad Platform Is Claiming the Same Sale: How to Stop Duplicate Attribution Inflating Your WooCommerce ROAS

The Scale of the Gap

When brands run conversion lift tests against their platform-reported attribution, the results are consistent and uncomfortable. On average, only 40-60% of attributed conversions are truly incremental — the rest would have occurred without the ad (Meta Conversion Lift Study methodology, 2025).

For WooCommerce stores, this translates directly to reported ROAS. Platform-reported ROAS averages 3-4x; independent lift measurement typically reveals a true incremental ROAS of 1.5-2.5x (industry incrementality benchmarks, 2025). The difference between a 4x ROAS decision and a 1.8x ROAS decision is usually a budget increase that doesn’t deliver.

Here’s the maths getting worse: when Google Ads, Meta, and a third platform like TikTok all run simultaneously, total claimed conversions regularly exceed actual WooCommerce orders by 200-300%. Every platform’s attribution model fires. Every platform claims credit. Your order count stays at 100 while your dashboards collectively report 300 conversions.

In that environment, a 4x blended ROAS might be a 1.2x incremental ROAS. And you’d have no way of knowing from dashboards alone.

How to Run an Incrementality Test Without an Agency Budget

Both Meta and Google offer lift measurement tools that cost nothing beyond your existing ad spend. Most WooCommerce store owners don’t know they exist.

Meta Conversion Lift Studies

Meta’s Conversion Lift tool is available in Ads Manager at no extra cost. It works by splitting your audience into two randomized groups: an exposed group that sees your ads, and a holdout group that does not. At the end of the test, Meta compares WooCommerce conversion rates between groups. The difference is your true incremental lift.

A holdout test is the core methodology here: a randomized control group is withheld from ad exposure, and conversions in the holdout group are compared to the exposed group to measure true lift (Google Ads Conversion Lift documentation). The holdout group is your baseline for “what would have happened anyway.” Without a holdout, you’re attributing everything to the ad — including the organic baseline.

Setup requires a minimum spend threshold (typically £1,000+ over the test period) and at least 15-20 WooCommerce conversions per week to reach statistical significance. It runs for 2-4 weeks minimum.

You may be interested in: WooCommerce Attribution Windows Explained: Why the Same Sale Gets Counted 3 Different Ways

Google Ads Conversion Lift

Google Ads offers Conversion Lift studies for Search and YouTube campaigns. The methodology is similar — a holdout group is unexposed to your ads, and Google compares conversion behaviour between groups. Unlike Meta’s tool, Google’s version runs at the campaign or ad group level, making it easier to test specific channel contributions.

For WooCommerce stores running Smart Bidding, this test is especially revealing. If your Target ROAS campaign shows 4x attributed ROAS but incremental lift is 1.8x, the bidding algorithm is optimising for the wrong signal — and every decision it makes is based on inflated data.

The Minimum You Need to Start

You don’t need a media mix model or a data science team. You need three things:

• Reliable conversion data: Your WooCommerce order count must be accurate and complete — not sampled, not blocked by ad blockers. This is your ground truth for the test.
• Sufficient volume: 50+ WooCommerce conversions per week minimum. Below that, statistical noise drowns the signal.
• Clean event data in BigQuery: If you plan to analyse results by cohort, customer segment, or channel combination, you need your event data in one place — not scattered across platform exports.

What Clean Data Makes Possible

The prerequisite for any credible incrementality analysis is complete, accurate conversion data. This is where tracking infrastructure stops being a technical detail and becomes a business decision.

If 20-35% of your WooCommerce conversions are missing from GA4 because of cookie rejection, your holdout group comparison is starting from an incomplete baseline. Your lift estimate will be off. You may conclude an ad is incremental when you’re actually measuring a data gap.

Server-side tracking solves this. When events fire from your own infrastructure directly to the analytics endpoint — bypassing browser restrictions and ad blockers — your conversion baseline is complete. The holdout comparison becomes reliable.

The Transmute Engine™ captures all WooCommerce conversion events server-side via the inPIPE™ plugin and routes them simultaneously to Google Ads, Meta CAPI, GA4, and BigQuery. Your BigQuery table becomes the complete first-party dataset that makes lift analysis viable — not a platform export that’s 20% smaller than it should be.

You may be interested in: Why GA4 Data-Driven Attribution Silently Fails Small WooCommerce Stores

The Question Worth Asking Every Quarter

Not “what’s our ROAS?” The question is: “what would happen to our WooCommerce revenue if we turned these ads off for two weeks?” If the honest answer is “probably not much would change,” that’s an incrementality problem — and no attribution report will surface it.

Incrementality testing doesn’t replace attribution. It calibrates it. You still use Google Ads and Meta dashboards to manage bids and creative. But you use lift measurement to decide how much budget actually deserves to go to each channel. Those are different decisions — and they need different data.

Key Takeaways

• Attribution ≠ incrementality: Attribution assigns credit based on touchpoints. Incrementality measures whether an ad actually changed a purchase decision.
• The gap is large: Only 40-60% of attributed conversions are typically incremental. Platform-reported 4x ROAS often hides a true incremental ROAS of 1.5-2.5x.
• Free tools exist: Meta Conversion Lift and Google Ads Conversion Lift are available to WooCommerce stores at no extra cost — most don’t know to use them.
• Complete data is the prerequisite: You need a reliable, server-side conversion baseline to run a credible holdout test. Cookie-based tracking with a 20-35% data gap produces unreliable lift estimates.
• BigQuery is your analysis layer: First-party event data in BigQuery lets you run incrementality analysis by channel, cohort, and campaign — without depending on platform exports.

Perfect tracking got you this far. Incrementality testing takes you further — to a place where your ad spend decisions are based on what’s actually driving growth, not what your dashboards are optimised to claim.