Meta Drives 2.3x Search's Incremental ROAS. 64% Are New Customers.
Founder, BTB Audits. $150M+ in ad spend managed across Meta and Google
The Measured data slots into a broader picture from the May 2026 summit. Meta's own Kantar research found Meta led every phase of the shopping journey (discovery, evaluation, purchase, and connection) for US consumers in 2025, while Haus's 640-experiment analysis found Meta's own click attribution under-reports its impact by 15 percent. Three independent data points, three different methodologies, all pointing in the same direction: Meta's actual contribution to D2C revenue is substantially larger than the dashboard shows.
What happened
What most operators will get wrong
The popular take on this data is: "Meta has 2.3x search's incremental ROAS AND delivers 64 percent new customers. Cut Google, scale Meta." This is the reaction Meta would obviously like operators to have, and it is the reaction many will have.
It is wrong.
The data does not say Meta is better than search. It says they do different jobs. Search captures existing demand. When someone Googles "best running shoes for flat feet," they have already decided to buy running shoes. The job of the search ad is to be there at that moment with the right offer. The conversion is high-intent, the funnel is short, and the customer was going to buy something in the category anyway. Search ROAS looks high because the search budget is harvesting demand that exists.
Meta does not capture demand. Meta creates demand. When someone scrolls past a Reel and sees a product they did not know they needed, that is a new customer journey starting. The conversion is lower-intent, the funnel is longer, and the customer was not going to buy in the category that day. Meta ROAS on first-touch can look weaker than search because it is doing harder work: creating purchase intent from nothing.
The incrementality data captures this difference. When Measured tests a holdout (some users see search, some do not), the search holdout group still converts at a high rate because they had demand and would have found another path to the brand (organic search, direct, another channel). So search's incremental ROAS is lower than reported ROAS. When Measured tests a Meta holdout, the holdout group does not convert because Meta was the source of the demand in the first place. Meta's incremental ROAS is higher than reported.
The right read of the Measured data is not "Meta wins." It is "your channels do different jobs, and the dashboard ROAS metric mostly measures one of them." A balanced D2C ad budget runs both, with the ratio set by where the brand is on the demand curve. A brand with strong existing search volume should weight search more. A brand still building awareness should weight Meta more.
The operators who lose on this update are the ones who treat it as an excuse to make a unilateral budget decision (usually away from Google, since Meta is the data's apparent winner). The operators who win are the ones who use it to start the right conversation: "What do our channels actually do, and what is the right ratio for our growth stage?"
What you should actually do
Run this 3-step check on your account this week. The numbers from Meta Ads Manager and Google Ads, combined with new-customer data from your back-end, will tell you whether the Measured data validates your current allocation or argues for reweighting.
The full diagnostic for what to do when Meta is over-spending without delivering new customers lives in the 10-stage Meta ad audit method, especially Stages 5 and 6 (audience health and budget allocation).
How this changes the audit method
The Meta audit method does not gain a new stage from this data. What changes is the benchmark at two existing stages.
Stage 6 (budget allocation) has always asked "where is your spend, and is it on the campaigns that drive value?" The Measured 64 percent number is now the new-customer benchmark for that stage. A healthy D2C account should see Meta delivering roughly 60 to 70 percent of net-new customers, with search and other channels splitting the rest. If your account is meaningfully below that, the audit asks whether the gap is real (search-heavy brand, unique vertical) or a measurement artifact (CAPI gaps, attribution under-counting).
Stage 8 (measurement reconciliation) gains a comparison benchmark. The 2.3x incremental ROAS ratio between Meta and search is a directional baseline. The audit can now compare what the brand's incrementality testing (if any) shows against the industry baseline, and flag gaps in either direction. A brand showing 4x Meta-to-search incremental ROAS likely has a measurement gap on the search side. A brand showing 1.1x likely has a measurement gap on the Meta side.
These are the only changes to the Meta ad audit method. The 10 stages stay in order. The benchmarks at Stage 6 and Stage 8 update with the Measured numbers as anchors.
Channel thinking: before and after
| Aspect | Before | After |
|---|---|---|
| Default CFO objection | "Search converts at 5x ROAS. Meta converts at 3x ROAS. Move budget to search." | "Search reports 5x, Meta reports 3x. On an incremental basis, Meta is 2.3x search. The reported ROAS comparison is the wrong question." |
| What the channels do | Both "drive conversions." Channel choice is a question of which delivers more conversions per dollar. | Search captures existing demand. Meta creates new demand. They do different jobs. The right question is the ratio, not which one is "better." |
| How to read first-touch ROAS | Higher is better. Channels with low first-touch ROAS get cut. | First-touch ROAS is heavily biased toward demand-harvesting channels (search, brand bidding). Channels doing demand-creation work (Meta, TikTok, display) look worse on first-touch but drive the 60 to 70 percent of new customers that compound. |
| What the audit checks at Stage 6 | Whether top-spend campaigns are also top-ROAS campaigns within each channel. | Same plus: cross-channel new-customer share. Is Meta delivering roughly 60 to 70 percent of new customers? If not, why? |
| Risk of misreading the data | Lower. Operators knew "attribution is hard" but had no specific cross-channel benchmark. | Higher if read as "cut search, scale Meta." The Measured data argues for the right ratio, not for unilateral reallocation in either direction. |
Frequently asked questions
Common questions
About the data
Who is Measured and how did they run the study?
Measured is an independent marketing measurement company used by major DTC and e-commerce brands. They run holdout-based incrementality experiments at the campaign and channel level rather than relying on attribution models. The 10,000-plus campaign analysis cited at the May 2026 Performance Marketing Summit reflects experiments run with their client base during 2025, spanning multiple verticals (apparel, beauty, food and beverage, household goods, others) and brand sizes ($1M to $500M annual ad spend).
Why is incremental ROAS different from reported ROAS?
Reported ROAS is total revenue attributed to a channel divided by spend on that channel. It includes conversions the channel actually caused AND conversions that would have happened anyway. Incremental ROAS measures only the conversions the channel CAUSED, using a holdout group (users who do not see the channel) as the comparison. For demand-harvesting channels like search, reported ROAS is much higher than incremental ROAS because search captures conversions that were going to happen. For demand-creating channels like Meta, the two numbers are closer because Meta is often the source of the demand.
Does the 64% new-customer finding apply to all categories?
Directionally yes, but the exact number varies by category. Categories with longer purchase cycles and lower repeat rates (furniture, mattresses, custom goods) tend to see even higher new-customer percentages on Meta because there are fewer existing customers to re-serve. Categories with strong subscription or repeat dynamics (beauty, supplements, pet food) may see lower new-customer percentages because Meta is also being used for cross-sell to existing customers. The 64 percent is the dataset-wide average and a good baseline for most D2C accounts to compare against.
What to do next
Should I cut my Google Ads budget?
Almost certainly not. Search captures demand the brand is already creating through other channels (organic, social, word of mouth, Meta). Cutting search budget means leaving high-intent buyers on the table. The right move based on the Measured data is to check whether your Meta-to-search spend ratio matches your new-customer-to-existing-customer revenue ratio. If it does, no change. If Meta is under-allocated relative to its new-customer contribution, shift incrementally (10 percent moves, not 50 percent moves), measure, decide.
What if I only run Meta or only run Google?
The Measured data is the case for diversifying. A Meta-only brand is leaving the bottom-funnel conversion capture from search on the table. A Google-only brand is missing the new-customer acquisition pipeline entirely. Most D2C brands benefit from running both, with the ratio set by growth stage. Brands earlier in their growth weight Meta more (because demand creation is the bottleneck). Brands at scale with strong search volume weight Google more (because demand capture is the bottleneck).
How does this interact with the Pixel auto-enrichment and one-click CAPI announcements?
The Measured data assumes Meta has reasonable signal coverage. Brands with broken Pixel events or no CAPI integration will see Meta under-performing the 2.3x and 64 percent benchmarks because the algorithm is making decisions on incomplete data. Fix the signal layer first (per the Pixel auto-enrichment post and the one-click CAPI post), then measure the channel ratio. Without clean signal, the Measured benchmark numbers are aspirational, not achievable.
The Measured data is the strongest third-party validation yet that Meta does the new-customer acquisition work most D2C operators have been quietly relying on it for. The 3-step check above tells you whether your account's allocation is in healthy shape or whether the data argues for reweighting. Either way, do not let this number become the excuse for a unilateral budget move.
If you don't have four to six hours, or you want a second pair of eyes that's managed $150M+ across Meta and Google, the Free Quick Scan is what I built for that. I'll record a private 5 to 7 minute Loom walking through the leaks I find on your account using public data only. You'll have it in 48 hours.
Get Your Free Quick Scan →Keep reading on Meta budget and measurement
Aditya Chaturvedi is the founder of BTB Audits. He has managed $150M+ in ad spend across Meta and Google for DTC, SaaS, and lead-gen brands ranging from $10K per month to $500K per month. Industry Updates from BTB Audits cover platform changes and what they actually mean for operators, not what the headlines say they mean. Read more on the BTB Audits blog.