For most of the last five years, running a conversion lift study in Google Ads required either a dedicated Google account rep or a minimum budget that ruled out most DTC brands. The floor was somewhere north of $100,000. Setup happened in a separate measurement product. Most accounts never touched it.
That changed in 2026. Google moved lift studies into the Experiments section of the Google Ads interface — the same area where you run A/B creative tests — and cut the minimum budget for conversion lift studies to around $5,000, per ALM Corp's reporting on the updated platform. For the first time, the infrastructure for proper causal measurement is sitting inside accounts that spend $150 to $500 a day.
This matters more heading into Q4 than it would at any other time of year.
What a lift study actually measures
Standard attribution — last-click, data-driven, or otherwise — tells you which ads got credit for conversions. A lift study tells you whether those conversions actually required the ad to happen.
Google's conversion lift study works by randomly holding out a percentage of eligible users from seeing your ads, then comparing the conversion rate of that unexposed group against the group that was exposed. The difference is your estimated incremental lift: conversions that happened because of the advertising, not conversions that were attributed to it.
These are often different numbers. For campaigns targeting users already in a purchase decision — branded searches, high-intent shopping queries — the incremental number is frequently lower than the attributed number. That gap is the whole point of the test. Knowing whether pulling back spend would actually hurt revenue, versus whether you'd keep those conversions anyway, is one of the few genuinely hard questions in paid media measurement.
Three study types, three different questions
The Lift studies tab in the Experiment Center currently offers three formats:
Brand Lift runs a survey comparing ad recall, brand consideration, and purchase intent between an exposed and a withheld audience. Most useful for YouTube and Demand Gen campaigns where the direct conversion path is long.
Search Lift shows whether your ads influence downstream search behavior — do users who saw your YouTube or Demand Gen ad search for your brand more than users who didn't? For DTC brands trying to prove upper-funnel spend creates lower-funnel demand, this is the measurement that attribution models structurally can't replicate.
Conversion Lift is the most directly actionable for performance advertisers. It measures whether your campaigns are driving purchases, leads, or other conversion events above the baseline you'd see with no ads running. At roughly $5,000 minimum (per ALM Corp, September 2026), this is now accessible to accounts that were previously locked out by budget floors.
The honest limitation
Google is measuring the performance of Google's own ads. That conflict doesn't disappear regardless of how the experiment is structured. The holdout methodology is sound — a properly randomized control group comparison is hard to game — but the choices around holdout sizing, confidence thresholds, and attribution windows are made by Google, not by you.
This matters in how you interpret results. Treat the lift estimate as a directional answer to a binary question — "is my incremental lift substantially above zero?" — rather than a precise ROAS multiplier to present in a budget review. If the study shows near-zero incremental lift on a campaign consuming significant budget, that's a real signal worth acting on. If it shows strong lift, that's useful validation but not proof.
Why running one now is worth more than in January
With forced AI Max migrations active across most Search campaigns and CPMs rising heading into October, this Q4 is not a good one to be flying entirely on attributed conversions.
A lift study started now — September 23 through late October — against a mature Search or PMax campaign gives you a causal data point before your peak-season budget scales. If incremental lift is real and substantial, you have evidence to justify scaling in October. If it's marginal, you have time to restructure campaigns or shift budget before Black Friday.
A conversion lift study started in November produces results in January. The timing window where the output is actually useful is right now.
What the new access level means practically
The Experiments section in Google Ads now shows a Lift studies tab alongside the standard A/B experiment options. For conversion lift, you select the campaign, set a holdout percentage (typically 10–15% of eligible users), define the test duration (minimum four weeks for most conversion volumes), and specify the primary conversion event you're measuring against.
For accounts spending $5,000–$20,000 per month on Search or PMax, this is now accessible without rep involvement. If you've been making budget decisions based purely on Google's reported conversion volume without any independent check on incrementality, this tool is the most direct way Google has offered to run that check — and for most of 2026, the budget floor kept the majority of advertisers out of it.
The cost floor change doesn't fix the conflict-of-interest problem in platform-owned measurement. But it removes the access barrier that forced most DTC brands to either trust Google's attribution wholesale or invest six figures in third-party incrementality vendors to get an independent read.
If you want to see how your current account's signal quality and conversion tracking setup affect the reliability of any lift study you run, the free audit gives you a baseline in three minutes.
Causal measurement is no longer reserved for accounts with a Google team on speed dial. Whether you use it before peak season is a choice you're making right now.
Sources: ALM Corp, Google Ads Help Center, Search Engine Land, September 2026.
