TL;DR
- John Chen, Google's senior director of product management for ads measurement, called the conversion lookback window arbitrary and said the difference between day 29 and day 31 is negligible.
- Chen named an attributed branded search, a site visit and an add-to-cart as the actions that qualify a click inside the seven-day window.
- Qualified Future Conversions was first introduced at Google Marketing Live on 20 May 2026 and detailed on the Ads Decoded episode of 2 September 2026, which runs 29 minutes 44 seconds with comments disabled.
- The episode places attribution on a daily update cycle while incrementality holdback studies and marketing mix models run quarterly or annually, and Chen declined to say which measure wins when they disagree.
- Google gave no rollout date, no country list, no accuracy figure and no statement on whether these volumes feed Smart Bidding.
Google has set out how Qualified Future Conversions works: a user who clicks an ad and then takes a further action within seven days unlocks conversion counting for the next 30, 90 or 180 days. The mechanism was described on the Ads Decoded episode "Build a measurement stack you can rely on to steer your campaigns", published by the Google Ads and Commerce Blog on 2 September 2026.
John Chen called the conversion lookback window arbitrary on 2 September 2026
On that episode, John Chen, senior director of product management for ads measurement at Google, said the conversion lookback window is arbitrary, and that the practical difference between a conversion recorded on day 29 and one recorded on day 31 is negligible: "we think that's probably a little bit too arbitrary." He appeared with Ginny Marvin, Google's ads product liaison. The episode runs 29 minutes 44 seconds and has comments disabled.
The quotation matters because it is a Google product lead saying out loud that a number every advertiser reports against is a convention rather than a measurement. Whatever you think of the replacement, that sentence describes the problem Qualified Future Conversions is built to answer. You can read the episode write-up on the Google Ads and Commerce Blog. The trade outlet PPC Land published the episode quotes on 5 September 2026.
How Qualified Future Conversions works, in two stages
Chen described a two-stage mechanism on the 2 September episode. Stage one is a seven-day qualifying period: within seven days of the ad click, the user has to take a further action that indicates the ad influenced them. Chen named an attributed branded search, a site visit and an add-to-cart as those actions. Stage two is the counting period: once the qualifying action happens, the system counts conversions occurring within the next 30, 90 or 180 days.
His worked example was a Nike ad followed within a week by a search on google.com for Nike or for Jordan 1s. The click alone does nothing. The branded search inside seven days is what licenses the longer horizon.
Everything Google put on the record about the mechanism fits in one table.
| Element | What Google stated |
|---|---|
| Qualifying period after the ad click | Seven days |
| Qualifying actions named by Chen | An attributed branded search, a site visit, an add-to-cart |
| Counting windows once qualified | 30, 90 or 180 days |
| First introduced | Google Marketing Live, 20 May 2026 |
| Mechanism detailed | Ads Decoded episode, 2 September 2026 |
Qualified Future Conversions was first introduced at Google Marketing Live on 20 May 2026
Google first introduced Qualified Future Conversions at Google Marketing Live on 20 May 2026, and the 2 September episode is where the mechanism was explained rather than announced. That gap is worth holding in mind when you read anything describing the feature as new this month. The idea is three and a half months old in public; the seven-day-then-180-day structure is what became public in September.
Google's own announcement page describes it as connecting early intent to future outcomes
Google's Qualified Future Conversions announcement page describes the feature as connecting early signals of intent such as branded searches to future financial outcomes, and says the system models the probability of a future purchase based on today's user actions. The page carries a "new for 2026" label. It attributes the prediction to Google's AI without naming a specific model on the page itself, and it states no time windows.
The episode frames attribution, incrementality and MMM as three gauges on different clocks
The 2 September episode positions three measurement approaches as instruments that refresh at different speeds. Attribution updates daily. Incrementality holdback studies land quarterly or annually. Marketing mix models run quarterly or annually at an aggregate level. The blog page for the episode puts the tension plainly, describing how tough it is to evaluate a media strategy when real-time attribution, periodic incrementality tests and media mix models conflict.
| Measurement approach | Cadence described in the episode |
|---|---|
| Attribution | Updates daily |
| Incrementality holdback studies | Quarterly or annually |
| Marketing mix models | Quarterly or annually, at aggregate level |
Reading the two tables together is the useful exercise. A 180-day counting window pushes attribution, the daily instrument, into a time horizon that previously belonged to the quarterly ones. A conversion attached to a click from five months ago is not the same kind of number as a conversion attached to a click from Tuesday, even when both appear in the same column of the same report.
Chen declined to say which measure becomes the north star when the three disagree
Marvin asked Chen directly which of the three measures should win when they disagree, and Chen declined to answer. That is the honest state of the guidance as of the 2 September episode: Google published a mechanism for counting conversions further out without publishing a rule for reconciling it against the slower instruments that were already there.
Chen also named Macy's and Amazon as examples of off-site transaction surfaces, the case where the sale completes somewhere the advertiser does not own. A brand selling through a marketplace has never had a clean click-to-purchase path to measure, which is the situation a predicted-future-conversion metric is aimed at.
What would change in your reporting, and what would not
This section is analysis rather than reporting, because Google published no rollout information. On the mechanism as described, a 180-day counting window can raise reported conversion volume without a single change to a campaign, a bid or a landing page. Upper-funnel activity gains the most, because that is where the gap between the click and the sale is widest and where a 30-day window truncated the most outcomes.
What would not change is the money. Bank revenue is bank revenue, and a metric that counts more of it inside a report does not create any. The risk in a longer window is not that the number is wrong but that two numbers built on different windows get compared as though they were the same measurement, which is a reporting discipline problem rather than a product problem. Anyone whose analytics and conversion measurement feeds a monthly client report has an interest in labelling which window produced which figure before the windows change underneath the report.
The other thing worth watching is bidding. Google did not say whether Qualified Future Conversions volumes feed Smart Bidding or sit in reporting only, and those are very different products. A metric that only reports gives you a better view of the same auctions. A metric that feeds bidding changes which auctions you enter, which is why the question is the first one to ask when the feature reaches an account you run Google Ads campaigns in.
What the source did not say
Google gave no rollout date and no country list for Qualified Future Conversions on either the 2 September episode or the announcement page. There is no statement on whether QFC volumes feed Smart Bidding or stay in reporting. There is no accuracy figure and no error band, so the reliability of the prediction is undescribed.
Google also did not explain how its model decides that a branded search was attributed to the ad in the first place, which is the step the entire mechanism rests on. A search for Nike a week after seeing a Nike ad may or may not have been caused by the ad, and the qualifying test is doing exactly that causal work without a published method.
The last gap is the one that lands on advertisers rather than engineers: Google published no guidance on comparing a 180-day QFC figure against a 30-day historical baseline. Every advertiser who adopts the metric has that reconciliation problem, and the source offers no method for it.
What this means for marketers in Thailand
Google named no countries, so this section is reasoning about implications rather than anything the source said about Thailand. Long-consideration categories carry a lot of Thai lead generation: condominiums, international schools, cosmetic clinics and insurance all involve a decision that takes months. A 30-day window has always under-credited paid search in those categories, because the sale lands well outside it.
If Qualified Future Conversions reaches Thai accounts, reported conversion volumes and the ranking of channels against each other can move with no campaign change at all. That matters most where a client report or a retainer is benchmarked against last quarter's cost per acquisition, since the comparison stops being like for like the month the window changes.
There is a sharper local point in the qualifying step. The seven-day qualifier leans on branded search, so a brand with thin Thai-language brand search volume generates fewer of the signals that unlock the longer window. Two advertisers with identical campaigns and identical sales can therefore report different QFC volumes purely because one of them is searched for by name in Thai and the other is not. Brand demand in Thai stops being only a brand metric at that point and starts being a measurement input.
FAQ
What is a Qualified Future Conversion in plain terms?
It is a conversion counted up to 180 days after an ad click, allowed only because the user took a qualifying action within seven days of that click. Chen named an attributed branded search, a site visit and an add-to-cart as qualifying actions on the 2 September 2026 episode, and the counting windows he gave were 30, 90 and 180 days.
Is this live in Thailand?
The source did not state it. Neither the 2 September 2026 episode nor Google's announcement page gave a rollout date or a country list, so no availability in Thailand or anywhere else was described.
Do I have to do anything?
Nothing is required of you today, because Google published no rollout information and no setup steps. The one piece of preparation that costs nothing is recording which lookback window produced the conversion figures in your current reports, so that a later comparison against a longer window is possible.
Will this change how Smart Bidding behaves?
Google did not say. The 2 September 2026 episode did not state whether Qualified Future Conversions volumes feed Smart Bidding or sit in reporting only, and the difference between those two determines whether the metric changes your spend or only your reports.
How accurate is the prediction?
Google published no accuracy figure and no error band for Qualified Future Conversions. The announcement page describes the system as modelling the probability of a future purchase from present-day user actions, and neither it nor the episode quantified how often that model is right.
One last thing worth saying plainly, because it will be asked: a longer counting window changes how many sales appear in a report, not how many happen. A rise in reported conversions after a window change is a measurement event rather than a commercial one.
The useful move while the rollout is unannounced is boring and cheap: write down the window behind every conversion number you currently report. When a seven-day qualifier and a 180-day horizon arrive, the accounts that kept that record will be able to tell a measurement change from a performance change, and the ones that did not will be guessing. If you want a second pair of eyes on how your conversions are counted, we are happy to look.
