Two reports, two answers, one truth: reconciling first-touch lag and click lag
I ran a lag analysis on a client account last week and got two answers to the same question. GA4 said roughly a quarter of paid-touched purchases took more than a week to close. Google Ads said about 6% did. Same account, same purchases, same week.
The usual instinct here is that one report is broken, and the usual next move is to trust whichever number supports the decision you already wanted to make. Both moves are wrong. Both reports were correct, and the distance between them turned out to be the most useful finding in either one.
Two clocks, two starting guns
The disagreement dissolves the moment you look at what each report actually measures, because they are not timing the same race.
GA4's Attribution paths report starts its clock at the first touchpoint it can see. For this client, a premium DTC apparel brand on Shopify with a $300+ average order value, that meant ad-involved purchase journeys averaging about 7 days and about 3 touchpoints from first contact to checkout. That is the full consideration journey: the first ad impression that registered, the return visits, the comparison shopping, the sleep-on-it nights.
Google Ads' conversion lag buckets start a different clock, at the last actionable ad interaction. From that click, most conversions landed fast. Only somewhere in the 5 to 10% range arrived beyond day 7. Measured from the click, this looks like a quick-closing account.
So one system times the whole journey and the other times the final sprint. Neither is lying. They are answering different questions, and if you read either one without knowing which question it answers, you will size budgets and judge campaigns against a number that does not mean what you think it means.
The gap is the finding
Here is the part worth the price of admission: subtract the two and you get a number neither report shows you directly.
If the full journey averages a week, but conversions come quickly once the final click happens, then most of the journey is happening before that click. The gap between first-touch lag and click lag is the pre-click consideration phase, measured. For this account it meant buyers were spending days forming a decision across multiple touches, then clicking an ad late, close to the moment they were already ready to buy.
That reframes what the credited click is. Last-click thinking treats it as the persuasion event. The lag math says it is often closer to a checkout door: the ad that happened to be standing there when a mostly formed decision walked through. The persuasion happened upstream, across touches that will never hold the attribution credit.
What this changes about where the money goes
The practical consequence shows up in budget meetings. If the credited click arrives late in an already-formed decision, then the campaigns that harvest those clicks, branded search above all, will always look like heroes in last-click reporting. The upper-funnel content, the prospecting campaigns, and the remarketing that did the slow work of forming the decision will look like they barely participated.
Cut them on that evidence and nothing breaks immediately. The branded clicks keep converting for a while, because the pipeline of formed decisions takes time to drain. Then the harvest thins, and the reports never quite explain why, because the reports were only ever watching the sprint.
This is the same family of problem I wrote about in the attribution windows piece: the measurement system quietly shapes what the machine learns and what you fund. A platform running a 7-day click window on this account would not only under-report the slow tail, it would train its bidding on fast closers while the lag data says the real work happens over a week of consideration. The window question and the two-clocks question are the same question wearing different clothes: does your measurement match how your customers actually buy?
When two systems disagree, neither is the enemy
The habit this account taught is the transferable part, and it applies well beyond lag.
When two reporting systems disagree, the common reactions are to declare one broken, to average them, or to pick the flattering one. All three throw away the information. The better move is to work out precisely what each system measures: where its clock starts, what population it counts, what it can and cannot see. Do that, and the disagreement usually stops being a data-quality problem and becomes the insight itself. Here, a 25% number and a 6% number disagreed their way into revealing the pre-click consideration phase, something neither report was designed to show.
GA4's own documentation is clear that Attribution paths measures from the first touchpoint of the journey, and Google's conversion lag reporting measures from the ad interaction. The reconciliation was sitting in the definitions the whole time. Most measurement mysteries are.
Run this on your own account
The diagnostic takes about 30 minutes. Pull GA4's Advertising section and read days-to-conversion and touchpoints for your ad-involved journeys. Then pull your Google Ads lag data and read the share of conversions landing beyond each bucket. If the two numbers roughly agree, your buyers decide fast and click fast, and last-click reporting is treating you fairly. If they diverge the way this account's did, you now have a measured pre-click phase, and a reason to defend the spend that fills it.
This is the analytics work behind the paid media work: making sure the numbers you steer by mean what you think they mean. It is most of what my analytics and reporting engagements are. And if two of your reports are contradicting each other right now, tell me what you're working on; that disagreement is probably worth more than it looks.