Your CFO just asked why Google Ads shows 30% more conversions than GA4 for the same campaign. You've been dreading this question because the honest answer involves explaining that both platforms are technically correct, just measuring different things with different rules. The January 2026 GA4 update finally gives you a way out of this conversation.
Google's January 16, 2026 release introduced something marketers have requested for years: the ability to adjust conversion attribution settings independently for every conversion action. This isn't a cosmetic change. It's the difference between a phone call conversion and a form fill being forced into the same 90-day lookback window, and being able to set each one according to how your buyers actually behave.
The Mismatch That Erodes Trust
The core problem has always been architectural. GA4 renamed conversions to "key events" in March 2024 precisely because the word "conversion" meant different things in GA4 and Google Ads. A key event in GA4 measures an important on-site action across all traffic and channels. A conversion in Google Ads is a key event that's been imported and re-attributed using Google Ads' own model, scoped only to ad interactions. The columns will never match exactly, and that's by design.
But the lookback window problem compounds this. GA4's default settings use a 30-day window for acquisition key events and 90 days for all other click-through key events. If your sales cycle is 45 days, that 30-day acquisition window means GA4 forgets the original touchpoint that started the journey. The conversion gets attributed to "Direct" or "Organic Search" instead of the paid campaign that actually created the demand.
For B2B companies with complex buying committees, this creates a specific failure mode: your upper-funnel campaigns look worthless because they fall outside the attribution window by the time the deal closes. You cut budget from the campaigns that create pipeline and over-invest in the ones that happen to catch the last click.
What Custom Windows Actually Change
The January 2026 update lets you set attribution windows at the individual conversion level. A phone call conversion can use a 7-day window because callers typically convert fast. A demo request can use 60 days because enterprise buyers take longer. A purchase confirmation can use 30 days because your e-commerce cycle is shorter than your lead-gen cycle.
The new Conversion Attribution Analysis Report, which entered beta in February 2026, adds visibility into how these windows affect your data. It separates assisted conversions from last-click conversions and categorizes touchpoints into Early, Mid, and Late stages of the conversion path. This matters because some campaigns primarily close simple, direct-response conversions while others nurture complex journeys involving six or more touchpoints over weeks.
The practical implication: you can finally build attribution dashboards that accurately value every part of your marketing funnel instead of systematically under-crediting awareness campaigns.
The Shorter-Window Argument
There's a counterintuitive case for using narrower windows than your sales cycle suggests. Some practitioners argue that in a world of consent banners, ad blockers, and privacy extensions, longer windows produce muddier data. A user who clicked your LinkedIn post yesterday and purchased today provides cleaner signal than an ad click from 80 days ago that may or may not have influenced the decision.
The logic: in multi-channel marketing, customers are exposed to multiple touchpoints over long periods. Not all touchpoints are equally valuable. A 90-day window means many channels can take credit for the same conversion, which inflates reported revenue and makes ROAS calculations unreliable.
This isn't a universal recommendation. It's a trade-off. Shorter windows undercount top-of-funnel work. Longer windows overcount it. The right answer depends on your sales cycle, your channel mix, and how much you trust the identity resolution in your tracking stack.
The CFO Conversation, Revised
Here's how to use custom attribution windows to build a board-ready measurement framework:
First, audit your actual sales cycles by conversion type. Pull CRM data on time-to-close for different lead sources and deal sizes. A $5,000 deal closes differently than a $500,000 deal, and they shouldn't share attribution settings.
Second, set lookback windows that match reality. GA4's attribution settings page lets you configure windows between 30, 60, and 90 days for key events. Match these to your documented sales cycles, not to Google's defaults.
Third, document your assumptions. When you present conversion data to finance, include a one-line note on the attribution window for each conversion type. "Demo requests use a 60-day lookback because our average enterprise sales cycle is 52 days" is a defensible position. "We used the default settings" is not.
Fourth, expect the numbers to change. When you adjust attribution windows, historical data recalculates. Your Q1 numbers will look different than they did before the change. Prepare a brief explaining why, and show the before/after comparison so stakeholders understand the methodology shift.
The Remaining Gap
Custom attribution windows don't solve the fundamental mismatch between GA4 and ad platforms. Google Ads uses data-driven attribution scoped only to ad interactions. GA4 uses data-driven attribution across all channels. If someone clicked a Meta ad, left, then came back through a Google ad and converted, Google Ads only sees the Google ad. GA4 sees both.
This means you'll still have two sets of numbers. The goal isn't to make them match. It's to understand what each one measures and use the right number for the right decision. Use Google Ads data for bid optimization. Use GA4 data for cross-channel budget allocation. Use CRM data for actual revenue.
Custom attribution windows give you one more lever to reduce the gap between reported conversions and business reality. They don't eliminate the gap. They make it smaller and more explainable.
The Two-Week Pilot
If you haven't touched your GA4 attribution settings since setup, here's a starting point:
Pull your CRM data on time-to-close by lead source. Identify your three highest-volume conversion types. Set lookback windows that match the 75th percentile of your sales cycle for each one (not the average, which gets skewed by outliers). Document the settings and the rationale. Run for two weeks. Compare the new attribution data to your CRM actuals.
If the numbers get closer to reality, you've found a better configuration. If they don't, you've learned something about where the measurement gap actually lives, and it's probably not in the attribution window.
