A 7-day click window isn't a measurement strategy. It's a default. Set each attribution window from your observed time to convert, keep platform and CRM windows separate, and reconcile both against revenue before you change a budget.
Attribution windows decide which ad gets credit when a conversion happens after an interaction. The wrong window can make retargeting look brilliant, hide the value of prospecting, and leave your CRM reporting out of sync with Meta or Google. In 2026, the fix is a small window policy tied to real conversion lag.
What is a paid media attribution window?
A paid media attribution window is the period after an ad interaction during which a platform can credit a conversion to that interaction. A click window covers post-click actions. A view window covers post-impression actions. A CRM lookback window covers earlier first-party touches. They answer related questions, but they aren't interchangeable.
- Attribution window
- The number of days between an ad interaction and a later conversion during which the interaction remains eligible for credit. The window is a reporting rule, not proof that the ad caused the conversion.
Why the default window creates bad decisions
Defaults are built for platform reporting, not your buying cycle. A short window misses delayed sales. A long window gives old touches too much credit. The result is predictable: paid media reports one number, the CRM reports another, and the team argues about whose dashboard is right instead of fixing the time rule.
Google Ads currently defaults to a 30-day click-through window, a 1-day view-through window, and a 3-day engaged-view window. Google says click-through windows can range from 1 to 90 days and recommends at least 7 days for richer conversion data. Meta's current website and in-store settings support 1-day or 7-day click, 1-day view, and 1-day engage-through windows.
Those settings aren't equivalent across platforms. A 7-day Meta click window can claim a different set of people than a 30-day Google click window. If your CRM uses 90 days for pipeline, platform ROAS will look cleaner than the customer journey really is.
How to choose the right window for your buying cycle
Choose a window from the distribution of days to conversion, not from the platform's default menu. Start with the median and 75th percentile for the same conversion event, then set a reporting window that captures most real conversions without turning every old touch into a win. Recheck the distribution by channel and offer.
- Conversion lag
- The elapsed time between a meaningful marketing interaction and the business outcome you care about. Measure it from first touch to conversion and from last eligible touch to conversion, because the two distributions answer different budget questions.
| Journey pattern | Starting click window | CRM lookback | Use this for |
|---|---|---|---|
| Fast DTC or low-consideration purchase | 1 to 7 days | 7 to 14 days | Purchases with most demand created and captured quickly |
| Considered lead generation | 7 to 30 days | 30 to 60 days | Booked calls, qualified leads, and short sales cycles |
| High-ticket service or wealth offer | 30 to 90 days | 90 to 180 days | Pipeline and closed revenue after nurture and sales |
| Enterprise or multi-stakeholder deal | 30 to 90 days | 90 to 180 days | Opportunities that need sales-stage and revenue reconciliation |
These are starting ranges, not universal benchmarks. Pull your own days-to-conversion report. HubSpot's May 2025 buyer journey research found that 87% of buyers complete purchases within a six-month sales cycle, which is a reminder that a 7-day media window can be too narrow for high-consideration offers even when the first lead arrives quickly.
Keep platform windows and CRM lookbacks separate
Your ad platform needs a window it can use for optimization. Your CRM needs a longer lookback to understand the path to revenue. Keep both, name both, and never present them as the same KPI. Platform ROAS is an operational signal. CRM revenue by source is a business outcome. They should be reconciled, not forced to match.
- Platform window: the click, view, or engage-through rule used for delivery and in-platform reporting.
- CRM lookback: the first-party period used to connect touches to lead, opportunity, and closed-won events.
- Finance window: the period used for payback, contribution margin, and cohort revenue decisions.
- Experiment window: the pre-defined period used to measure lift against a control group.
Salesforce's May 2026 attribution documentation treats the lookback window as a configurable rule alongside identity resolution, signals, and conversion metrics. That is the right mental model. A window is part of the data contract, not a hidden setting buried in one ad account.
Why offline conversion timing matters
A long CRM lookback only helps if downstream events reach the platform while the click identifier is still usable. Send approved events as soon as they happen, preserve the original event time, and log late or rejected uploads. A 90-day reporting window cannot recover an event that the platform rejected after its transport deadline.
Meta's Conversions API guidance says offline events can be up to 7 days old when sent, transactions should be uploaded within 62 days of conversion, and the maximum deduplication window is 7 days. Google says its GCLID is kept for 90 days and recommends uploading offline conversions more frequently. The rule is simple: a long measurement window needs fast transport.
The four checks that stop window drift
Window governance is a recurring control, not a one-time setup. Review the rule whenever the offer, sales cycle, channel mix, or conversion event changes. Then compare platform results, first-party events, and CRM outcomes by age bucket so you can see exactly where credit is accumulating or disappearing.
- Publish one window policy per conversion event, including click, view, CRM, and finance rules.
- Report days to conversion in buckets: 0 to 1, 2 to 7, 8 to 30, 31 to 90, and over 90 days.
- Compare platform-attributed conversions with CRM-accepted and closed-won outcomes by channel.
- Change only one window at a time, annotate the effective date, and avoid comparing mixed settings.
| Check | Healthy signal | Warning signal |
|---|---|---|
| Days to conversion | Most outcomes fit the declared window | Large late tail outside the window |
| Platform to CRM match | Gap is stable and explained | Gap widens after a tracking change |
| Event freshness | Events arrive in real time or daily | Events arrive days after the CRM update |
| Window change log | Every change has an owner and date | Teams compare reports with hidden settings |
The practical 30-day window audit
You can audit attribution windows in 30 days without buying another tool. Start with one conversion event and one major channel. Pull raw timestamps, calculate conversion lag, document platform limits, and compare the result with CRM revenue. Then lock a rule that operators can explain in one sentence.
- Days 1 to 5: export ad interactions, first-party events, CRM stages, and closed revenue for one offer.
- Days 6 to 10: calculate median, 75th percentile, and late-tail conversion lag by source.
- Days 11 to 15: document platform windows, upload limits, identity fields, and deduplication rules.
- Days 16 to 25: set the window policy, repair the event path, and annotate the effective date.
- Days 26 to 30: publish a reconciliation report and create the next monthly review.
Frequently asked
Is a 7-day click window always the right choice for Meta?
No. It can work for fast conversion journeys, but it can undercount delayed demand. Use your conversion-lag distribution and keep the CRM lookback separate.
Should my Google and Meta windows match?
They should be comparable and documented, not blindly identical. Platform rules differ, so reconcile both against the same first-party and CRM outcomes.
What window should a high-ticket service business use?
Start with 30 to 90 days for platform click reporting and 90 to 180 days for CRM revenue analysis, then replace those ranges with your observed sales-cycle data.
Does a longer window prove an ad caused the sale?
No. It only makes the touch eligible for credit. Use experiments or incrementality tests when you need a causal answer.
What should $1M to $100M+ brands serious about growth do first?
Pick one revenue-linked event, measure its conversion lag, document every window, and reconcile platform data to CRM outcomes before changing spend.
Moonshot is the agency for $1M to $100M+ brands serious about growth. We build the first-party data architecture, paid media systems, and CRM operating rhythm behind better decisions. FlowOS is the SaaS product in that ecosystem, not the agency itself.