What a 6-Month Attribution Window vs. a 30-Day Window Does to Reported ROAS

The window can change the report—not the business

Your attribution window determines how long an advertising interaction remains eligible for conversion credit. Changing that window can materially change reported return on ad spend even when actual revenue and ad spend stay exactly the same.

That matters because reported ROAS often drives budget decisions:

Reported ROAS = attributed revenue ÷ ad spend

A 30-day window may exclude revenue from leads that take several months to buy. A six-month window may capture those sales—but could also credit an old ad interaction when referrals, repeat visits, sales follow-up, or another channel did more of the work.

Google Ads documents configurable conversion windows, while Google Analytics documents lookback-window controls for attribution reporting. Meta also applies attribution settings to the conversions shown against ads. These platform settings and definitions are not necessarily identical, so a Meta ROAS and a Google ROAS should not be compared until their windows and conversion rules are aligned.

The key point: a longer window does not create more revenue. It assigns more existing revenue to earlier interactions.

First separate three different clocks

Before comparing a 30-day report with a six-month report, identify which clock is changing:

1. Ad-platform conversion window: How long Meta or Google can credit a conversion after an eligible click or view. 2. Analytics lookback window: How far an analytics tool looks back when assigning conversion credit across known touchpoints. 3. CRM sales-cycle window: How long you continue matching leads to closed jobs in your own customer data.

These are not interchangeable. A six-month CRM analysis can show that a lead eventually became revenue without proving that an ad platform deserves full credit for the sale.

Also document whether reports use click-through, engaged-view, or view-through interactions. A 30-day click-based figure and a broader report containing view-through credit are not a clean window-length comparison.

Use one written definition for every decision report: interaction type, attribution model, window length, conversion event, revenue source, and reporting date. Without that record, ROAS can move simply because someone changed a setting.

Model 1: A short purchase cycle

Consider an emergency plumbing campaign, where customers usually need help quickly.

Illustrative model—not a measured benchmark:

• Ad spend: $10,000
• Revenue matched to eligible ad interactions within 30 days: $32,000
• Additional matched revenue appearing during months two through six: $6,000
• Reported 30-day ROAS: $32,000 ÷ $10,000 = 3.2x
• Reported six-month ROAS: $38,000 ÷ $10,000 = 3.8x

The longer window raises reported ROAS by 0.6x in this model. But the analyst still needs to inspect the extra $6,000. Did those customers return because of the original ads? Were they existing customers? Did an email campaign, referral, or branded search close them later?

For an urgent service, a large amount of late revenue can be a warning sign rather than proof that the six-month number is better. The 30-day figure may more closely reflect the campaign’s immediate economic role, while the longer view is useful as a secondary cohort report.

Model 2: A long-consideration service

Now consider a remodeling company. Prospects may research, request estimates, compare contractors, arrange financing, and schedule work well after the first click.

Illustrative model—not a measured benchmark:

• Ad spend: $15,000
• Closed revenue within 30 days: $18,000
• Additional closed revenue from the same lead cohort during months two through six: $42,000
• Reported 30-day ROAS: $18,000 ÷ $15,000 = 1.2x
• Six-month cohort ROAS: $60,000 ÷ $15,000 = 4.0x

A 30-day report could make the campaign look weak and prompt a budget cut before qualified opportunities mature. The six-month cohort view gives a more complete picture of revenue, provided the business reliably connects the original lead to the eventual customer.

This is why lead cost alone is insufficient. Cheap leads that never close can underperform expensive leads that produce profitable jobs. Evaluate revenue, gross profit, customer acquisition cost, and eventual customer value together. For that broader calculation, see LTV:CAC Benchmarks for Local Service Businesses.

Where mismatched windows misallocate budget

Suppose Meta is judged on a six-month CRM cohort while Google is judged on a 30-day platform report. Meta receives more time to accumulate revenue, so its reported ROAS may look stronger even if Google assisted many of the same purchases.

The reverse can happen when paid search captures the final branded query. Google may receive last-touch credit for demand initially created elsewhere, causing the business to over-invest in closing demand and under-invest in generating it.

Overlapping attribution creates another trap. If Meta claims $40,000 and Google claims $35,000, that does not prove the campaigns jointly generated $75,000. The same sale may appear in both reports.

Use these controls:

• Compare channels over the same customer cohort and maturity period.
• Reconcile attributed sales to actual CRM revenue.
• Deduplicate orders, customers, and closed jobs across platforms.
• Separate new customers from returning customers.
• Report both platform-attributed ROAS and deduplicated business-level ROAS.
• Avoid changing budgets based on immature cohorts.

Campaign structure also affects interpretation. One Google Ads Campaign or One Per Service? explains why separating services can improve budget control, while Minimum Remarketing Audience Size for Local Ads helps frame whether remarketing is operationally viable in a small local market.

A practical budget decision model

Use a maturity table instead of choosing one supposedly perfect window. For each monthly lead cohort, calculate revenue observed by day 30, day 90, and month six. Those checkpoints are analytical choices, not universal benchmarks.

Then apply this process:

1. Find the sales-cycle curve. Determine what share of eventually closed revenue is typically visible at each checkpoint using your own CRM history. 2. Set a primary decision window. Use the shortest window that captures a stable, meaningful share of outcomes for that service. 3. Keep a longer validation view. Use mature cohorts to detect delayed revenue and test whether early indicators predict final results. 4. Reconcile to economics. Evaluate attributed revenue, gross margin, refunds, cancellations, and repeat revenue—not just form submissions. 5. Change budgets incrementally. Hold targeting, conversion definitions, and attribution rules steady long enough to separate real performance changes from reporting changes.

A useful dashboard shows both 30-day observed ROAS and mature cohort ROAS, clearly labeled. If the gap is consistently small, shorter feedback loops can guide decisions. If the gap is large and repeatable, short-window reports need a maturity adjustment.

The right attribution window is not the one that produces the highest ROAS. It is the one that best matches the purchase cycle while minimizing unsupported credit. If you want help building a channel-level revenue model from ad-platform and CRM data, book a call with Nika Spark.