What a Mismatched Attribution Window Actually Costs: How 7-Day vs. 30-Day Click Windows Warp Your Reported ROAS

The Problem Isn't Your Ads. It's Your Measuring Stick.

Local business owners optimizing Google Ads make one expensive assumption constantly: that the ROAS number in the dashboard reflects what actually happened. Often, it doesn't — not because the platform is lying, but because the attribution window you're using is a filter, not a fact.

An attribution window is simply the time period after a click during which a conversion gets credited to that ad. Change the window, and you change how many conversions get counted. Same spend. Same real-world sales. Different reported ROAS.

This mechanical distortion routinely leads local businesses to pause campaigns that are working, scale campaigns that aren't, and report performance to stakeholders using numbers that aren't comparable week-over-week. Let's build a model that makes the math concrete.

Platform Defaults You Need to Know First

Before the model, a quick grounding in documented platform behavior:

• Google Ads default: 30-day click attribution window for most conversion actions (search and shopping campaigns). This is Google's published default as of 2024.
• Google Ads 7-day window: Available as a manual selection; some automated bidding strategies also evaluate performance on a shorter rolling window internally.
• Meta Ads default: 7-day click, 1-day view — a fundamentally different default that makes cross-platform ROAS comparisons unreliable without normalization.

The core issue for local businesses: if you or your agency ever changed the attribution window — or if you're comparing ad accounts set up at different times — your ROAS trend line may be measuring different things across different periods. That's not optimization data. That's noise.

The Labeled Model: Same Campaign, Two Windows

Let's walk through a straightforward illustrative model. All figures below are clearly labeled as illustrative.

Campaign inputs (illustrative):

• Monthly ad spend: $3,000
• Average job/service value: $400
• Real-world conversions generated in the month: 30 booked jobs
• Real-world revenue: $12,000
• True ROAS: 4.0x ($12,000 ÷ $3,000)

Now, here's where the window matters. Not all 30 conversions happen on the day of the click. For local services — think HVAC, dental, legal, home services — there's often a consideration lag: someone clicks, visits your site, then calls or books 8–20 days later after getting a second quote or checking availability.

Illustrative conversion timing distribution for a local service business: | Days After Click | Conversions Captured | |---|---| | Day 1–7 | 14 | | Day 8–14 | 9 | | Day 15–30 | 7 | | Total (30-day window) | 30 |

Reported ROAS under 30-day window: 30 conversions × $400 = $12,000 revenue reported $12,000 ÷ $3,000 spend = 4.0x ROAS ✓ (matches reality)

Reported ROAS under 7-day window: Only Day 1–7 conversions count → 14 conversions 14 × $400 = $5,600 revenue reported $5,600 ÷ $3,000 spend = 1.87x ROAS ✗ (understates reality by more than half)

The consequence: An owner or media buyer looking at 1.87x ROAS might pause the campaign or dramatically cut budget — on a campaign genuinely delivering 4.0x. The campaign wasn't failing. The window was lying.

The Flip Side: When a Wide Window Flatters a Bad Campaign

The distortion runs both directions. A 30-day window can also overstate ROAS on campaigns with high assisted-conversion overlap — particularly when you're running both branded and non-branded campaigns simultaneously.

Here's the scenario: a user clicks a non-branded ad on Day 1, does nothing, then searches your brand name and converts on Day 28. A 30-day window may credit the non-branded campaign. A 7-day window credits neither (the click is outside the window). A last-click model credits your branded campaign.

For local businesses running both paid search and local service ads, this kind of double-counting inflates non-branded ROAS and makes it look like top-of-funnel keyword spend is doing more work than it is. If you're also running ad schedule bid adjustments (a topic we break down in Ad Schedule Bid Adjustments vs. Flat Bidding for Local Ads), a bloated ROAS figure in your reporting will cause your automated bidding to misprice bids — a compounding error.

The fix isn't always changing your window. It's knowing which window you're using and being consistent.

How to Audit Your Own Attribution Setup (A 4-Step Process)

Here's the decision framework we use before touching any optimization lever on a client account:

Step 1 — Document the current window In Google Ads: Tools & Settings → Conversions → [Conversion action] → Attribution model + window. Write it down. Check every conversion action — they can be set differently.

Step 2 — Map your actual sales cycle length For a local service business, look at your CRM or booking system. What's the median time between first contact and booked job? If it's consistently under 7 days, a 7-day window is defensible. If it's 10–21 days (common for higher-ticket services like remodeling, dental implants, or legal consultation), a 30-day window is closer to reality.

Step 3 — Check for window changes in account history In Google Ads change history, filter for conversion setting changes. If the window was shortened at any point, your ROAS trend is not apples-to-apples. This is one of the most common sources of "our ads stopped working in [month]" diagnoses we see — and it's also worth cross-referencing against landing page CVR data, which we cover in Warm vs Cold Traffic CVR: Local Service Landing Pages.

Step 4 — Normalize before you optimize Before changing bids, pausing ad groups, or restructuring campaigns (see RSAs vs SKAGs: Which Lowers CPA for Local Services? for how that affects downstream quality), make sure your measurement baseline is stable. Optimizing on distorted data produces confidently wrong decisions.

One Cited Benchmark Worth Anchoring To

Google's own published guidance acknowledges that attribution windows meaningfully affect conversion volume reporting, and recommends aligning the window to the typical conversion lag for your business category — not defaulting to the shortest available window simply because it reports faster.

Separately, research by Wordstream / LocaliQ (2023) on local services Google Ads benchmarks consistently shows that service-category conversion rates vary widely by vertical — which means the shape of your conversion timing distribution (how much of your volume comes in days 1–7 vs. days 8–30) is not universal. A roofing company's consideration lag looks nothing like a pizza delivery shop's. Applying a 7-day window across both is categorically wrong for one of them.

(Note: The LocaliQ/Wordstream benchmark report is publicly available and updated annually; exact conversion rate figures vary by category and year — verify the current edition for your vertical before quoting specific numbers.)

The Practical Takeaway

Attribution windows are not a technical setting you configure once and forget. They are a strategic choice that defines what your ROAS number means — and therefore what decisions you're allowed to make confidently from it.

Quick rules of thumb:

• Sales cycle under 7 days → 7-day window is defensible
• Sales cycle 8–30 days → 30-day window is closer to truth
• Mixed channels running simultaneously → normalize to a common window before comparing ROAS figures
• Never compare ROAS across time periods where the window changed without flagging it explicitly

The model above is illustrative, but the mechanic is real. The gap between a 1.87x reported ROAS and a 4.0x true ROAS is the gap between pausing a winning campaign and scaling it.

---

If you're not certain which attribution window your campaigns are running — or whether your reported ROAS reflects your actual business performance — that's exactly the kind of audit we run in the first phase of every engagement. Book a call with Nika Spark and we'll tell you what your numbers actually mean before we touch a single bid.