Your Conversion Window Is Wrong - And Smart Bidding Is Paying for It

Why the Window Matters

The conversion window tells Google Ads how long after a click to attribute a conversion. A 30-day window means any purchase within 30 days of a click gets credited to that click. A 7-day window means only purchases within 7 days count.

This isn't just a reporting setting. It fundamentally changes what Smart Bidding optimises for. The algorithm evaluates each click by its outcome: did this click lead to a conversion within the window? If your window is misaligned with your actual purchase cycle, the algorithm is working with systematically wrong data.

A too-short window undercounts conversions. A too-long window over-attributes them. Both distort bidding - in opposite directions, with different but equally costly consequences.

The Default 30-Day Trap

Most accounts use the default 30-day conversion window because nobody changed it. This default is Google's compromise - long enough to capture most purchase cycles, short enough to maintain some attribution integrity. But "good enough for everyone" means "optimal for no one."

If you sell fast-fashion at £15-25 price points, your purchase cycle is 1-3 days. A 30-day window means Smart Bidding is counting conversions from clicks that happened four weeks ago - clicks that had minimal influence on the purchase decision. The algorithm thinks your historical traffic is more valuable than it actually is and bids too aggressively.

If you sell high-consideration furniture at £500+, your real purchase cycle might be 45-60 days. A 30-day window cuts off a significant portion of genuine conversions. Smart Bidding sees lower conversion rates than reality, concludes your traffic underperforms, and bids too conservatively. You're losing traffic because of a settings page, not a strategy problem.

When the Window Is Too Short

A window shorter than your purchase cycle creates a systematic undercount. Smart Bidding sees fewer conversions per click, calculates a lower conversion rate, and bids more conservatively. The consequences cascade:

• • Reduced traffic: Lower bids win fewer auctions. Impression share drops. You lose visibility on searches you should be capturing.
• • Biased product mix: Products with shorter consideration periods (lower price, impulse buys) appear to perform better because their conversions fit within the window. Higher-value, longer-cycle products get systematically deprioritised.
• • Incorrect ROAS: Your reported ROAS is artificially low because real conversions aren't being counted. Decisions based on this data - budget cuts, campaign pauses, product exclusions - are based on incomplete information.
• • Downward spiral: Conservative bidding reduces traffic, which reduces data volume, which makes Smart Bidding less confident, which makes it bid even more conservatively. A self-reinforcing loop driven by a settings error.

The most dangerous part: a too-short window makes underperformance look like a campaign problem. Agencies restructure campaigns, rewrite ad copy, and rebuild audiences - when the actual issue is a setting that takes 60 seconds to change.

When the Window Is Too Long

A window longer than your purchase cycle creates over-attribution. Clicks that had marginal or no influence on a purchase decision get credited with the conversion. This inflates performance metrics and causes Smart Bidding to overbid:

• • Inflated ROAS: Conversions attributed to clicks from weeks ago inflate the apparent return on those clicks. Your data says traffic is more valuable than it is.
• • Aggressive bidding: Smart Bidding sees high historical conversion rates and bids aggressively - paying more per click than the genuine incremental value justifies.
• • False scaling signals: The account looks like it can absorb more budget because ROAS appears strong. You increase spend, but the incremental traffic converts at lower rates because the attributed performance was partially fictional.
• • Repeat buyer inflation: A long window captures repeat purchases that would have happened anyway. A customer clicks an ad, buys immediately, then buys again 25 days later. With a 30-day window, both purchases are credited to the ad. Only the first was influenced by it.

Over-attribution is harder to detect than under-attribution because everything looks good. ROAS is strong. Conversions are up. Budget utilisation is high. The P&L tells a different story - but by the time finance catches the discrepancy, you've been overbidding for months.

Finding Your Real Purchase Cycle

Your conversion window should match your actual purchase consideration period. Here's how to find it:

• • Google Ads Path Length report: Under Attribution → Paths → Path Length, see how many touchpoints precede a conversion and the typical time between first click and purchase.
• • GA4 Purchase Journey report: Time to Purchase shows the distribution of days between first session and conversion. Look for where 80-90% of conversions have occurred - that's your effective window.
• • Shopify order data: Compare first site visit (from UTM tracking) against order date. Export 3 months of data and calculate the median and 90th percentile time-to-purchase.
• • Google Ads Time Lag report: Shows the distribution of conversion lag in days. If 85% of conversions happen within 7 days, a 7-day or 14-day window captures the vast majority without over-attributing.

The goal is to capture 85-90% of genuine conversions without inflating the window beyond what's defensible. A window that captures 100% of conversions almost certainly includes over-attributed conversions from stale clicks.

It Varies by Category

Purchase cycles aren't uniform across your catalogue. Typical ranges by product type:

If you sell across categories - a home brand selling £15 candles and £800 sofas - a single conversion window can't serve both. You need account architecture that separates these categories with different conversion actions and different windows. One-size-fits-all is one-size-fits-none.

Setting Windows Correctly

Practical steps to fix your conversion windows:

• • Create category-specific conversion actions: Instead of one "Purchase" conversion, create "Purchase - Impulse SKUs" (7-day window) and "Purchase - Considered SKUs" (30-day window). Assign campaigns to the relevant conversion action.
• • Use conversion action sets: Each campaign or campaign group can use a different conversion action set. This lets different campaigns optimise against different windows without creating separate tracking implementations.
• • Test before changing: Switching from a 30-day to a 7-day window will cause a temporary "drop" in reported conversions. Smart Bidding will re-calibrate over 2-3 weeks. Don't panic - the data will be more accurate, and bidding will improve once the algorithm adjusts.
• • Monitor the transition: Track conversion volume and CPA for 3-4 weeks after any window change. Compare against conversion-by-time data to ensure the new window aligns with actual purchase behaviour.

The Bidding Impact

When you align conversion windows with real purchase cycles, Smart Bidding gets cleaner signals. The effects:

• • Accurate product valuation: High-consideration products get properly credited for delayed conversions. Low-consideration products stop claiming credit for coincidental repeat purchases.
• • Budget allocation improves: Smart Bidding distributes spend based on genuine performance, not inflated or deflated signals. Products that actually convert profitably get more budget.
• • ROAS becomes meaningful: Your reported ROAS reflects genuine click-to-purchase attribution within a defensible timeframe. Finance can trust the numbers. Marketing can make decisions on them.
• • Scaling becomes predictable: When you increase budget, the incremental ROAS matches the historical average - because the historical average was accurate to begin with.

This is one of the highest-ROI changes you can make in Google Ads. Zero additional spend. No creative needed. No feed changes. Just a setting that matches your commercial reality - and an algorithm that finally has the data to do its job properly.

Next Steps

Pull your Time Lag report right now. If more than 15% of your conversions happen in the last 3 days of your window, the window is probably too long. If your conversion-by-time data consistently exceeds your click-date conversions, the window might be too short. Either way, Smart Bidding is working with wrong data - and every bid reflects that error.