Google Ads Conversions vs Conversions by Conversion Time

Both columns count the same attributed conversions. The only difference is the date each conversion is filed under. “Conversions” files a conversion on the date of the ad interaction it was attributed to. “Conversions (by conv. time)” files it on the date the conversion itself happened.

That is the whole distinction. Neither column is a sales ledger, neither is automatically your P&L, and choosing which one to look at does not change how your campaigns bid. Google documents the behaviour of both columns in its own help centre: About conversion time reporting.

A Worked Date Example

A hypothetical single conversion, to make the date basis concrete. Someone clicks your ad on 31 August. They come back and purchase on 2 September, with a conversion value of £100. Google Ads attributes that purchase to the 31 August interaction.

One purchase. One attribution decision. Two different months, depending on the column. Illustrative figures, chosen to isolate the date basis.

The Two Columns Compared

What Conversion Time Is Not

• • Not all your sales. It contains only conversions Google Ads attributed to an ad interaction, within your attribution model and conversion windows. Orders with no ad interaction, or falling outside the window, appear in neither column.
• • Not cash or P&L. It is closer to transaction dates than click-date reporting, which makes reconciliation easier. It is still tracked conversion value, not booked revenue.
• • Not a bidding change. Adding a column changes what you see, not what the bid strategy optimises towards.

Reconciling Against Your Shop Platform

Conversion-time reporting removes the date-basis difference between Google Ads and your shop platform. It does not remove the other reasons the numbers differ. Before you conclude anything is wrong, check each of these:

• • Which conversion actions are in the column. Filter to the purchase action you are comparing, rather than every action counted in the account.
• • Timezone. Your Google Ads account timezone and your shop platform's reporting timezone may cut the day differently.
• • Currency. Multi-currency stores may convert at a different rate or moment than the values sent to Google Ads.
• • Tax, shipping, discounts and refunds. Whether these are inside the conversion value you send is a configuration decision, and refunds may not be netted off unless you adjust conversions.
• • Attribution scope and windows. The attribution model and conversion window decide which interactions get credit, and for how long.
• • Modelling and deduplication. Some conversions are modelled rather than individually observed, and counting settings decide whether repeat conversions from one interaction are counted once or every time.

Expect a residual difference and agree, in advance, how large a gap is acceptable. That single decision removes most of the “our numbers don't match” conversations between marketing and finance.

Adding the Columns

1. Open the campaign, ad group or conversions report you want to read.
2. Open the column editor and search for “conv. time”.
3. Add “Conversions (by conv. time)” and “Conv. value (by conv. time)” alongside the standard columns, rather than instead of them.
4. Segment or filter by conversion action so you are comparing the purchase action, not the account total.
5. Save the column set so the same view is used every time the report is run.

Google's help centre documents the exact column names and behaviour, and is the source to check if the interface changes.

Conversion Lag vs Reporting Delay

These are two different things and they are constantly confused.

• • Conversion lag is real-world behaviour: the time between the ad interaction and the purchase. It is a property of your products and buyers, and it differs by category, price point and audience.
• • Reporting delay is systems behaviour: the time between a conversion happening and it appearing in the interface, including any import or upload schedule you run.

Both make recent periods look worse than they will eventually be, so both argue for patience with fresh data. Neither has a universal settling period. Measure your own: compare the same date range pulled a week apart and see how much it moved. That measured figure is the only defensible basis for saying when your data is stable, and it is worth revisiting when your mix or conversion window changes. The related problem of acting on unsettled data is where most avoidable interventions come from.

Two Reporting Mistakes to Avoid

Mixing date bases in one ratio. Dividing the spend for a calendar period by conversion-time revenue for the same period is not a cohort ROAS. The spend belongs to interactions in that period; the revenue belongs to conversions in that period, some of which came from earlier interactions. If you want a cohort view, hold the cohort constant and let the revenue mature.

Treating a column change as a fix. If reported values are wrong at source, both columns are wrong, on two different date bases. Conversion setup comes first; which date the number is filed under comes second.

Next Steps

If marketing and finance are reading different columns and nobody has agreed a reconciliation tolerance, the disagreement will repeat every month. Settle the measurement basis first, then judge the campaigns.