A quick primer
Google's own troubleshooting guide names this delay as the main factor to account for when evaluating a channel: conversion delay matters, some channels are affected more than others, and you need to adjust your date range to exclude the delay period to fairly judge ROI at the channel level. Without that adjustment, a channel's CPA or ROAS (and the whole campaign's) will look worse than it really is until the delayed conversions get reported.
The mechanics are documented separately: conversion delay is the gap between someone clicking an ad and completing a conversion action (an app install, a purchase, adding something to a cart, and so on). Since Google Ads reports conversions by query date, the most current numbers aren't always fully in yet, which can make CPA look inflated and ROAS look deflated. A related official page confirms this is inherent to the system: conversion delay happens because a user clicks and then takes time before converting, and because conversions are reported by impression date, the freshest figures are systematically understated at the moment you're looking at them.
Google gives you a specific tool to size this delay for your own account: conversion lag reporting lets you hover over your conversion metrics ("Conversions," "Cost/conv.," "Conv. rate") to see the average delay and a forecast of how many conversions haven't been reported yet. The gap with your daily figures isn't random; it's predictable, and the interface will tell you the number if you know where to look.
What to check before you touch anything
- Have you hovered over the conversion metrics to see the average delay and the forecast of unreported conversions. This is the fastest way to size the gap for your own account, not a general guess.
- What date range are you comparing daily figures against. The most recent three to seven days are almost always incomplete purely because of the delay, regardless of channel.
- Does delay differ meaningfully by channel. Channels with a longer path to purchase (YouTube or Display, typically) usually show a bigger delay than direct Search traffic.
- What attribution model you're using. The report supports whichever model you've assigned to each conversion action, and different models can spread out when a conversion is credited.
Possible approaches
- Exclude the last few days of your window, matching your account's average delay, instead of comparing raw daily figures against the finished report.
- Use conversion lag reporting (hovering over the metrics) as an account-specific estimate rather than a generic rule of thumb about how many days to wait.
- When judging a specific channel, remember that channels with a longer path to conversion will naturally look worse in fresh data purely because of the delay; don't jump to a conclusion about weak performance before the numbers settle.
- Reconcile final figures only after the delay for your selected period has fully cleared, usually one to two weeks depending on your business's conversion cycle, rather than in the moment.