If you're asking whether you're wasting money on retargeting, you're already thinking about it more carefully than most advertisers do. Most brands allocate 20-40% of their digital budget to retargeting, see a 5-8x ROAS in the dashboard, and never question it. The ROAS looks real. The attribution logic behind it does not.
The problem is not that retargeting does not work. Some retargeting does. The problem is that the metric most brands use to evaluate it (attributed ROAS) is almost perfectly designed to make retargeting look better than it is.
Most retargeting spend targets people who would have purchased regardless
Why retargeting ROAS is misleading
Retargeting targets people who already visited your site. By definition, these people already know your brand and showed enough interest to arrive at your website. A meaningful portion of them were going to buy without ever seeing another ad from you.
When they convert, the retargeting ad sitting in the attribution window gets the credit. The ad did not necessarily cause the conversion. It was present when the conversion happened, which is all that attribution requires.
The structural problem: the higher the organic purchase intent of your retargeting audience, the more misleading your attributed ROAS becomes. A broad retargeting pool of all site visitors in the last 90 days will look much better in the dashboard than it performs incrementally, because that audience is full of people who were already going to buy.
A quick sanity check without running a formal test
You do not need a holdout experiment to get a rough estimate of whether your retargeting is incremental. You need two numbers: your site's organic conversion rate and your retargeted audience's conversion rate.
If your site converts 2% of visitors organically (without ads), and your retargeting audience converts at 4%, the ads appear to be adding approximately 2 percentage points of lift. Roughly 50% of retargeting conversions are incremental. That is a respectable result.
If your organic conversion rate is 3.5% and your retargeted conversion rate is 4.2%, only about 17% of retargeting conversions are incremental. You are spending to claim credit for conversions that were going to happen anyway at a rate of roughly 83%.
This is a rough calculation, not a controlled experiment. Your retargeted audience is not a random sample of your full audience, it skews toward more engaged visitors. But the direction is usually right, and the gap between organic and retargeted conversion rates is a useful early signal.
Signs your retargeting is probably not worth what you're spending
Your retargeting ROAS is 5x or higher but your overall MER (total revenue / total ad spend) is 2x or below. The gap between channel-level ROAS and business-level efficiency is one of the clearest signals that your retargeting is consuming budget without adding proportionate incremental revenue.
Pausing your retargeting for a week barely moves total revenue. If your business performs nearly identically with and without retargeting over a 2-week window, that is evidence the ads are not driving much incremental demand. The carryover effect from the prior weeks muddies a short pause, so a two-week observation period is the minimum.
Your retargeting audience is defined broadly. All visitors in the last 90 days, all video viewers, anyone who engaged with an ad. These audiences contain a wide range of intent levels and the majority are not close to purchasing. You are showing ads to people who visited your site once two months ago and have no active purchase intent.
You are spending heavily on retargeting people with minimal engagement signals: single-page visits, time-on-site under 15 seconds, no product page views. These visitors never had meaningful purchase intent and retargeting them rarely drives incremental conversions.
Signs your retargeting is probably working
Your MER improves noticeably when you increase retargeting spend and softens when you reduce it. Business-level efficiency moving with channel spend is the strongest signal that the channel is generating genuine incremental revenue.
Your holdout tests show 50% or higher incremental conversion rates. If you have run a formal holdout and found that more than half of retargeting conversions are incremental, the spend is justified.
You are targeting tight, high-intent audiences: cart abandoners in the last 7 days, product page viewers in the last 14 days, past purchasers with high repeat-buy likelihood. These segments have much higher organic intent, but they also have a clearer reason to need a reminder. The incrementality tends to be higher because the intervention is more timely.
What to do
Tighten your audience criteria. Remove all visitors and replace with segmented, intent-based lists. Cart abandoners with 7-day recency. Product page viewers with 14-day recency. High-LTV customers approaching a typical repurchase window. Each of these segments has a more coherent reason to respond to a retargeting ad.
Shorten your retargeting windows. A 90-day window is almost always too long. Most incremental influence from retargeting happens within the first 7-14 days after a site visit. Visitors beyond 30 days have, in most categories, already made their decision and moved on.
Run a four-week holdout test. Suppress ads to 15-20% of your retargeting audience and compare backend conversion rates. This will give you a real incrementality number rather than a back-of-envelope estimate.
If your incrementality test shows that fewer than 30% of conversions are incremental, reallocate a portion of that budget to prospecting. Cold audiences have lower organic intent and tend to show much higher incrementality. The attributed ROAS will look worse. The incremental ROAS will often be better.