AdRoll Ads ROI: Turning Spend Into Revenue You Can Track

AdRoll is a retargeting and prospecting platform popular with e-commerce brands, especially on Shopify, BigCommerce and WooCommerce. It runs display, native and social ads across its network and partners such as Facebook and Instagram, plus email retargeting. The AdRoll Pixel tracks site visits and conversions and builds audiences from them.

In the AdRoll dashboard you will see impressions, clicks, spend, attributed conversions, attributed revenue and ROAS. Attribution includes:

• Click-through conversions: a purchase after someone clicked an AdRoll ad within the click window.
• View-through conversions: a purchase after someone saw, but did not click, an ad within the view window.
• Multi-touch credit across channels, depending on your attribution settings.

View-through conversions are the main reason AdRoll’s reported ROAS can look much higher than what GA4 shows. Retargeting audiences are, by definition, people who already visited your site, and many would have returned to buy without seeing an ad.

ROAS divides attributed revenue by ad spend. ROI asks whether the campaign made money. The difference is easy to show with an illustrative example.

• Ad spend in a month: 3,000 USD.
• AdRoll-attributed revenue: 15,000 USD, a reported ROAS of 5.
• Of that revenue, click-through conversions account for 6,000 USD; the rest is view-through.
• Gross margin on products: 40 percent.

If you count everything, gross profit on attributed revenue is 6,000 USD against 3,000 USD of spend. If you count only click-through revenue, gross profit is 2,400 USD, which means the campaign loses money. Neither figure is the true answer, because some view-through sales were influenced by ads and some click-through sales would have happened anyway. That is why incrementality testing matters.

An incrementality test compares people who could see your ads with a similar group who could not.

• Geographic holdout: pause AdRoll in a set of comparable regions for three to four weeks and compare total sales trends with regions where ads continued.
• Audience holdout: exclude a random share of your retargeting audience from ads and compare their conversion rate with the exposed group.
• Time-based test: pause campaigns for a period and watch total revenue, accepting that seasonality makes this the weakest method.

Ask your AdRoll account representative whether a lift study is available for your account size, as AdRoll has offered incrementality measurement for some advertisers. Whatever method you use, decide the success threshold before the test starts.

Several settings decide whether AdRoll spends on people likely to buy or on people who would buy anyway.

Attribution windows

Shorter view-through windows reduce the credit AdRoll takes for sales it did not influence. Review the click-through and view-through windows in your attribution settings and choose ones that match your typical purchase cycle. Use the same settings when comparing periods.

Audience segmentation

A single ‘all visitors’ audience treats a bounced visitor the same as someone who abandoned a full cart. Segment instead:

• Cart abandoners in the last 7 days: highest intent, worth the highest bids.
• Product viewers in the last 14 to 30 days: dynamic product ads showing the items they viewed.
• Blog or homepage visitors: lower intent, better suited to brand or offer messages with lower bids.
• Recent purchasers: exclude from acquisition campaigns, or target only with cross-sell offers after a sensible delay.

Frequency caps and burn pixels

Set frequency caps so the same person is not shown your ads many times a day. Make sure the conversion pixel removes buyers from retargeting audiences quickly, so you stop paying to chase people who already purchased.

Once measurement is honest, improvements usually come from a short list.

• Shift budget from low-intent to high-intent segments.
• Refresh creative every few weeks; dynamic ads with clear prices and product images tend to outperform generic brand banners for retargeting.
• Use email retargeting for known contacts, which can be cheaper than display for the same audience.
• Test prospecting campaigns separately, judging them on new-customer revenue rather than total ROAS.
• Pause placements or channels with spend but no click-through conversions over a meaningful period.

For example, a home fragrance brand might discover that most of its click-through revenue comes from cart abandoners, while homepage visitor campaigns produce mostly view-through credit. Moving budget toward the cart and product-viewer segments typically improves ROI even if reported total ROAS falls.

AdRoll is strongest for e-commerce brands with enough site traffic to build meaningful retargeting audiences. With very low traffic, audiences stay below the size needed to serve, and fixed costs such as platform fees eat a larger share of spend.

It can also overlap with what you already run. If Meta and Google Ads both retarget the same visitors, adding AdRoll may simply mean three platforms paying to reach one person, each claiming the same sale. Before adding or keeping AdRoll, list every platform that retargets site visitors and check the combined frequency a typical visitor experiences.

• Under a few thousand monthly site visitors: retargeting audiences may be too small to justify a dedicated platform.
• Heavy overlap with Meta and Google retargeting: consider consolidating or assigning each platform a distinct segment.
• Long B2B sales cycles: account-based tools may suit better, unless you use AdRoll’s B2B features specifically.
• Products bought once with no repeat purchase: cross-sell segments have little value.

None of these rule AdRoll out, but each should be a deliberate decision rather than a default left running from an old setup. Revisit the decision every quarter with the incrementality results in hand.