If you’ve searched “does product video increase AOV,” you’ve probably landed on the same number we did: 57% of retailers say video pushed AOV up by 50% or more. It’s real. It’s also from 2015, and it shows up in dozens of posts today with no date attached.
We went looking for what’s actually been published since. The honest answer is more useful than the borrowed one: the literature on video’s specific effect on AOV is thinner than marketers assume, but the mechanism behind it, and the data on adjacent metrics like engagement and returns, is solid enough to build a real test plan on. That’s what this guide covers, step by step, for Shopify merchants deciding whether product video is worth adding to your PDPs and how to know if it’s actually working.
Our finding: Almost every “video boosts AOV by X%” statistic circulating in shoppable video marketing content traces to one of two places: the 2015 Liveclicker survey, or a cluster of unsourced numbers (73%, 144%, 37%, “80% better”) that appear on content-mill roundups with no named study, sample size, or link back to primary research anywhere in the chain. We couldn’t verify any of the second group. Treat them the way you’d treat any statistic with no source: as marketing copy, not evidence.
Step 1
What Does the Actual Data Say About Video and AOV?
By the end of this step, you’ll know which video statistics are worth citing to your team and which ones to drop from the deck.
In 2015, a retailer survey run by Liveclicker found that stores with video on 50-75% of their product pages averaged roughly $300 in AOV, versus about $177 for stores with video on 0-25% of pages. 57% of respondents reported AOV gains of 50% or more (Liveclicker, via MarTech, 2015). It’s the single most-repeated AOV-specific figure in the category. It’s also the only one: no independently replicated, methodology-disclosed study since has isolated video’s effect on AOV the same way.
What’s been published more recently measures adjacent things. In 2026, Baymard Institute’s large-scale product-page usability testing found 41% of shoppers interacted with or watched product video when it was present, but 35% of major ecommerce sites place video so poorly, buried in tabs, below the fold, or missing a play icon, that shoppers never get the chance (Baymard Institute, “Embedding Product Page Videos”, 2026). That’s an engagement ceiling, not a conversion or AOV number, but it matters: video can’t move a metric it’s never seen.
Vendor-authored reports fill some of the gap, with the usual caveat that the vendor sells the product being measured. Firework’s 2025 State of Video Commerce report, drawn from 600+ enterprise brands, claims shoppable video drives 5x higher add-to-cart rates than traditional PDPs (Firework, “State of Video Commerce 2025”, 2025). It also claims video-commerce leaders generate 43% of total GMV through video and a 275% lift in repeat purchases. Directionally useful, not independently audited.
Step 2
How Does Video Actually Move AOV, If It Does?
By the end of this step, you’ll understand the three mechanisms worth testing on your own store, instead of hoping a borrowed percentage applies to you.
Video doesn’t raise AOV by magic; it raises it (when it does) through a small number of specific mechanisms, and knowing which one you’re testing changes what you should measure.
Cross-sell and bundle discovery. A shoppable video that tags a full outfit, a bundle, or a “goes with this” companion product gives a shopper a reason to add a second item they weren’t already looking for. A static image gallery for one SKU can’t do that; a video showing the product styled or used alongside others can. This is the mechanism closest to what the 2015 AOV data was likely capturing, even though that study didn’t isolate it.
Better fit and use-case information, fewer surprise returns. U.S. ecommerce return rates run 19-20.5% overall, with apparel averaging 25% (shoes 31.4%) and electronics around 11% (Eightx, citing Statista Consumer Insights and Capital One Shopping data, 2026). Video that shows scale, fabric drape, or a product in use sets expectations a photo can’t, which should reduce mismatched-expectation returns. No study isolates video’s causal effect on returns specifically; the widely repeated “video cuts returns 35%” figure traces only to unsourced content-mill pages, so we’re not citing it as fact. Treat your own realized AOV (after returns, not gross) as the metric that matters, since a return silently erases whatever lift you thought you got at checkout.
Purchase confidence at a higher price point. Wyzowl’s 2026 survey found 85% of consumers say video has convinced them to buy a product, and 83% of video marketers say video has directly increased sales (Wyzowl, “Video Marketing Statistics 2026”, 2026). That’s a general video and purchase-intent stat, not AOV-specific or product-page-specific, but it supports the idea that video reduces hesitation, which matters more on a $180 item than a $20 one.
What we found: Across Whatmore’s own shoppable video customers, the AOV gains that show up in published case studies tend to cluster around brands that used video for cross-sell and bundle tagging specifically, not just as a passive PDP asset. Nasher Miles shipped 130+ shoppable videos across 1,300+ tagged products and reported a 65% AOV uplift. Philips India reported a 20% AOV increase after adding shoppable video. Beauty by Bie saw AOV rise 12% within 30 days. These are first-party, brand-specific results, not a controlled experiment against a matched control group, so treat them as evidence the mechanism works, not proof of a universal multiplier.
Step 3
Where Does Video Placement Actually Change Basket Size?
By the end of this step, you’ll know which surfaces to prioritize before you build out a full-catalog video program.
Not every video placement has the same shot at moving AOV. A hero video on a homepage builds brand awareness but rarely touches basket size directly. The two placements with a real, testable line to AOV are the product detail page itself and post-add-to-cart or cart-page cross-sell moments.
On the PDP, video that shows a product styled with complementary items, or tagged with a “frequently bought together” set, gives the shopper the closest thing to an in-store associate’s cross-sell nudge. See winning product detail page for the broader PDP layout principles this sits inside.
| Placement | Primary lever | What to measure |
|---|---|---|
| PDP, single-product video | Purchase confidence, fewer mismatched-expectation returns | Conversion rate, realized AOV after returns |
| PDP, cross-sell/bundle-tagged video | Basket size directly, via a second tagged item | Items per order, AOV at checkout |
| Cart or post-ATC video | Last-chance upsell before checkout | Attach rate, incremental revenue per session |
| Homepage or collection hero | Awareness, brand fit | Session-to-PDP click-through, not AOV directly |
Prioritize the second and third rows if AOV specifically, not just conversion rate, is the metric you’re trying to move. A single-product PDP video is still worth having, but it’s aimed at closing the sale, not growing the basket.
Step 4
How Do You Set Up a Test That Actually Proves a Video-Driven AOV Lift?
By the end of this step, you’ll have a test design that won’t fall apart the moment someone on your team asks “but how do we know it wasn’t just seasonality?”
Pick your 5-10 highest-traffic PDPs and record baseline conversion rate, add-to-cart rate, and realized AOV (after returns, not gross checkout AOV) over the prior 30 days. Add cross-sell-tagged shoppable video to those pages, leaving the existing image gallery in place; this is additive, not a replacement. Run the test for a full 30-day cycle minimum to smooth out day-of-week and traffic-source noise, and don’t stop the test the moment it first crosses statistical significance.
That last point matters more than it sounds. Stopping an A/B test as soon as it first crosses the standard 95% confidence threshold, instead of committing to a pre-set sample size upfront, can inflate the real false-positive rate by 5-10x over the nominal 5% you think you’re getting (Johari, Koomen, Pekelis, and Walsh, “Peeking at A/B Tests,” KDD 2017). In plain terms: if you check your dashboard daily and declare victory the first time the number looks good, there’s a real chance you’re looking at noise, not a real lift.
- Set your success metric before launch: realized AOV after returns, not gross checkout AOV.
- Commit to a minimum sample size and test duration before you look at results, not after. Thirty days is a reasonable floor.
- Track cross-sell-tagged video separately from single-product video, since they test different mechanisms.
- Wait out the full committed window, even if an early read looks good.
- Compare against your recorded baseline from the same PDPs over the prior 30 days, not against an industry benchmark you found in a blog post.
For the deeper statistical traps specific to shoppable video testing, see is your shoppable video lift real?, and for the broader benchmark context across the category, see the shoppable video benchmarks 2026 report.
Step 5
What Tools Do You Need to Run This Test on Shopify?
By the end of this step, you’ll know what to look for in a shoppable video app if AOV, specifically, is the metric you’re optimizing for, not just conversion rate or watch time.
Not every shoppable video app reports the metric you need. A plain video embed gives you play rate and watch time, engagement signals that don’t tell you whether the video sold anything. What this test needs is a platform that tags individual products inside the video, so a shopper can add a second, cross-sold item without leaving the player, and reports video-attributed add-to-carts and AOV separately from the rest of your PDP traffic.
Whatmore’s product page video and video insights features are built around exactly that: cross-sell tagging within a clip, plus reporting that isolates video-attributed AOV from the rest of the page. Pricing runs on a free tier up through plans priced by monthly video clicks rather than raw traffic; check the current pricing page for exact tiers. If you’re comparing platforms on tagging depth and AOV reporting, see Whatmore vs. Vimotia.
Where teams go wrong
Common Mistakes When Measuring Video’s Effect on AOV
1. Citing the 2015 Liveclicker number as if it’s current. It’s the most-repeated AOV statistic in the category and the least fresh. Use it with its date attached, or don’t use it.
2. Measuring gross checkout AOV instead of realized AOV after returns. A video that sells a second item that gets returned two weeks later didn’t actually move your AOV; it moved your return rate. Track realized AOV, not the number at checkout.
3. Testing a passive product video and calling it a cross-sell test. A single-product video without cross-sell tags can support conversion and purchase confidence, but it has no mechanism to move basket size. If AOV is the metric, the video needs to tag more than one product.
4. Peeking at results daily, or crediting video for a lift that came from elsewhere. Stopping the moment the number looks good can inflate your real false-positive rate by 5-10x over the nominal 5% (Step 4). Just as common: crediting video for a lift that actually came from a promotion or seasonal shift running in the same window. Isolate the test to a fixed set of PDPs and a fixed window, with nothing else changing on those pages.
Next steps
Next Steps
The honest version of “does product video increase AOV” is smaller than the recycled 2015 statistic, but it’s actionable: the mechanism (cross-sell tagging, better fit information, purchase confidence) is real, the adjacent engagement data holds up, and Shopify brands like Nasher Miles, Philips India, and Beauty by Bie have published genuine AOV gains after adding shoppable video. The only way to know what it does for your catalog is to run the 30-day test in Step 4. If you’re ready to try it, see how Whatmore’s shoppable video platform tags products for cross-sell and reports AOV separately from the rest of your PDP traffic.
Ready to install? Try Whatmore free on your Shopify store — no credit card required.
Book a Demo →
