492 starts at a grocery end-cap. 91 percent watched to the end, then acted.
Dr Pepper, the Charlotte Hornets and Harris Teeter put one QR code on the side of an end-cap. It opened a video of O’havia and Hugo the Hornet that greeted shoppers by name, unlocked by pointing a phone at a ten-pack. SparkleTree built the adaptive assets; VideoBomb.co delivered them in AR. Here is what happened, and what went wrong.
Numbers measured and supplied by VideoBomb.co, which hosted the WebXR experience and gave permission to publish them.
The setup.
Dr Pepper wanted shoppers to notice a new drink and to feel the Charlotte Hornets partnership at the point of sale, in Harris Teeter, the Hornets’ hometown grocer. The usual answer is a printed end-cap and a coupon. This activation turned the end-cap into a surface a shopper could interact with, and measured every step.
SparkleTree was the engine behind the video. Hornets talent O’havia and Hugo the Hornet recorded one take in front of a greenscreen. SparkleTree composited that take onto a backdrop it created, and generated more than 900 variants of the clip, one for each name a shopper might type, so that every viewer was greeted personally. VideoBomb.co hosted the WebXR experience that delivered the right variant in AR over the ten-pack, and measured what happened.
One QR code went on the side of the Dr Pepper end-cap. That was the whole footprint in the store.
What the shopper did.
- Scan. The QR code opened a WebXR app in the phone’s browser. Nothing to install.
- Find the ten-pack. The app asked the shopper to find a Dr Pepper ten-pack and, if they liked, add their name to a rolling list on screen.
- Point the camera. Image tracking recognised the side of the ten-pack and played the shopper’s own variant over it: O’havia and Hugo the Hornet greeting them by name, inviting them to a game and introducing the new drink.
- Watch. The clip ran 30 to 35 seconds.
- Act. When it ended, the shopper got a set of calls to action: open the Hornets’ and Dr Pepper’s Facebook links, like Instagram posts, and more.
What happened.
Of the 492 starts, 91 percent watched the entire clip and then performed at least one follow-up action. That is a view-through rate measured from starts, and it is the number to remember. It is not a click-through rate on impressions, and it should not be quoted as one.
For a half-minute of branded video with no reward attached, on a phone, in a grocery aisle, a completion rate in the nineties is a different order of attention from a social feed. The shopper had chosen to scan, walked to the product and held up a phone. Having done that, almost all of them stayed to the end and did something after.
What went wrong, and what it taught us.
The number of starts was lower than the store traffic should have produced. The reason was placement: the end-cap carrying the QR code stood more than a hundred metres from the ten-packs the experience needed. Shoppers who scanned at the end-cap had nothing in front of them to point the camera at, and many did not go looking. The experience was designed for the two to be together.
The lesson is now the first check on any in-store surface SparkleTree runs: the trigger and the product have to be within arm’s reach of each other. Even with that walk, 91 percent of starts ran to the end and acted. Putting the two together is how more shoppers start.
What this looks like on SparkleTree today.
The engine that rendered 900+ personalised clips from one take is the same one that now tailors the assets and the copy for each visitor in real time, on screens, phones, print and the web. This activation is the pattern SparkleTree productises for stores, bars, venues and festivals. A QR code on an end-cap, a table, a coaster or a stage banner opens a campaign surface a person can watch and act on, with their questions answered in the brand’s voice. Scans, starts, completions, questions and follow-up actions are counted, so the brand paying for the activation gets a report and the store gets a reason to sell the next one.
The same campaign runs before the visit on social and the web, during it on screens and phones, and after it in the online store. Read how it works for brands, retail stores and festivals, or the explainer on always-on digital promoters.
Three things to take from one end-cap.
People finish what they choose to start.
A view-through of 91 percent on a half-minute clip is not a social feed number. The shopper opted in, found the product and held up a phone. Once they had done that, they stayed to the end and did something after.
The trigger and the product have to be together.
The end-cap with the QR code stood more than a hundred metres from the ten-packs the experience needed. Many shoppers who scanned had nothing to point the camera at. Starts were lower than the store’s traffic should have produced.
A surface in the store can be measured like a web page.
Scans, starts, completions and each follow-up action were counted. That is the report a brand needs to pay for the next activation, and the report most in-store promotion never produces.
Questions about the numbers.
Is 91 percent a click-through rate?
No. It is a view-through rate: of the 492 starts, 91 percent watched the entire clip and then performed at least one follow-up action, such as opening a Facebook link or liking an Instagram post. Click-through would be a fraction of all impressions, and this was measured from starts.
Who built and measured the activation?
SparkleTree built the adaptive assets: it took one greenscreen recording of O’havia and Hugo the Hornet, composited it onto a backdrop SparkleTree created, and generated more than 900 personalised variants so the clip could greet each shopper by name. VideoBomb.co hosted the WebXR experience that delivered those variants in AR and measured every step. The numbers on this page come from VideoBomb, which gave permission to publish them.
Why were only 492 starts recorded?
Because the end-cap carrying the QR code was placed more than a hundred metres from the Dr Pepper ten-packs the experience needed, and the two were meant to be together. Shoppers who scanned at the end-cap had nothing nearby to point the camera at, and many did not go looking. It is the reason placement is the first thing SparkleTree checks on any in-store surface.
Can my venue or store run something like this?
Yes. A QR code on an end-cap, a table, a poster or a stage banner opens a campaign a person can watch and act on, and that answers their questions. Every step is measured. It runs on SparkleTree for brands, retail stores, bars, venues and festivals.
Put a surface like this in your store or venue.
A 30-minute walkthrough built around your actual aisle, end-cap or room.