Produced a pre-recorded wine tasting video featuring an expert reviewing 12 wines, shot on a green screen and built out with sourced imagery for immersion, accessed via QR codes on direct mail and a landing page. The campaign delivered 5x the sales of its control.
→ A/B test: QR code + tasting-video landing page vs control delivered 5x the sales.
Problem
The COVID lockdowns took every in-person wine tasting event off the calendar overnight, removing a channel the business relied on for customer engagement and acquisition, with no in-person alternative available.
Context
I produced a pre-recorded video of a wine expert giving a tasting review of each of 12 wines, shot on a green screen and built out with sourced imagery so it felt immersive rather than a static talking-head recording.
Business objective
Produce a wine tasting video campaign as a new acquisition channel, proven against a control through an A/B test rather than assumed to work.
Customer/user objective
Customers needed a genuinely engaging at-home tasting experience they could access easily, not a lesser substitute for the in-person event they could no longer attend.
Constraints
There was no existing infrastructure for producing content like this, so the video shoot, the green screen production, the sourced imagery, and the QR codes on the direct mail campaign all had to be coordinated from scratch under lockdown conditions.
Stakeholders
The wine expert presenting the tasting reviews, the production and technical delivery team, and customers accessing the video via the QR code to taste along at home.
Research and discovery
I worked out what the video needed to feel genuinely immersive rather than a flat recorded talking head, which is what led to the green screen production built out with sourced imagery.
Options considered
I considered a simple recorded video with no additional production versus a green-screen production built out with sourced imagery to make it feel more immersive. Rather than assume the more produced version would perform better, I ran it as an A/B test against a control, so the result was evidence, not a guess.
Prioritisation
Setting up the A/B test came before the campaign went out, not after, so results could be judged against a clear control rather than assessed retrospectively.
Delivery
I coordinated everything including the video production, the green screen shoot of the wine expert reviewing each of the 12 wines, sourcing and adding the imagery to make it immersive, and the QR codes on the direct mail element of the campaign, giving customers access to the video via direct mail or the landing page. I designed the full customer journey from invitation through to follow-up, built the campaign plan across email, CRM and digital to drive attendance, and A/B tested the QR code and landing page against a control group.
Decisions made
Running this as an A/B test against a control, rather than producing the video and assuming it worked, was the decision that mattered most, it meant the business got proof the production investment actually drove the extra sales.
Trade-offs
Building out the video with a green screen shoot and sourced imagery took more time and coordination than a simple recorded talking-head video would have. I judged that worthwhile given how much the immersive feel mattered to engagement, and the A/B test proved it out.
Business outcome
The A/B test compared the QR code and landing page carrying the tasting video against a control, and the test variant delivered 5x the sales of the control, giving the business clear evidence to justify continued investment in the format.
Customer outcome
Customers got an immersive tasting video reviewing all 12 wines, accessible via a QR code on their direct mail or the landing page, rather than a lesser substitute for the in-person event.
Lessons learned
Testing a new production format against a control, rather than producing it and assuming it works, is what turns a one-off campaign into evidence the business can actually act on.
What I'd improve today
I would build in a simpler repeat production process from the first video, so scaling the format for future tastings required less rebuilding each time.
