How smarter packaging lifted our average order value.

TL;DR

The diagnosis

One flat price was underpricing our best feature.

ArtistWorks sells one promise no lesson library can copy: a real instructor watches you play and answers you personally. That's Video Exchange. So why was almost nobody using it?

When we dug in, fewer than 15% of members had ever submitted a video. The main plan bundled personal feedback in for everyone. The 12-month plan ran $279 a year with “unlimited video submissions to your teacher” baked in. Most members were happy using us as a lesson library. But the small group who did submit videos were getting the most valuable thing we offer, personal coaching from a world-class instructor, for the same flat price as everyone else. That's the thing we were underpricing.

The members who did submit videos were our most engaged people. And we were letting them down. Instructors had a growing pile of submissions and a fixed number of hours, so feedback got slower and slower. The single most differentiated thing we do was breaking for the exact people who cared about it most.

Why the split actually works

The base plan has no human on the other end.

Once personal feedback moved to Premium, the economics of the two plans split apart.

Every new Standard member is almost pure margin, so we can grow that plan as far as marketing can take it. The cost that doesn't scale is instructor time, and it now sits behind Premium, paid for by the members who actually submit. Same feature, completely different math.

Setting the price

How we landed on the number.

I didn't want a price pulled out of the air, so I worked with our data team and ran a Van Westendorp Price Sensitivity Meter as part of a Voice of Customer study of ArtistWorks members. It asks four questions: at what price is it too expensive to consider, getting expensive but worth a thought, a great deal, or so cheap you'd doubt the quality?

Every respondent rated each price band four ways. Plotted together, the sentiment shifts cleanly as the price climbs. “Too cheap” sits at the bottom, and “too expensive” doesn't take over until you pass $251:

How members rated each annual price for an unlimited-feedback plan. “Too expensive” doesn't take over until past $251.

And when we asked them to name a single fair price, the answers clustered the same way. A clear plurality landed in the $251–500 band:

Where members placed a fair annual price for unlimited Video Exchange

Then came the judgment call.

The data team spotted a gap. Members had a clear “bargain” price and a clear “too expensive” ceiling, with open room in between, a “high but doable” band we weren't pricing into at all. So the question got simple: where in that gap do we land?

The price range members gave for each sentiment ($/yr). The gap between “bargain” and “high but doable” was the room we had. The data team recommended the bottom of it; I pushed to the top.

The data team recommended starting at the bottom of that range. I pushed to the top:

Two reasons:

1. We're a premium brand, so price like one. ArtistWorks isn't a budget library. Pricing our flagship feature at the top of the range signals that, and the people who want personal coaching read price as a proxy for quality.
2. You can always come down. Start too high and a discount fixes it. Start too low and you're stuck raising prices later, which is far harder and burns the members you already have. Start high and leave yourself room.

We launched Premium at the top of the range and watched what it did to average order value. It went up 11%. The split sent the people who valued personal instruction to a higher price and gave everyone else a cheaper plan that matched how they actually used us.

How I led it

Three things kept this from turning into a six-month debate.

1. I led with the member problem. The pitch inside the company was simple. We were underpricing the thing our most engaged members love, and slowly breaking it for them. Fix that and the revenue follows. It's also a lot easier to get a team behind a member problem than behind a margin goal.
2. I made pricing a data exercise. The Van Westendorp survey took the number off the table as a thing people could argue about by feel. We argued about the data instead.
3. I protected the instructors. Moving heavy Video Exchange usage to a tier that pays for it means we can finally staff and prioritize that queue instead of letting it quietly rot.

Then we executed:

• Ran the Voice of Customer study with the data team and built the price recommendation off the curves.
• Split the single plan into Standard and Premium, and drew the exact line between what each one gets.
• Rebuilt the plan-selection page so the reason to go Premium reads in a couple of seconds.
• Launched to production and tracked AOV.