Reciprocation effect or how we increased purchase conversion 2x times

The last time I told you about the hack that helped us improve our purchase value. But as you might remember we had no changes in purchase conversion, so we were struggling to improve it, since 1.9% purchase conversion was not acceptable for our user economics.

I was reading Cialdini’s book “Influence” when a new idea came to my mind: Mr. Cialdini was explaining the reciprocation effect (“the old give and take.. and take” as Cialdini calls it) when people feel obliged to do something for you if you do something for them. The idea was simple: to gift our users the premium features for the 1st flight for free, so they could try out the premium features and, probably, decide to purchase one of the packs. We decided to implement a quick test: allow users to apply for promo code giving out their email for exchange. This way we would have more levers to retain users, while they get the premium features for free… We were expecting the purchase conversion to increase…

But in the short term this decreased our purchase conversion, since those interested in premium features used the promo code and didn’t purchase pack. Thanks God we expected this short term effect and being system thinkers decided to wait longer. And as it turned out in the longer term their conversion was 7% with the average days to purchase about 28 days!

We were excited BUT only 10% of people noticed this “Apply for promocode” button at the bottom of Purchase screen, so the overall conversion has not increased much. So we asked us an obvious question: how to make it so 100% of users get the promocode and try out the premium features.

Any ideas how we did it? Yep, exactly, we just gave out premium features for 1st flight right from the start! No special buttons, no barriers — you just get it the moment you add the 1st flight. We expected the overall purchase conversion would be around 7%, but in reality it hit only 4%. Anyway increase from 1.9 to 4 was a big achievement for us!

P.S. Any ideas why the purchase conversion is not 7%, but only 4%? What was wrong with the initial assumption?