Booking Fails To Replicate Airbnb's Signup, This Experiment Suggests

Booking shifted their headline in this experiment from: (A) creating a listing to the (B) discovery of how much one could make from listing their home. This seems like a good direction, however Booking might have also mistakenly "delivered" on their promise right on the same screen. Notice how they actually provide the answer to that question by showing a dollar amount.  

Comparing this to Airbnb, I'm going to make a hypothesis that Booking's headline is rather shallow. Airbnb on the other hand makes a deeper promise of actually earning money. Comparatively, the value is demonstrated but not delivered just yet (it requires more action first).

In the variant (B), Booking then shows a city field which plays a role in affecting the promised dollar amount that a landlord could generate on a monthly basis. And then, Booking also asks potential landlords how much they would like to charge their guests per night. I suspect this to be the root cause of why the experiment might have broke down for two reasons:

1) Showing the day rate ($100) and the monthly earning potential ($1500) together might hint at questions. If someone does some rapid math ($100 day rate x 30 days = $3000), then this simple calculation might make it seem that Booking takes 50% of the revenue. This of course is not true, but it could be subtly implied.

2) More so, asking users about their desired day rates and then using that to project monthly income, doesn't feel like users are being educated about their potential. Instead, this feels like the calculation itself is really on the user to figure out on her/his own. 

Comparing this to Airbnb, they ask about 3 critical factors that influence the projected price: location, the type of apartment (room, full, shared, etc) and number of guests. Using this, they provide a more realistic price range. Airbnb educates potential landlords, without raising eye brows over how big of a cut they will take. To me thats's a solid approach.

Here I just wanted to criticize the quality of the implementation - orphaned elements. We can see two unnecessary and wasteful line breaks before elements that could be tighter (the "?" tooltip, and the word "earning"). Such orphaned elements may hint at a subtle sloppiness to some picky designers, but more specifically they may also push the call to action lower. Could this have played some microscopic role in the measured effect?  

Finally, in the B variation, Booking swapped the first price discoverability component with the next step of the registration process - containing the name and email fields. From what I remember about the experiment, there might not have been any way to move back (if someone wished to adjust anything for whatever reason). Such an inline change also raises questions about "perceived progress" - do users really see that they went to step 2 after submitting a form? Not sure.

What we do know comparatively, is that Airbnb uses a clearer and more dramatic modal for their next step. This might have the added benefit of 1) being able to collapse ihe model and move back a step, as well as 2) hinting about progress more visibly. Here is what the step 2 modal looks like on their site: