Bookingkit is an online booking solution and management tool for Tours and Activities providers.  As a UX / UI designer at bookingkit, I was part of the Conversion Team, focused on optimizing the conversion for the widget that vendors insert to their website, and therefore increase the number of online bookings.

Part of the challenge of this ongoing process was the high level of customization the widget has; vendors can adapt the widget to their own brand, changing the colors of all the elements, including the CTA and choosing their own fonts. In addition, bookingkit has a big variety of customers, which means that the same widget will be used in very different cases: city tours, escape rooms, yoga, cooking classes…

A/B Testing

Together with the Product Manager and a Developer, we brainstormed different hypothesis, we documented them and assigned a priority according to the impact we believed they might have and the required effort from development.

We ran tests continuously, with a special focus on mobile, where we received most of the traffic.

Success Stories

The conversion rate of the last quarter of 2018 (blue) and 2017 (orange).

Most of the hypothesis tested were success stories. The variants proved to convert better than the original process. In total, we improved the conversion by 15% in one year.  These are some of the tested hypothesis where the variant proved to convert better than the original.

Appointment selection



We wanted to simplify the way visitors select an appointment for they visit. In the original process, users need to select a starting time from a dropdown menu. Some of the vendors could have dozens of appointments starting every day. This makes the dropdown very large and hard to scan. The suggested variant was to display the starting times in a grid layout.

Hiding the Coupon Code



The coupon code input is often a conversion killer. We couldn’t remove completely the input field since it’s required to have it for the nature of the business of tours and activities where the participants very often buy vouchers that they redeem later.

The conversation team worked on the idea of building a completely different checkout for the redemption of vouchers, so the conversion on the normal checkout won’t be affected, but before spending time on such a big project we needed to prove beyond doubt that the Redemption input field was a conversion killer. That’s why we decided to run a test where we slightly hide the input field.

Adding urgency texts

We decided to add urgency texts on the detail page and the checkout. For this reason, we run a test with different variants. All the urgency texts were based on the real data, which means that some messages would display more often than others. “Three people looking at this experience” would probably display more often than “Last booked one hour ago”. Part of the challenge was to gather enough data to have a significant test result.

Since the text is dynamic, at the end of the process we decided to assign each variant to a priority and always display the highest possible.

Improving the participants’ selector



We wanted to simplify the way users select and add participants to the cart, especially on mobile. On the original set up, the user can add participants from each price category with a dropdown. Our hypothesis was that having a plus and minus would make tapping easier on mobile. In addition, we made the number an input field to make it easier for big groups.  

Small Wins

Not all the tests required big changes. Some of the tests had a positive impact just by adjusting the text. For example, “Buy a Gift Voucher” proved to convert better than the original “Buy a voucher”, or adding verbs to the CTA: “Go to Cart” instead of “To Cart”.

Success stories (with unexpected results)

As it is expected, not the 100% of our hypothesis were correct. In some tests, the original converted equal to or better than the proposed variant. This doesn’t mean that the test failed since we gather evidence that the original version was converting better and it gave us clues on what hypothesis we could test next; for this reason, we can consider the tests success stories with unexpected results.