186% increase in orders - PragueOnSegway conversion rate optimization

PragueOnSegway is a company that does private tour guides on a Segway in Prague (as the name suggests). According to TripAdvisor, they are the number 1 activity in Prague, 6 years in a row.

The business need

PragueOnSegway came to us with a request for a website redesign. The competition grew 4 times since last year and the client suspected that his web was not performing to its best in terms of both revenue and usability.

Old web of PragueOnSegway

Old homepage with a slider

The challenge

The web had a lot of usability issues such as complicated order process, too many call to actions, lack of social proof. On top of that, it was coded in a table layout which made both the download time and updates slower. The traffic was not such that we could test different elements, so an overall redesign with best practices in mind (and follow-up testing) was in order.

The solution

Complete redesigns are always tricky. You are basically going with best practices, quantitative research and your gut feelings. None the less with the right process in place it can be done so that the conversion rate raises instead of lowers. After some initial ideas and research we proceeded with information architecture and basic wireframes. For smaller sites it is far easier to just draw them than use a software such as balsamiq or axure. Also clients love hand drawings :). At least these ones did, because they chose a hand drawn icons for their tours.


skinzone homepage wireframe

homepage wireframe

skinzone category wireframe

tours wireframe



new homepage design

tours page

new categories page

Split (A/B) testing the order process

The order process was the first thing we tested. The control (version A) had 13 form fields. After about a month of pushing for this and persuading the client I was able to reduce the number of fields to 3. My clients' argument was that certain fields would disqualify people not ready and interested (hotel name) and that some information is useful for the guides so they can prepare in advance (country, both names).

(Dis)appearing comment box

The other problem was that when someone filled in the comment box, the order was not automatically confirmed, but had to wait for one of the owners to respond to and confirm by hand. And since the tourist business is very competitive, sometimes by the time the confirmation was received the users had already booked somewhere else. The comment box was important so that some arrangements could be made, such as 3 people going on a tour or a request for a specific tour guide. But after analysing the content of the comment box, most people that filled it in said: We are looking forward to visiting Prague...

The solution came to me after watching Luke Wroblewski's video on form design where some form fields were hidden behind a link. And that is exactly what I did. I also put in a "receipt" of the tour details next to the order form so users would not have to scroll back to the top to see what exactly did they order.

version A (control)

order form - version A

version B (challenger)

simplified order form - version B

42% lift in orders

Results of the AB test proved my hypothesis that a simpler form would lead to more orders. The clients kept their eye on cancellations to see if the simpler form made more people order and then cancel, but that proved insignificant.

Version A conversion rate was: 5.33%.
Version B (simpler form) conversion rate was: 13.74%.

Revenue lift

The increase in orders was not that dramatic since the competition increased considerably - from 3 to 19 competitors within a year. None the less I was able to improve the number of orders by 42% and the conversion rate by 34% within a few months.

The decline in order volume is due to number of tourists that peaks in summer and weather conditions. Only a few people like to go on a tour in freezing temperatures.

814% increase in traffic

42% increase in orders.

More visitors use a mobile device than a desktop

Over the past year I have seen a huge increase in mobile and tablet visits. Last year the percentage of mobile and tablet using visitors was 70%, this year it is 40%. So the majority of visitors are now from mobile devices.

A second reason to design a responsive version is conversion rate. It is naturally lower for mobile and to a certain extent tablet, but having a responsive site that looks nice in all devices usually increases the conversion rate as well.

Responsive wireframes

It is best to design for 4-5 different scenarios.

  1. mobile portrait,
  2. mobile landscape,
  3. tablet portrait,
  4. tablet landscape / notebook
  5. large monitor (full-HD).

Always check your analytics for your site before deciding. I chose to not go for bigger monitors as the number of visitors using them was very small.

Tour page wireframes and designs

wireframes and designs of the Tour pages

wireframes and designs of the Tour pages

64% increase in mobile conversion rate

Since the responsive version launch, the stats have been phenomenal. Compared to the previous non responsive period, there was a 38% increase in orders and a 43% increase in conversion rate. Cart abandonment went down by 10% as well.

43% increase in orders.

43% increase in orders.

425% increase in mobile orders compared to last year

Compared to the same period last year, there was a 425% increase in orders and a 97% increase in conversion rate for the mobile segment. The overall stats look good as well.

814% increase in traffic

425% increase in mobile orders compared to last year