Pocket Diner is a content-managed mobile website solution for restaurants that we created back in 2011. It’s an interesting cross section of over 1400 restaurant sites, from single location independents to some large chains.
We thought it would be interesting to look at and share this anonomised data, to see if there are any useful insights into the browsing habits of restaurant customers on a mobile device.
So, the headlines…
- Total page views since launch: 133 million
- Total unique visitors: 14 million (that’s about 21% of the UK population!)
- Monthly unique visitors: 741K
- Monthly table bookings: 23K / 82K covers
- Average conversion rate: 3.1%
OK, so they are some pretty big numbers considering that's purely mobile device traffic!
Let’s look a little more in depth…
Perhaps one of the most surprising stats is the number of people using a phone on a mobile network, so 3G, 4G, GPRS, Edge etc. I usually use my phone to browse when connected to Wifi (at home or the office) and looking at the stats of our non-restaurant clients that’s fairly typical behaviour, with only 10-15% of mobile device traffic coming via a mobile network.
However a massive 51% of Pocket Diner users are on a mobile network, and therefore relatively slow network speeds. When we work those numbers back into total traffic stats it shows the difference between restaurant customers' mobile use, which tends to be higher than average anyway, and most other sites (for the purposes of 'other sites' we are looking at averages of our other clients that aren't in the restaurant sector)
We're treating customers booking a table as a conversion. These can vary quite considerably between restaurants (the 3% is an average over all of them) but looking at our own clients where we can check against desktop stats, we see a very similar conversion rate between desktop and mobile for those that use Pocket Diner. For those without any sort of mobile optimised solution, we see a 80-90% drop in the conversion rate on mobile.
This is quite interesting to compare to a desktop restaurant site, where, excluding traffic driven by marketing activity, daily visits are quite consistent with no real spike and often a slight dip at the weekend. With Pocket Diner we see almost an inverse graph with a significant peak on a Saturday, where people tend to be out and about shopping and looking for somewhere to eat.
5-6pm is the busiest time no matter what day, with a smaller spike at 12:00
There is a consistently higher proportion of women compared to men (72% vs 28%) suggesting that women take the lead in researching where to eat and booking tables.
The age bracket 25-34 accounts for nearly 50% of all traffic. There is only 14% of traffic from 18-24, and only 5% over 55.
Pocket Diner has a pretty straight forward navigational path, however it can be difficult to analyze due to restaurant price points, which means you have to choose a location before you can see a menu with some brands.
That said, if we include people that click a location and then go straight to a menu, we see that a total of 70% of all users are clicking through to the menus.
For bookings, 87% that booked a table did so from clicking ‘Book a table’ as their very first click.
Devices & browsers
The iPhone is still by far the most popular phone hitting our server. This seems at odds with sales figures, where, depending on the report you read, Android are either on a par, or in front of Apple.
When looking back over trends over the last couple of years, iPhone has actually remained pretty static, with the only notable changes being a 10% switch from Android to Chrome, probably due to Samsung making it the default browser.
The dominance of iPhone and Samsung does make you wonder how on earth the other phone makers can ever catch up, and raises the question of is it worth (in terms of ROI) optimising a mobile site for anything other than Safari and Chrome.
We spent quite a lot of time creating a basic version of Pocket Diner for customers not using a smartphone. Looking at the stats even from launch, this was probably wasted effort...although there was one person out there in the last month browsing on a Nokia N97 who hopefully appreciated it!
- Safari - 63%
- Chrome - 22%
- Android - 10%
- IE - 2%
- iPhone - 65%
- Samsung Galaxy - 18%
- Sony - 3%
- HTC - 2%
- Nokia Lumia - 2%
- Blackberry - 1%
Hopefully this will be of use for others when looking at the mobile strategy for restaurants. The most interesting discovery for us is the high volume of people on slower connections. That really does mean that a 'cookie cutter' responsive site, where the same content (and page weight) is served up to all devices may not be the best approach (unless you have very minimilist branding!). Instead, a great deal of thought needs to go into server side detection and optimisation, both in terms of file sizes and customer journey, to ensure that the customer experience, and therefore table bookings, isn't compromised.
It also seems to be the case that customers are more task-based on their phones. Whilst on desktop sites people might browse around a little and then book, on mobile they head straight for the booking link. Also the timing suggests that people tend to use their phones in the afternoon and evening, with a sharp peak around 5pm, whereas desktop useage is a much flatter graph throughout the day.
There are a lot of similarities with desktop users though - menus still receive the largest amount of traffic and outright conversion rates are similar, so the challenge is how to create a great user experience on all devices, and crucially for restaurants, at all network speeds!
Google have recently announced that they are going to start adding ‘mobile friendly’ tags to search results, based on a number of set criteria. There’s a tool on the site to check if a site complies, and we’re delighted to see that it follows detection redirects, both server side and client side, meaning that Pocket Diner restaurants will all receive this tag :)