Join Localytics' Director of Solutions Engineering Lee Isensee as he details the different key metrics for mobile and web analytics.
Hi, I'm Lee Isensee, Director of Solutions Engineering here at Localytics and thank you for joining us for today's Whiteboard Wednesday. Today we're going to be talking about web versus mobile and, specifically, the different metrics involved.
We're going to focus in on three specific topics: base KPIs, user identification and sessions.
The first thing that we really want to look at is the differences between web and mobile metrics that we typically see. One important element is the mobile world's concept of users, which is more than just a semantics issue. We'll come to that a little bit later. In contrast, the web world focuses on the concept of a visitor.
Mobile sees sessions instead of web visits. Both worlds use events, but whereas websites rely on pageviews, there's a concept of 'screens' in the mobile world which is very different than pageviews. Screens are the frame in which the user is looking at the application, not necessarily the amount of content that they viewed.
So as we look into this a little bit further beyond those base metrics we have to think ultimately about the user itself. We try to focus on an individual person as they engage with your application or your website, and mobile and web technologies have very different ways of calling them out as individuals.
In the web world of course we can use very simple methods such as user agent and IP address. Alternatively we can use more sophisticated cookies - first party or third-party - or if your web site has authentication, you can identify users via a login screen or social engagement.
In the mobile world we actually have fewer options but they tend to be a lot more accurate. In the mobile world we don't worry about user-agent, IP address and cookies, as we're able to look at the device itself and get information about it. So whatever mobile device is in the person's pocket provides us with a general idea that it's that individual device. As many mobile applications are now taking advantage of social authentication, we can also leverage that to identify a user, and this really helps bridge the gap across additional devices as well. If the person's engaging with you on the web, social authentication can make the connection between their web usage and their mobile usage.
Now with all these great opportunities to understand the user there are obviously some down sides that happen from time to time. In the web world we have to worry about roaming users; people who are on their laptops at home, a coffee shop, at work. If you're using a user agent and IP address method to identify your users, you're going to get a different one every time.
Another thing to think about is multiple browsers. In this case, beyond the different user agent, you also have the concern of different cookies.
Even with an authenticated user, we have to think about the different ways they're logging in - are we resetting that login? Are we able to persist that?
Going back to cookies, is your user clearing their cookies regularly? Users can easily remove cookies manually, and some even set up automated, scheduled cookie deletions.
In the mobile world there are similar issues. As I mentioned a moment ago a person might have multiple devices - multiple mobile phones, perhaps also a tablet. They may also have different setups at home, where they're sharing devices across users.
Like clearing cookies, mobile marketers need to be cognizant of users resetting their mobile identifiers. In the iOS world one feature announced with the switch to advertiser ID is the availability for users to reset their device identifier. It is, however, fairly well hidden - certainly less likely for someone to reset a device identifier than resetting a cookie.
Finally the last thing that we wanted to touch upon is sessions. This is more about how the user engages with your digital assets.
In the web world we understand some users may come to your website, maybe view a couple pages, get up, grab a coffee - maybe they're gone for ten, fifteen minutes. They come back, browse a couple more pages. Maybe they open up another tab and go to a different website, then come back.
Ultimately as long as their browser is pointed at your website within the traditional thirty minute timeout window, all of this activity constitutes a single, perpetual session. You're left to understand the sequence of those events in the context of one session.
I like to think of this as a "walk-away metric." How long did it take the person before they completely walked away from your site?
For mobile, it's difficult to really use thirty minutes as that cutoff point. Imagine holding your device in front of you for thirty minutes. That's really difficult to do for most people and they don't have that attention span.
When we think about the experience of the mobile user we have to ask ourselves, "is my application in front of that user, and for how long are they engaging?"
What Localytics has done, in looking at all of our clients, we found that an appropriate timeout period is not thirty minutes, but rather fifteen seconds. Once a user is out of the application for fifteen seconds, their attention is broken, and their session is over.
Really, it's about how users engage with your application. If they get an email do they quickly scroll to the email, view it, and immediately back to your application? Maybe it only took ten seconds and they haven't lost their train of thought.
However, if an email comes and they respond to it, that may take them away from your app for a minute, two minutes, five minutes, or more. When they come back to your application, they've really lost that train of thought - effectively, they're starting a new session.
So these are a couple things that you really want to think about when you're considering web vs. mobile analytics.
I'm Lee Isensee, Director of Solutions Engineering here at Localytics, and I'm glad to be able to deliver conversation today about web versus mobile metrics.
Thanks for signing up. Look for your first email shortly!
We’ll reach out shortly to schedule a time to talk.