This article was first published on November 5, 2015.

Well, it’s official. Mobile has become inescapably vital. Even way back in 2001, during the era of WAP and tiny telephone screens, we were hearing that the mobile Internet was gonna be big. It took nearly 15 years to get there, but here we are. We’ve gone from being “mobile-friendly” to “mobile-first” – and some of us are even “mobile-only”. And unless some buzz-thirsty guru declares that we should enter the “forget mobile” phase, it’s likely that websites, mobile sites and mobile apps will continue to coexist peacefully for many years. In this article, I’ll outline an approach to measuring mobile that goes beyond simply comparing overall traffic to give you a greater understanding of user behaviour. To get there, you’ll first need a holistic view of how users are interacting with your entire digital presence.


Integrated measurement of how desktop and mobile devices are used is key

Studies on device usage are multiplying. For exemple, a study by Criteo indicates that “mobile is becoming the preferred purchase device for cross-device shoppers: 4 in 10 transactions now involve multiple devices”. Another  study from AT Internet reveals that 2 of every 5 visits to French media sites are done via smartphones. The interest in these surveys speaks to the importance of comparing performance of different devices and properties. It’s key to understanding how digital behaviours are evolving.


But what about your site? Is mobile also representing a larger share of traffic to your digital properties?
I want to emphasize the importance of measuring all elements of your digital presence, ideally with a single audience measurement tool.
Companies often offer the same content on different digital properties. But a user can act using different devices and channels at different times, in different contexts. For example, actions taken on a smartphone can change or cancel actions that were taken on a PC.
It is therefore quite dangerous to evaluate the performance of a property independently of others, without comparing its performance to that of your entire digital presence. You risk coming to faulty conclusions which do not account for cross-device and cross-channel usage.
However, there’s a need to go further than just a single overall comparison of how properties are used. And that’s the core of the approach I’m going to suggest here.


The introduction of mobile has created new expectations

People now have a choice. Every day, they decide to use certain devices over others to access your services.
For what reasons does someone prefer to use your mobile site, or your mobile app, or the desktop version of your site?
For people who only have a smartphone, there is no question, as they’re limited to using their mobile phone. Same thing for people who don’t have a latest-generation smartphone – the computer remains the device to use in order to access the Internet. When on the go, we might naturally use our mobile phones to take care of something immediately. But what about when we’re at home and have access to both a mobile phone and a computer? What might make us choose to use our smartphone over our computer?
These questions are essential. And their answers are complex.


To find these answers, we must no longer consider mobile as just another device. Mobile has given rise to new behaviours… new behaviours that would be impossible without mobile. As observed by Thomas Husson, vice president and principal analyst at Forrester Research, Inc., during his talk at AT Internet’s 2015 Digital Analytics Forum, “Mobile is much more than just a channel – it’s a revolution […] We’re now referring to ‘mobile moments’ in certain contexts.”
With mobile, people are focused on the task at hand. We no longer need to wait until we’re home or at work, behind a desktop computer. It’s immediacy that takes precedence, anytime, anywhere.
We’re also hearing more and more about micro-moments. These are short moments, lasting just a few minutes, where we’re able to quickly accomplish a task thanks to our mobile phones. There are 4 main types of micro-moments: “I want to do something”, “I want to buy something”, “I want to go somewhere” and “I want to know something”. In this way, mobile is transforming our behaviour by creating the need to accomplish things instantly, on-the-spot, as soon as they pop into our heads… while commuting to work, while waiting or taking a break, while with friends, or while browsing through a store. This approach forces us to examine the context and the events that trigger the desire to accomplish a given task.

This view of mobile as a catalyst for new behaviours colours how we approach the measurement of mobile performance.
Rather than only measuring different devices’ overall performance, it’s more strategic to also examine performance according to the task to be accomplished. By going down one more level – measuring tasks at hand – we can shed light on user behaviours and improve our understanding of their expectations.


The “task-centric” approach: measurement based on the task at hand

This involves evaluating the performance of each device based on indicators related to the successful completion of a task. This approach has the merit of measuring the specific task, and not just overall usage.
Each site or app offers several “calls-to-action”. We often lose sight of this, but a site does not boil down to just one action. For example, an e-commerce site’s function is not limited to purchases alone, even if this is the main objective.
To explain further, here are some examples of tasks that can be accomplished on an e-commerce site:

– Purchase
– Check on delivery status
– Reserve a product
– Check product availability in store
– Find the closest store
– Read customer product reviews
– Get a promo code
– See store hours

The objective of this approach is to know which tasks users undertake most on mobile. This will allow us to identify the tasks for which the mobile site is not satisfactory for users, or discover if certain devices are more conducive to certain actions.

There are 3 steps to this analysis:

– Make a list of all tasks that can be done on your different sites or applications.

– Check the right tracking of all online actions confirming the completion of a task. In the large majority of cases, completion of a task will be measured by the loading of a page, or a click on a link. If there are no possible measures for knowing if a task has been completed, we must implement the required tag.
– Extract analytics reports from our web analytics tool.


For a comprehensive view of user task completion, the following 3 analyses are extremely helpful:

  • Breakdown of tasks completed by channel and device:

breakdown of tasks by channel and device

This report shows the breakdown of usage of each device and channel, based on the task at hand.
You’ll notice that 30% of people who made a purchase during the month did so using the mobile application.
The report indicates users’ preferred channels for performing each task. We can therefore consider that these channels and devices are the most practical for them to use in each case.
If someone finds a certain device to be most practical, does that mean s/he will not use any other devices at all? The rate of exclusivity tells us. In our table, 8% of customers buy exclusively on the mobile application. These customers adopted the app and feel that it helps them save time when making a purchase; they won’t use another channel to buy. These are surely very loyal customers with a high lifetime value. They should therefore be treated with much care.


  • Evolution of this breakdown over time

Evolution of task completion by device and channel

This report shows the evolution of task completion over time.
You’ll notice that between Q1 2014 and Q2 2015, the share of computers being used to reserve a product has dropped by 13 points, from 64% to 51%.
This report reveals high and low trends on each device/channel. These trends reflect the underlying behavioural trend of people increasingly using new devices to accomplish new tasks, based on their equipment and their maturity.
This report also allows us to evaluate the impact of certain optimisation initiatives: creation of the mobile site, transferring all content to mobile, shifting to responsive design, adapting content to a mobile version,…


  • Breakdown of task completion within each device and channel

task-completion tab evolution

This report shows the ranking of tasks most often done using each device and channel.
You’ll notice that 6% of people who accessed the mobile app ended up using it to make a purchase. 23% of people who went on the mobile site did so to find the closest store.
This report enables us to answer the following question: What tasks are completed on each device/channel? It reveals which devices and channels are most efficient for each task, and for which tasks each device and channel is failing to appeal to users.

This “task-centric” approach to mobile performance measurement paves the way for better understanding of user expectations regarding mobile. This approach helps us verify that our usage of each channel and digital property is aligned with our objectives, and enables us to focus our actions on tasks and prioritise new initiatives based on results.


Want to know more? For tips on building an effective multi-device strategy, check out this guide from AT Internet: The Guide to Mobile App Analytics: From Tagging to Insights




Head of Client Success – Generaleads Benoit has a master’s degree in Economics from the University of Bordeaux and 10 years of web analytics experience developed while at AT Internet. In early 2015, Benoit joined the Google AdWords specialist agency GENERALEADS as Head of Client Success. In parallel, he’s working on the start-up GetLandy, the first landing page creation tool designed for traffic managers.

Comments are closed.