It seems all technology is getting smaller and more efficient. It’s certainly true for computers, as smartphones are progressively overtaking their larger counterparts.
According to Dazeinfo research, there were about 1.13 billion smartphone users in 2012. This number increased by 27.1% in 2013 to 1.43 billion, and by 2017, nearly half of global mobile users are likely to own a smartphone.
Band-aiding a mobile experience is no longer a possible solution, as 70% of mobile searches lead to action on a website within 1 hour of searching.
Marketers need to face mobile optimization head on by identifying the barriers that prevent your users from converting and taking action against them.
From here, you can run tests and use those results to strategize the best solutions for your customers.
To find what barriers your mobile website may face, let’s look at the limitations of mobile, how to compensate for them, and what improvements you can implement to ensure the best user experience, increase conversions, better retarget potential customers, and more.
The Mobile Gap & The Problems It Causes
Coined by Jason Goldberg over at Razorfish, the mobile gap is the time between consuming media on mobile devices (which is relatively high) and the amount of time, effort, and money brands spend working on and optimizing mobile (which is relatively low).
This “gap” first came to fruition in 2009 when retailers decided that the best way to attract mobile users was to create a mobile application.
While apps proved to be sufficient for some brands, surveys later showed that users preferred to use a mobile site rather than an application. In fact, a survey by Quiex states that 49.7% of the respondents dislike using mobile apps. This is because 26.1% feel they eat away at device storage while 23.6% of them feel that they have slow and inconsistent performance.
Although this gap is decreasing due to the rise of mobile website optimization, conversion rates for these websites are still poor when compared to their desktop counterparts.
Before digging into the tactics to help increase these conversion rates on mobile, let’s look at the problem areas that were created by the mobile gap:
- Moneyball Syndrome
- Cross-Device Attribution
Inspired by the movie ‘Moneyball,’ MoneyBall syndrome refers to the fact that companies aren’t looking into the right metrics when measuring the effectiveness of their mobile website.
Some marketers try looking at conversion as a solo metric of determining how their CTAs perform on mobile. But conversion rate alone is not a success metric.
Marketers should also be looking at acquisition (visits, unique visitors, pageviews), behavior (pages per visit, time on site, bounce rate), and revenue (if you’re eCommerce). Once this data is acquired you can compare the metrics against desktop behaviors to see how user interactions and behavior change on different devices.
Also realize that some users visit the site without the intention to purchase or convert. These users may be visiting to check inventory, look at reviews, read blog articles, or even compare pricing. This ultimately helps skew the overall conversion rate.
Bottom line: recognize that conversion rate can’t be looked as a stand-alone metric. One must also take into account user behavior, user intentions, and comparisons to desktop.
According to Criteo, 40% of eCommerce transactions involve multiple devices along the path to purchase. This is what cross-device attribution refers to; users who initially view products on one device to research them, only to eventually purchase them or convert later on another device.
(e.g. Someone viewing a shirt on mobile, but purchases it later on desktop).
Although the same person might have visited the page twice, the fact that it’s on a different device makes it so it looks like two unique visitors have visited the page.
The first device it was viewed on gains no credit for that purchase and marketers see it as a bounce or a missed conversion opportunity. This, too, results in skewed data.
For example, according to Mona Elesseily, “conversions in the health vertical have the best ROI during work hours and into the evening on tablet devices.”
This means companies in the health vertical need to recognize that the majority of its users are coming from a mobile device. This might be because they’d prefer viewing their personal medical information while relaxing at home rather than worrying about coworkers seeing it.
Using a unique identifier or a compilation of specific demographic and behavioral data, we can begin to accurately track the same user on the variety of the devices they use. While this technology is still relatively new, it is slowly becoming more refined for the future.
Overcoming The Mobile Gap: Advanced Mobile Optimization Strategies to Boost Conversions
Despite the challenges the mobile gap creates, marketers can overcome these obstacles.
This not only involves making visual website tweaks, but also analyzing your users’ behavior on mobile, the reasons they come to your mobile website, and how they typically behave and interact with it.
1. Revamp Mobile Metrics to Combat Moneyball
A study by Econsultancy and Adobe found 49% of the client-side marketers and 53% of agency executives said they did not measure mobile user engagement and mobile return on investment (ROI), let alone tackle attribution. This means marketers and executives have no way of knowing what their mobile users are actually doing on their website.
To start, both parties need to look at mobile usage separately and develop a more detailed variety of analytics for their mobile campaigns.
Some analytics and metrics you should begin measuring to help capture a greater range of data are:
- Behavioral Cohorting
- Session Lengths and Drop-off
- Event Tracking
1. Behavioral Cohorting
Behavioral cohorting is defining a group of users based on actions they have or have not taken with your product. These behaviors can be useful in investigating how different behaviors impact retention and churn, funnel conversion rates, and revenue.
Think Facebook’s 7 Friends in 10 Days correlation. It’s all about teasing out behaviors that correlate with higher conversions, retention, or whatever metric you’re optimizing.
Once you define your cohort, you can begin comparing how different variations of it affect user behavior and their retention. To delve deeper, you can analyze how these behaviors vary from each device.
Say your behavioral cohort is all users who have read reviews prior to purchasing a product on mobile. You might be interested in analyzing if users who read reviews:
- Have a higher conversion rate than those users who don’t read reviews.
- Are the users more engaged – longer sessions, more time in app, fewer drop-offs.
You can then begin testing to see if users who visited your testimonials or reviews page ended up eventually making a purchase.
To help start your analysis, Amplitude has created a worksheet that helps you define your behavioral cohort and the impact their behavior has on your core metrics. Using this can help you monitor what your users are doing to help you form hypotheses to improve retention.
2. Session Lengths and Drop-off
Sessions are a basic unit of measurement to monitor a user’s engagement (traditionally within apps).
In the case of your website, a session is counted every time a person views a specific page you are monitoring.
Analyzing session length, or the amount of time the website or app is being used, can be helpful in finding mobile trends, flaws, and ultimately, how to improve users experience.
You primarily want to focus on the average session duration of your mobile users in combination with your conversion rate.
For example, you may find users are spending a significant amount of time on your pages, but aren’t converting. Are they debating making a purchase or researching? Are they having trouble using the page? Maybe they can’t find what they’re looking for.
You can combine this data with tools like HotJar or MouseFlow, that allow you to 1) see heatmaps of where users click and move on your page and 2) watch actual users interact with the page..
Of course, all of this is just data that you need to inform A/B test hypotheses. And once you set up tests, you can analyze your sessions length vs. conversions. You may find session lengths and conversions both increased, or decreased sessions lengths with no change in conversions rate (great SEO, but issues when converting).
From here, we’ll look into how using you can using event tracking will help you better analyze how users are spending their sessions on your mobile app/website and how to find where issues within it are.
3. Event Tracking
Tracking any action that is important in converting on mobile gives you insight into how effective they are and offers insight into your users’ behavior on-the-go.
Initially, you want to focus on events that are most critical to your core conversion areas, such as consultation CTAs on your success stories page, or how users are interacting with your MOFU (middle-of-the-funnel) tools.
Custom events simply allow you to see user behavior more granularly.
Google Analytics, HubSpot, and Appsee are great tools to help you set up event tracking so you can ultimately discern between which elements aid and hinder your user experience to help you increase your bottom line.
Track no more than 30 events for a website, and 200 events for an app, so you have the ability to strategically monitor what’s happening and draw conclusions on the behaviors relatively easily.
2. Solving Cross-Device Attribution
The best way to start doing this is with Cross-Device Attribution or Identity –the ability to know that a person who uses smartphone X is the same person using tablet Y and desktop Z.
Think of it like this: say you found a pair of shoes online on your mobile phone. You liked them, but decided not to purchase them. When you arrive home, you check the shoes again on your tablet, but still avoid buying them. Finally, you cave in and buy those shoes on your desktop computer the next day.
Due to the change of device, marketers can’t tell that you visited the site three times on three different devices. Instead, they see it as three unique users on three different devices, but the one on desktop converted, while the other two people didn’t.
Now imagine another 100 people doing the same thing. Marketers are left with analytics that don’t match up with the actual amount of users on their site. This is why understanding which user is the same across multiple devices is so important.
Once we are able to make that distinction, we can begin to segment these users and deliver specific messaging, layouts, or retargeting campaigns that best resonate with buying habits on those devices.
This article recommends many solutions for cross-device attribution.
Revising Our Approach to Tracking Customers Across Multiple Devices
There are 3 steps marketers can follow to start connecting specific users’ mobile and desktop behavior:
- Aggregate audience data
- Develop individual personas based on the data
- Create campaigns that connect with these personas
1. Aggregate Audience Data
To get started with solving cross-device attribution, you’re going to need to collect all your non-PII (personally identifiable information) data so you can segment your audience into appropriate personas.
You’ll ideally need some sort of Data Management Platform (DMP) such as Oracle BMP or Marketo where you’ll combine your first-party website data (information generated directly from your website) and third-party data (behavioral and demographic information). At a minimum, your data sets here should contain:
- Customer data from your CRM
- First-party URL shortening and widget sharing data
- User search activity and keywords used
- Purchases (if applicable)
- Demographics and Geography
- Frequented blog topics or category pages
With this information, you can begin to track patterns and analyze what groups of people are actually looking at your website. This is where persona-based targeting comes into play, though you want to make sure you’re hitting the right user regardless of the device used.
To do so, you need to combine this data with a unique identifier that tracks the user.
2. Develop Individual Personas Based on That Data
Since DMPs can only track users based upon their desktop interactions or when they submit their information on a form, we have no way of knowing what they are doing on their mobile devices. You need a unique identifier.
Most user activities on portable devices tie into one of the three unique identifiers listed here:
- Statistical IDs: Algorithms operating off the user’s device, but using information provided by it, and/or by the gateway it uses to access the Internet.
- Google’s Advertising ID: A user-resettable, unique, anonymous ID for advertising, provided by Google Play services for Android devices.
- Apple’s Identifier for Advertisers (IDFA): A unique ID for each iOS device that mobile ad networks typically use to serve targeted ads. Users can choose to limit ad tracking by turning off this setting on their devices.
Setting up these personal IDs and connecting your anonymous users with one of them is the key part of recognizing the same person across multiple devices. For these IDs to be the most effective, however, they need to collect information from one of two types of matching methods: “deterministic” and “probabilistic.”
Deterministic tracking, the more exact of the two, relies on personally identifiable information (PII) to match the user across devices, in other words, when said user uses the same email to log into an app and a website.
This allows marketers to target and follow the user across multiple screens so long as they are logged in on the device, allowing for near perfect precision and accurate data collection.
While effective, this can create privacy concerns, since it is traditionally done without the user knowing or without the capability for users to opt out (as Verizon once had issues with).
For those who don’t want to use login mechanics or deal with these concerns, there’s Probabilistic matching which is achieved through an analysis of multiple groups of data, including location, device type, operating system, etc. to create statistical matches between devices.
The caveat here is that the math isn’t exact. Software such as Drawbridge and Crosswise that specialize in this data say they can match devices with a 71%-91% rate of accuracy. This leaves a sizable margin of error, but it still gives marketers a more consolidated view of their audience for them to work with.
3. Create Campaigns That Connect With These Personas
Now that you are able to recognize the same individuals on multiple devices, a DMP’s segmentation becomes significantly more powerful.
Now, marketers can now begin to monitor specific users’ behavior across multiple devices and build campaigns around them, and strategize ways to make those user tasks easier for them to complete.
According to RadiumOne, for example, “data from advertisers in the travel and financial services illustrate the first-hour after consumer signals intent to be a “sweet spot” in terms of willingness to convert.”
For the advertisers, delivering 13% of ad impressions during the “sweet spot” actually drove 81% of conversions, their highest conversion rates.
Since consumer intent takes time for DMPs to process, you need a more efficient way to receive the information so marketers can respond properly with an ad or content. This is where using a Demand-Side Platform (DSP) — inventory across a variety of ad exchanges, networks, and publishers– becomes incredibly useful.
Retargeting with a DSP
Since conversions happen so close to intent signals, utilizing this system helps activate real-time data about consumers in milliseconds, rather than minutes or hours. This helps capture opportunities quickly that otherwise would usually be lost.
For the most flexibility, your integrated DMP/DSP should reach individuals across all major real-time channels:
- Display real-time bidding (RTB)
- Mobile RTB
- Video RTB
- Social RTB
- Mobile push.
To show this in action, consider the following scenario – what would a marketer want to happen if a user visited their eCommerce site on a mobile device, then left?
Naturally, the marketer would want to quickly reach the person with an ad that brings them back to the site to make a purchase such as the one Kohl’s just showed me on Facebook below:
For this to happen, the DMP would use first-party data to see the user left without making a purchase. Then, using the personal ID, the DMP would pass them through a retargeting segment. The ID will tell the DSP how to retarget the individual, regardless of device.
The DSP will then find the next possible opportunity to reach the individual within an hour though networks such as through Facebook, the person will hopefully click the ad and the marketer will score a conversion.
There is no perfect methodology or tool to help keep your users on mobile buying your product, especially when some are just looking to get educated on your offerings or content.
With that being said, don’t use your energy trying to find the perfect formula; it’s all about testing and discovering what experience works for your users and improves your metrics using the tips from above:
- Define mobile behavioral cohorts based upon hypotheses you believe will connect to users reasons to convert.
- Use sessions lengths and duration to understand how long users are spending on your website and event tracking to identify what pages users are spending the most time on.
- Use a unique identifier to begin connecting your mobile website users and app users so you can better target these users on their other devices with appropriate ads/content.
Once you feel you have a grip on your core converting/high traffic areas, you’ll have a better grasp on how your mobile website is perceived by your users and its functionality.
Join the conversation
Add your comment
Good article as a whole, but very strange usage of the “Moneyball” syndrome. The idea behind Moneyball is using the right metrics to choose players, without paying big.
Actually, Moneyball is based on maths, analytics and algorithms – things we normally see in marketing and sales – transferred to sports. Not the other way around.
My point is that you are using a buzzword just to attract attention. The headline has little to do with the whole article. Not cool.
Thanks for reading. I actually used the term Moneyball here in reference to Jason Goldberg’s (senior vice president of commerce and content at Razorfish) keynote presentation at the ChannelAdvisor Catalyst Americas conference. Within it, he expands upon how “retailers need better metrics than simply relying on conversion. Conversion isn’t a KPI; profits are.” I had no intention of using moneyball as a buzzword and apologize if it came across that way.
In my presentation at Catalyst (Channel Advisor annual conference), that the author referenced, I referred to “Moneyball” in the context of e-biz leaders using “traditional” metrics like AOV/Conversion/Traffic/App Downloads/Etc… instead of understanding what metrics lead to actual wins. Just as Billy Beane shifted focus from “Looking Good in the Uniform” to “Total Bases”, I was encouraging e-biz leaders to be shift focus to annual customer value or lifetime customer value.
In fairness, the Moneyball reference wasn’t intended to be clever (or Clickbait), Billy Beane was the speaker prior to me at the show :)
Christin and Jason, thank you for the explanation. Totally agree with the rest of article, just the title sounded strange.
Comments are closed.