KEY POINTS
- Mobile users want information delivered to their phones in a more convenient, sophisticated, and personalized way. The days of searching a keyword in a desktop search browser are long gone. More and more, consumers expect their devices and apps to anticipate their needs.
- The mobile experience has become cluttered, and customized services can streamline it. Mobile users have an average of 33 apps downloaded to their devices and launch at least eight apps daily. In turn, they spend some 2 1/2 hours per day within apps. There's a huge opportunity for apps that can make it easier to access the information and services you need, without a lot of effort.
- There's a boom underway in consumer usage of apps that help them customize their smartphone experience. On Android, personalization apps saw a 1,325% increase in user sessions between early 2013 and early 2014.
- A wide array of mobile technologies are already on the market, each capable of offering personalized features and services: app launchers, custom lock screens, smart personal agents, intelligent personal assistants (e.g., Google Now), and task automation tools. But the biggest market for personalization is probably within specific apps — technologies and algorithms that can help show users information tailored to their specific needs.
- Personalization usage and demand is skyrocketing. Android app-launcher sessions are exploding, Google Now is available on nearly 1 billion devices, and several companies are looking to imitate Apple's Siri. The app If This Then That, or IFTTT, which automates tasks across apps — for example, it emails you a copy of every picture you upload to Instagram — has seen its "recipes" grow 80% since adding mobile support.
- Finally, there is a massive marketing opportunity in personalization. The same technologies that can power personalized information and notifications within apps can be used to customize the ads shown to app users and mobile Web surfers. This is the kind of individualized hypertargeting that is marketers' Holy Grail. Marketers need to better understand how end users are implementing personalization, and how to be engaging without being intrusive.
Introduction
The way consumers discover and absorb information has transformed dramatically thanks to mobile devices. Their expectations have also changed radically.
The Web browser and the search engine aren’t necessarily at the center of the mobile experience, as they are on desktop. Smartphone users increasingly expect their devices to push information to them. Or, they search within apps for information and increasingly expect content within these apps to be relevant to them as individuals — not just as members of a vaguely defined consumer segment or audience.
The ability to be relayed information anytime, anywhere, and to hold dozens of apps in the palm of your hand has made content discovery and search inherently personal. This has created an entirely new opportunity for developers and marketers to reach consumers.
"What's happening in content discovery and search is that new technology is being developed to allow people to find information through contextual signals, like voice, and location, and other signals," remarks Tim Tuttle, CEO of anticipatory computing company Expect Labs. "And that represents the new frontier of search and content discovery."
There are many buzz-worthy terms — contextualization, artificial intelligence, smart assistants, anticipatory computing — that overlap with the basic concepts of personalization and customization which we'll be looking at here in this report.
But each approach has its own strengths and weaknesses.
In this report, we'll be using the blanket-term personalization, which might include any of the following technologies (which also overlap with one another):
- App launchers and lock-screen/home-screen customizers — software that showcases the apps that are most relevant to the user given a specific time, location, or other context. At work, a user's home screen may show news and email apps; at home, Facebook and YouTube may be featured prominently.
- Digital concierges — Autonomous, algorithm-driven technology that studies user behavior and preferences to deliver an ongoing feed of personalized information, content, and task management. An example is Google Now, which delivers information-rich "cards," based on a user's past search history and actions. An example might be a card that lists movies a user might be interested in seeing over the weekend.
- Algorithms in general — This term doesn't require much elaboration, but an increasingly wide subset of apps, advertising platforms, and services are investing in algorithms that help them deliver personalized information, recommendations, features, or advertisements. An example is TripAdvisor's algorithm that determines results for the app's "Near me now" results, taking into account distances as well as friends' ratings and reviews.
- Voice-activated assistants— Technology that uses vocal recognition capabilities to perform specific, one-time tasks as instructed by the user's voice. These are often integrated with smart personal assistants. Ask for the weather, and the assistant will pull up the weather app showing climate in your location.
- Task automation tools— software that connects applications so that when a user completes one task within a specific app, another task is triggered and completed automatically in a separate app. For example, when a user posts a photo to Instagram, that photo will be automatically uploaded to their Flickr account. Actions within apps might also be automated.
Later in this report, we'll take a more in-depth look at each of these categories, and examine some of the well-known apps or technologies that fall under each, including Google Now, Apple's Siri, and IFTTT (If This Then That), among others.
Essentially, we'll be talking about any mobile technology that accommodates individual user preferences and behaviors in order to deliver more relevant information, enable discovery, or allow task-automation.
Click here to download the charts and data associated with this report in Excel »
Click here to download the report as a PDF »
The Need For A Streamlined Experience
A big reason why there's a growing desire for a more tailored experience is the increasing pervasiveness of apps.
They are quickly becoming the be-all and end-all to the smartphone experience.
These days, the modern U.S. smartphone user with apps on their device spends some 2 1/2 hours per day within apps according to Flurry. And that excludes voice calls and email activity.
Beyond games and social networks, it seems almost every major brand has created a dedicated mobile app. Smartphone apps have become powerful tools for entertainment, shopping, research, and productivity.
But such widespread app availability has created a problem: app complexity. To conduct complex tasks or searches, most users have to navigate through a chain of apps.
For example, a user searching for last-minute travel deals while planning a trip likely might have to
book a flight in the Kayak app, search for hotel rooms on Hotel Tonight, and book a car with the Zipcar app.
According to data from Flurry and Google, the average U.S. smartphone user carries about 33 apps on their smartphone, and launches about eight of those apps every day.
There's a parallel problem in the world of advertising — a cacophony of ad formats vying for users' attention: in-app banner ads, Pandora radio ads, social media in-stream ads, etc.
Most of the services we'll discuss in this report cut through this noise and complexity. They do this by doing what the best consumer-facing services have always done — either by providing concierge-style individualized attention, or by providing one-stop-shop type services that help users tailor their experience themselves.
Lock-Screen Apps And App Launchers — Aviate And Cover
Apps are inherently siloed.
The goal is to keep users within a single app for as long as possible. In contrast, Web browsers afford desktop users the opportunity to toggle between multiple tabs or pages instantaneously.
App launchers, lock-screen and home-screen apps are trying to bring an analogous experience to mobile screens, by allowing users to more easily access important apps at the right times or to switch between apps more easily.
Many of these apps create customized, personalized home screens and/or lock screens — showcasing the apps that individual users are most likely to use. They often use algorithms to filter the displayed apps according to the most-used apps, time of day, or location.
Home screen and lock screen customizers are an Android-only phenomenon. The more restrictive nature of iOS precludes any ability for developers to modify these screens.
App launchers, however, are also found on iOS — an example is Launch Center Pro, a $4.99 paid app. (It's also worth noting that the ability to toggle between apps for multitasking was improved with iOS 7.)
Increasing consumer interest in customization has led to a boom in demand for these apps, particularly on Android. It may be that by allowing for more customization, Android has done a better job than iOS in encouraging personalization app adoption:
- Among the 4,500 Android personalization apps operating on Flurry's app analytics platform (most of them home-screen and lock-screen customizers), there were more than 6.6 billion user sessions in the first quarter of 2014 — even with a few weeks left in the period.
![AndroidPersonalizationAppUsage]()
That's compared to just 468 million personalization app sessions in the same quarter a year ago. That means usage has skyrocketed 1,325%.- The 6.6 billion app sessions so far this quarter amount to more personalization app sessions than took place in all of 2013. There were about 5.8 billion app sessions last year.
With consumers responding favorably and usage taking off, expect many more mobile app companies to enter the field, particularly on Android.
Each will be looking to become dominant in the field of app launchers and customization.
The increased consumer interest in personalization has led to the beginnings of a race among big tech companies to acquire the apps that have already developed expertise and carved out a user base and differentiated technologies. These recent deals have also elevated these apps a bit in the eyes of investors and analysts.
In early January 2014, Yahoo announced it had acquired a startup specialized in creating an "intelligent home screen": Aviate.
Aviate surfaces relevant apps to a user's home screen based on where they are, or what time of day it is. It studies their phone's behavior based on signals like Wi-Fi usage, GPS, and accelerometer activity.
Aviate's technology may prove vital in that it will allow Yahoo — which is keen to become more mobile-centric — to push its own apps and content directly to mobile users.
Twitter made its first Android-specific acquisition with another lock-screen and home-screen app: Cover. Much like Aviate, the app picks out relevant apps to feature on a user's home screen after recognizing where the user is.
There's no real word yet on how Cover will fit into Twitter's broader Android strategy, but it indicates that the Android Twitter experience may end up being separate from what it will be on iOS.
If Twitter stays in the home-screen customization business, it will be interesting to see how it prioritizes its own app and content on users' home screens.
Digital Concierges — Google Now
Smart personal agents are designed to automatically execute tasks with little to no direct interaction from the user.
The smart personal agent's work is ongoing, is done autonomously in the background, and information is relayed intuitively and instantly.
The foremost smart personal agent on mobile is Google Now. The technology was first developed by Google as an addition to Android for the launch of version 4.1 Jelly Bean, which debuted officially in July 2012. About a year ago, the service was also made available on iOS devices.
Google Now works by analyzing the way you use your smartphone.
It's embedded within Google's broader Search app, so typical behaviors it pulls from include repeated search queries, common search locations, frequently visited sites and apps, and routine calendar appointments. For example, if you check to see what time your train arrives for your daily work commute, Google Now will learn that behavior and deliver train times to you every morning automatically.
After a user has opted-in to Google Now and pulls up the Search app, the Google Now feed is displayed prominently in the form of "cards," or boxes that feature information for different categories. One card can show your commute time and the next card will show you stock quotes for companies you frequently search. The only real interaction with this feed is that users can swipe cards from the feed away that they don't need. Otherwise, Google Now will intuitively deliver any relevant cards.
The other a way user can interact with Google Now is through voice recognition capability. Users can speak into the device, stating a specific command or asking a specific question, and Google Now will instantly respond with a specific answer by making a relevant recommendation or by completing a desired task, like scheduling something in a user's calendar.
Turns out Google may ultimately become the "Google" of personalized mobile experiences.
When it launched on iOS in April 2013, Google opened up the market for its smart personal agent service to almost 300 million more active mobile devices. That's on top of our estimate that there are more than 600 million active Android devices in the world.
That puts the market for Google Now at just under 1 billion active devices.
It's important to remember that while Android makes up about 80% of the global market for smartphones, only about 52% of the market is made up of Google-licensed Android phones. Google Now is only available to those running Google-licensed Android, not forks. For example, Google Now would not be available on a Xiaomi handset.
Of course, not every single device capable of running Google Now actually is, and there is no hard data to support just how many devices utilize Google Now.
But, we do know that Google Now operates within Google's Search app, and that Google holds an overwhelming 95% share of the mobile search market, meaning 95% of all global mobile search activity is funneled through Google.
In addition, Google also recently launched Google Now within the desktop Chrome browser, which will give it an even greater ability to integrate and learn from user activity.
Google has a serious upper hand in the data collection that powers smart personal agents. We believe that puts Google Now in position to be the go-to mobile personalization experience.
Voice-Activated Assistants — Apple's Siri and Microsoft's Cortana
Intelligent personal assistants perform specific, one-off tasks from a user's voice command.
A user can open the intelligent personal assistant app, say, "What's the weather?" and the software will recognize the question and instantly respond with the weather tailored to their specific location.
Apple brought the intelligent personal assistant into the limelight when it introduced Siri as a new feature on the iPhone 4S in October 2011. Siri has been an integral part of every update to the iOS platform since iOS 5.
Siri can handle a number of iOS tasks from triggering voice calls, utilizing voice-to-text for messaging, and adjusting your iOS calendar. It also supports search within Google, Google Maps, Bing, Yahoo, Yelp, and Wikipedia, among others. For example, a user can set up Siri so that when they want directions somewhere, the user can say where they're going and Siri will automatically launch Google Maps. Google Maps will then provide directions to that location.
Siri is one of the oldest and most recognizable personalization technologies on the market. And the market for Siri is pretty big considering that 98% of Apple devices run iOS 6 or iOS 7.
But has that large market translated to extensive usage?
Surveys suggest that initial reaction to Siri was enthusiastic, but as time has passed, demand for the service may have tumbled.
Nearly nine of out 10 U.S. iPhone 4S owners reported talking to Siri at least once a month back in March 2012, according to a Parks Associates survey. What's more, Parks Associates found that about 70% of all respondents claimed they were satisfied with Siri's capabilities.
Conversely, a more recent survey from Intelligent Voice found that 85% of U.S. consumers had not used Siri in Apple's most recent platform update, iOS 7. The survey was conducted over the course of October 2013, among 2,330 U.S. adult Internet users, and the methodology may be a bit flawed in that it surveyed total Internet users, opening up the population to users of platforms not compatible with Siri, like Android.
Nonetheless, given that Apple had a 41% share of U.S. smartphone users in September 2013, it still looks like Siri isn't being adopted by an overwhelming number of iPhone users.
By one measure, though, Siri has certainly succeeded. It has bred a number of imitation technologies that other smartphone vendors are incorporating into their own hardware, indicating that developers and manufacturers think the potential for voice-activated personal assistants is considerable.
These imitators include LG's Voice Mate, Samsung's S Voice, and HTC's Hidi, though none has been especially successful.
But the most high-profile competition to Siri is Microsoft's new intelligent personal assistant, Cortana. The technology was unveiled in February 2014 and was released as part of the Windows Phone 8.1 update.
In an in-depth review of the technology, CNET found that out of the gate, Cortana can hold its own against Apple's Siri, and praised Microsoft for getting an early jump on third-party app integration.
Even if personal assistant usage hasn't truly caught on yet, there are several intriguing possibilities that can take this technology to the next level.
The mobile industry is expanding to include wearable devices and connected cars, and reports claim that Apple is making a big push to revamp Siri for these purposes.
In wearables, Apple is opening the Siri platform to more third-party app integration, likely with its impending, though not yet confirmed, iWatch in mind.
One of the reported ideas is to use Siri to bring relevant apps to the forefront of the wrist's small screen, which can only show a few apps at once. Siri would surface apps based on what it detects the user is doing. For example, if you're jogging, your Siri-powered watch could automatically pull up your Nike+ fitness app.
One of Siri's pitfalls is that Apple has been notoriously slow moving in opening the technology up to third-party app integration. But consumers do want to leverage Siri for more services. In fact, a few college students at the University of Pennsylvania have developed GoogolPlex, a Siri modification. The hack creates integration for third-party apps currently not supported by Siri, like Spotify, Nest smart home appliances, and even Tesla vehicles.
In cars, Apple has already introduced its CarPlay technology, which extends iPhone and Siri usage to the car's dashboard. In the battle for OS supremacy in the car, Siri is Apple's ace. The technology has an obvious upper hand here, given that it’s a hands-free, voice-command service.
"When incoming messages or notifications arrive, Siri provides an eyes-free experience by responding to requests through voice commands, by reading drivers’ messages and letting them dictate responses or simply make a call," said Apple to MarketWatch.
Task Automation Tools — If This Then That (IFTTT)
Task automation tools help users connect apps to perform a string of tasks that can be automatically completed in one fell swoop.
For example, you may be checking your email on your way to work when you receive an important attached file from a client. Click on that attachment from your phone and a task automation tool would theoretically know to automatically file that document to your team's Dropbox account for your colleagues to access.
While these task automation tools may require more direct user interaction for initial setup than, say, Google Now, the end-goal is to lessen the burden of having to navigate through multiple apps to get basic, related tasks completed.
Originally launched in late 2010 as a way to connect Web applications, If This Then That (IFTTT) made a splash in the mobile industry by adding iOS support in July 2013.
In the beginning, users were able to create specific personal tasks to be conducted on the Web. But IFTTT's technology progressed into shared "recipes" as the service grew in popularity. The idea now is users can create a task and make it a “recipe” by sharing it with the IFTTT community for others to utilize.
These recipes originally came out on desktop but have become particularly relevant on mobile. Recipes are conditional statements linked together by one trigger function and subsequent action. To simplify the setup for the user, IFTTT has dedicated channels for different apps (i.e., Instagram, Dropbox, Gmail), and users can quickly and easily create a recipe by linking these channels. These recipes are then shared through IFTTT for other users to easily integrate into their devices.
For example, in the above example of sending an email attachment that is opened on a smartphone to Dropbox, the trigger would be downloading the email attachment. The subsequent action would be that attachment being saved to Dropbox.
These shared recipes make different apps part of one cohesive task.
Since they were first introduced in September 2011, IFTTT users have created more than 160,000 different shared recipes, according to data provided to us by the company.
What's more, in the nine months since the introduction of IFTTT for iOS, we estimate that IFTTT recipe creation has grown nearly 70%.
The Marketing Opportunity
Personalization brings an inherent opportunity for hypertargeting ads and services, and this is consistently the Holy Grail for marketers.
As easily as a user can be served up details about the day's sunny weather, marketers can use the contextual and behavioral data gathered by personalization apps to deliver that same user an advertisement for the local sunglass shop.
Of course, the challenge here is how to properly channel promotional material to the user within what is built to be an innately personal service — to be engaging without becoming intrusive.
Here's a look at how personalization could fit into the overall mobile ad market.
Mobile is still a growing advertising arena, but using mobile as an ad platform is not a new practice. In 2013, U.S. mobile ad revenues grew 110% to $7.08 billion, according to new figures released by the Interactive Advertising Bureau (IAB). That makes 2013 the sixth year in a row in which mobile ad spend has at least doubled.
But mobile's share of ad spend still significantly lags behind time spent on the devices. That's in large part because no one has figured out which ad formats really foster engagement on small mobile screens.
Personalization could help narrow the gap in mobile usage versus mobile ad spend. If marketers can effectively use the data and services offered through personalization services to serve the absolute most relevant content to users, the cost of ads will likely rise to match effectiveness.
But just as there are specific mobile ad formats that have proven more successful than others in generating consumer engagement, similarly, the marketing opportunity is different within each specific personalization technology category described throughout this report.
On customized lock-screen apps, for example, the idea is to engage with the user from the instant they turn on, or wake up their smartphone.
"The average user is going to unlock or wake up their phone at least 100 times per day. This is a huge opportunity because they have to see the lock screen whenever they're going to use their phone," said Luke Ahn, CEO of Fronto, a company that specializes in lock-screen advertising.
The unique aspect of Fronto's technology, according to Ahn, is the option users have to engage with the ad, or not. No matter what, the ad will be present when the user unlocks their phone. But the user can then swipe one way to engage with the ad, or the other way to bypass it entirely.
This may provide a refreshing amount of power to the user compared to clunky, ever-present mobile display ads. "Users have full authority to control their lock screen or home screen, not like traditional interstitial ads," added Ahn.
The smart digital concierge and voice-activated assistants present another advantageous, personalized platform for ads.
"With traditional ad-serving, it's a little special-purpose, and a little more about understanding the user, what the user's demographics and interest is. Whereas when we're talking about creating an intelligent assistant, there's more emphasis centered around trying to understand the knowledge base or corpus of data," remarks Tim Tuttle of Expect Labs.
Since both technologies are rooted in search, an obvious fit is paid search.
Mobile search activity is already driving a growing share of paid search clicks in the U.S. Just over 34% of paid search clicks originated from smartphones and tablets in December 2013, according to a new report from Marin Software. That's up from a 21.8% share of paid search clicks at the beginning of the year.
One example of a paid search ad format that would fit seamlessly with a digital concierge or voice-activated assistant is click-to-call and location-aware ads that help consumers move through the shopping process, from online to offline. Looking up location information for offline retailers is one of the top retail activities on smartphones.
For both personalized ads and mobile ads in general, mobile ad spending will increasingly move to newer formats that work best on the constrained mobile screen, such as native in-stream advertising, app install ads, rich media, and mobile video ads.
In fact, the mobile ad format mix is already trending that way.
According to mobile ad network Millennial Media, the post-click action that is most widely used on mobile ad campaigns is prompting a subsequent app download. About 34% of all mobile platform campaigns during 2014 integrated an app download. Following closely behind that was site search, which made up 29% of all campaigns. Video ads made up 18% of all campaigns.
When leveraging personalization apps and technology for targeted campaigns, these are the formats that will likely drive the most interest among marketers.
THE BOTTOM LINE
- Mobile users want information delivered to their phones in a more relevant, sophisticated, and personalized way. The days of searching a keyword in a desktop search browser are long gone.
- The mobile experience has become cluttered. Mobile users on average have 33 apps downloaded to their devices and launch at least eight apps daily. In turn, they spend more than 2 1/2 hours per day within apps. But amid all that time spent, it's still not easy to toggle back and forth between apps to complete a task.
- A bevy of technologies are already on the market, each capable of various personalized functions that can make it easier to access the information and services you need, without a lot of effort. These technologies run the gamut from app launchers, to custom lock screens, to smart personal agents, to intelligent personal assistants, to task automation tools.
- Personalization usage and demand is skyrocketing. Android app-launcher sessions are exploding, Google Now is available on nearly 1 billion devices, several companies are looking to imitate Apple's Siri, and If This Then That (IFTTT) has seen its recipes grow 80% since adding mobile support.
- As personalization use skyrockets, a massive marketing opportunity in personalization is rising alongside it. These technologies could offer the kind of hypertargeting that marketers are perpetually after. The goal for marketers is to understand how people are utilizing personalization, and how they can enter into that realm without being intrusive.