In August of 2013 Google suffered a rare two minute outage. During that infamous two minutes 40% of the web went dark. Forty percent! This event was a clear reminder of the ultimate control Google has over the web and web traffic.
Since the early days of the web, control of discovery and web traffic has been the focus of many. In the mid- to late-nineties, such control was in the hands of ISPs such as AOL and Earthlink. In the late nineties, the control shifted to portals with Yahoo and its MyYahoo service in the lead position. (I recall a conversation in 2000 with a former colleague of mine asking for an introduction to the MyYahoo team: “If you are not on MyYahoo, you simply don’t exist. It costs too much to promote your site.”) In the early to mid 2000s, Google dis-intermediated Yahoo and stole that position with Search and AdWords. Google has since reigned supreme over web discovery, organic and paid traffic acquisition. An entire industry has been built around SEO and SEM and both of these acronyms have become synonymous with the name Google.
Then came mobile and its apps. In just over five years since the launch of the Apple AppStore, apps have taken the mobile industry by storm. Today, 86% of time spent on mobile devices is spent inside applications, with just 14% left for the mobile web. App discovery has been a major challenge over the past five years. That led to the fast rise of the paid App Install market. In fact, Marketers are expected to spend over $2.4 billion USD in the US alone to get their apps noticed.
Climbing into a Walled Garden
Discovering relevant content and services within apps has also been a big challenge. Consumers want such a service. As an example, if a consumer is searching for a flight to Las Vegas, in an ideal scenario the phone or discovery service would launch whatever flight booking app is installed on the phone and send the consumer directly to the reservations page. But apps (alongside the content and services in them) can’t be crawled. They are not just about index and links, and hence are hard to “page rank.” As a result it is difficult for a search engine to discover apps and offer consumers an entry point.
Consumer behavior on mobile is very different than the desktop web. The de-facto behavior on the web is to launch the browser to google.com, type a sentence and be re-directed to content. The de-facto behavior in mobile is to launch an app (previously installed on the device) and enjoy the comprehensive experience offered within it. You are rarely, if ever, linked out to a mobile webpage from an app or sent to another app. Each experience is essentially an island unto itself, completely reliant on the consumer to come ashore.
The industry and some major mobile players are looking to solve this problem and the battle is raging for the “gateway” for mobile apps, content and services. In short, the Google position for mobile is up for grabs and billions of dollars are at stake.
At Flurry, we are in the middle of that battle and have built solutions to help apps get discovered and acquire traffic. We watch the space very carefully. We have seen a lot of companies, large and small, try to solve this problem: Apple with Siri, Google with Search and Google Now, Facebook with App Install and App Conversion Ads, and myriad others. But recently we have been most intrigued by Android Personalization apps, also called Launchers, especially after the creation of Facebook’s Home and Yahoo’s acquisition of Aviate. So we have taken a closer look at this category and the chart below shows the very fast adoption and usage of these personalization apps on Android.
The majority of these apps offer a personalized homepage for Android, and do some form of app discovery and app launch.
There are over 4,500 apps in this category on the Flurry platform and the data shown represents the aggregate usage of these apps. It shows the number of app sessions recorded by Flurry in apps in the Android Personalization category, by quarter, since the beginning of 2012. The growth in that category is astonishing. In fact, Q1 2014 (and the quarter is not over yet) usage is higher than all of 2013. While the cumulative reach of these apps is still relatively small (30 million monthly users in the US), the usage growth is eye catching.
This data set is telling us something. In fact, this data set is telling us something big. It is telling us that consumers are eagerly waiting for an innovative service that help them discover apps, content, and services around them in a personalized way. While the numbers are still small, the fast adoption of these apps can’t be ignored. The battle for the mobile homescreen has begun. The battle for the “Google position of mobile” has entered a new phase, and it is “must-win” battle for many.
Developers devote a lot of attention to the time immediately after an app is launched. How quickly is it growing? Will it go viral? How is it ranking in app stores? While that launch period is critical, managing apps well throughout their entire lifecycle means also paying special attention to what happens after an app peaks. Does it decline precipitously or manage to hold much of its audience for a long time?
We explore the post peak phase of the lifecycle of mobile apps using the concept of a half-life, which for purposes of this analysis we define as the point at which an app’s monthly users (MAU) have declined to 50% of their lifetime peak.
Usage of Many Apps Decays Rapidly The First Few Months Post Peak And Then Levels Off
We start by looking at the overall rate of decay for 26,176 apps that peaked in the first half of 2011, 2012, and 2013. (That allows enough time for decay even for apps peaking in 2013 and also controls for the time of year in which the app peaked.) The chart below shows what proportion of apps were at what percentage of their peak MAU in each of the first ten months after they peaked. Apps in the gray band fell below 25% of their peak MAU by the month indicated, apps in the orange band had fallen below 50% but were still above 25%, apps in the green band were below 75% but above 50%, and apps in the blue band were above 75% of their peak MAU in the month shown in the horizontal axis.
As shown in the chart, decay is fairly rapid in the first few months post peak. A quarter of apps fall below half of their peak MAU the month after peaking (see the top of the orange band in the chart). That percentage grows to 43% in the second month post peak. In the third month post peak, just over half (52%) of apps fall below half of their peak MAU. The fourth month post peak is a critical benchmark: From this point on, MAU decay continues, but at a much slower rate (note the flattening of the color bands after the 4th month).
The overall pattern of app user decay has been remarkably stable over the past three years, but there is variation based on app category, size, and operating system. The table below shows median app half-life for apps within each grouping. This corresponds to the month in which the top of the orange band and the bottom of the green band intersect at the 50% line in the previous chart.
The biggest influence on the rate of app decay is category. Apps in utilitarian categories such as news and health decay at a much slower rate than apps in the more trend-driven games and social categories.
On average, apps that peak at 10,000 MAU or more enjoy a half-life that is two months longer than apps that don’t reach that level.
iOS apps typically reach their half-life a month later than Android apps.
Managing Decline Is As Important As Managing Growth
Many developers give a lot of thought to how to grow their apps, but we suspect fewer devote the same attention to managing their app’s decline. In some ways, this is analogous to people looking after their health as they age. The right habits can prolong longevity. In the case of apps, that means paying close attention to user retention and re-engagement – particularly as MAU growth begins to plateau and decline.
Those first few months post peak are key since the rate of decay levels off after about four months. More than half (56%) of apps that manage to hold more than half of their peak users for the first four months after they peak are still holding on to more than half of them ten months post peak. That could have a sizeable revenue impact. For example, consider an app that peaks at 100,000 monthly average users and has average revenue per user per month of $2.50. Holding at least half of those users through month four means the app has a 56% chance of holding them for another six months and that translates into revenue of $750,000 (100,000 users* half of them remaining *$2.50 per user per month * 6 months).
Just as actuarial tables help insurance companies anticipate costs, understanding typical app decay patterns can help developers and those who fund them anticipate app revenue streams. For example, assuming similar development costs and average revenue per user per month, a game app would need to peak with many more active users than a news app to have the same profit potential since its life can be expected to be a lot shorter. And it pays to invest in growing MAU for any app to the highest peak possible since that is also likely to mean the app will enjoy a longer half-life.
In November 2013, Benedict Evans, a well-respected and widely followed analyst, shared an updated version of his famous slide deck called “Mobile is Eating the World”. This deck quickly made the rounds on social media and was highly referenced by industry and financial analysts who cover mobile. We can’t help but agree with Benedict’s conclusion. For the past five years, we have watched mobile disrupt every industry, in every country, and continue to break its own records year after year. 2013 did not disappoint.
According to Flurry Analytics, overall app use in 2013 posted 115% year-over-year growth. (In this context, we define app use as a consumer launching an app and recording what Flurry defines as a session.)
Every single app category has shown growth over the last twelve months. In the chart below, we have focused on the categories of interest to most. Utilities and Productivity apps posted 150% growth in use year-over year, as smartphones and tablets became personal computers and productivity apps, such as Evernote and Quip, gained sophistication and adoption. Even Gaming, which was feared to reach saturation levels in 2013, posted 66% year-over-year growth in use.
However, the segment that showed the most dramatic growth in 2013 was Messaging (Social and Photo sharing included). The growth in that segment should not come as a surprise to many, given the attention that messaging apps such as WhatsApp, WeChat, KakaoTalk, LINE, Facebook Messenger and SnapChat have received in the press. What is surprising, however, is that the rate of growth (tripling usage year-over-year) dramatically outpaced other popular categories. This type of growth could explain the high valuation Facebook has allegedly put on SnapChat, or Facebook’s rush to add direct messaging in Instagram, an app frequented by teens.
Killer Apps or Killer Platform?
While some of these apps, such as Korea-owned and Japan-based LINE, are enjoying great revenue growth, there is still a debate about whether these apps are simply experiences or more of a platform. 2013 saw a few examples of these apps becoming more of the latter.
In March of 2013, and just three months after launching it game distribution platform, LINE announced that it had delivered over 100m downloads to its gaming partners.
Tencent’s WeChat has conducted an experiment with China’s emerging device manufacturer Xiaomi demonstrating WeChat’s potential as an m-commerce player. In that experiment, Xiaomi launched a new smartphone to WeChat users. The result: 150,000 new smartphones sold in under 10 minutes through a messaging application.
Such examples, coupled with Facebook’s own successful entry into the paid mobile app install market, have demonstrated the potential messaging and social applications have to become a mobile storefront for digital and physical goods.
It is these simple yet very promising anecdotes, coupled with an over 200% year-over-year growth, driving the frenzied land grab in the Communications market. 2014 will be a crucial year for these applications and will determine whether they will remain independent, but highly frequented applications, or become killer mobile platforms and distribution channels. TIME magazine has already placed its bets on the latter. At Flurry, we are fascinated by the growth, the retention rate, the reach and the frequency for this category. We see the potential for what it might become, but before declaring that mobile is all about messaging apps, we would like to see more of these marketing/commerce experiments a la Xiaomi/WeChat happen at scale and in many countries.
Another explosive growth year in mobile has passed. On December 31st, 2013 at 11:59 pm, Flurry Analytics tracked a record 4.7 Billion app sessions in a single day, for a total of 1.126 Trillion sessions for the whole year. Those are some very, very big numbers. One minute later the counter went back to zero. A new year has begun, and if the first few days of January are any indication, the mobile world is looking at another major growth year and yes Benedict (and Fred Wilson too), mobile is continuing to eat the world.
As we said in previous posts, connected devices have become part of our Christmas and Thanksgiving traditions. This marks a change from just a few years ago when smartphones and tablets represented exciting new additions to our lives that we couldn’t wait to unwrap, activate, and start filling up with apps.
Flurry’s data showed another record-breaking level of app downloads this Christmas, but it also provides further evidence that, in the early mobile markets at least, devices are evolving from being our new shiny toys to our everyday companions.
For this report, we looked across more than 400,000 apps that Flurry tracks globally to see what happened around the Christmas tree. We found that Christmas downloads were up by 91% compared to an average day in the first three weeks of December. That is a large increase, but as shown below, the size of the Christmas download spike is diminishing over time as the app market matures and globalizes.
More Signs the App Market is Maturing
2013 was the biggest Christmas yet for mobile app downloads. As the chart below shows, overall app downloads increased by 11% on Christmas 2013 compared to Christmas 2012. Over the same one year time span, there was a 25% increase in app downloads on an average December day. As the data illustrates, both the overall rate of year-on-year growth and the year-on-year Christmas growth have slowed considerably in the past year (from 97% and 90% respectively between 2011 and 2012).
The slowing growth rates and smaller Christmas Day app download spike signal market maturation. Many consumers in Western Europe and English-speaking countries -- large mobile markets where Christmas is a big holiday -- already have a smartphone and / or a tablet. Fewer people are coming online with mobile for the very first time. Consumers who are on second, third or fourth devices have apps that they like and trust, and while they still download new apps, there isn’t much more impetus to do so on Christmas than any other day when they have a little downtime. New device activations do still spike on Christmas, but that spike is waning compared to years past, and it comes on top of a much larger installed base. That means that when new devices are loaded with apps, the overall impact on app download volume is not as big.
The biggest growth in mobile now is coming in countries where Christmas is a less significant holiday or not celebrated at all, so new device activations and app downloads come at different times of the year in those places. And because those high-growth areas are joining an already large global market, overall growth rates are less striking than when the mobile market was new.
Leisure-Oriented Apps Experience the Biggest Christmas Lift
To look at the types of apps that see the greatest increase in downloads at Christmas, we indexed the numbers by comparing Christmas to an average day in the first three weeks of December to eliminate inflation due to the continued growth of the overall connected device installed base.
After doing that, we see a similar phenomenon to what happened on Thanksgiving. Because our devices have become so intertwined with our lives, on a day like Christmas when most of us are relaxing, the apps we download reflect that. Games and social apps were downloaded on Christmas at twice the rate they were on a typical December day. The other categories that experienced the largest surge in Christmas downloads were media (photo, video, music) and lifestyle (sports, books, magazines, entertainment).
Are Prospects Changing For App Developers?
Maturation of the app market means that changes are in store for app developers. The gold rush days of huge jumps in the overall size of the connected device installed base on Christmas, followed by a dizzying rush of app downloads are fading. Of course, there will still be apps that tap into some vein of consumer interest or amusement and developers who strike it rich as a consequence. But overall, the successful developers in 2014 and beyond will be those who put in the hard work of identifying a target group of users, creating apps that work well for them, and continually refining and reinventing mobile experiences to profitably retain those users. We’ll talk more about that in early 2014.
According to IBM, two out of five online retail visits in the U.S. on Black Friday were made from mobile devices, directly generating more than 20% of online retail sales. On Thanksgiving Day, smartphones and tablets accounted for an even greater share of online retail visits and sales (43% of visits and 26% of sales). These results show the extent to which connected devices now influence retail sales, but Flurry’s own Thanksgiving weekend data demonstrates that our relationships with our smartphones and tablets go well beyond picking up a Black Friday deal while picking at Thanksgiving leftovers.
App usage overall in the U.S. spiked by 25% on Thanksgiving compared to the previous Thursday, and not just as a result of shopping app use. The overall pattern of app usage over the Thanksgiving weekend demonstrates that smartphones and tablets have become the first truly personal computers, changing their function as we change our routines. To illustrate that point in this post, we take an in-depth look at U.S. app session starts from the day before Thanksgiving through to Cyber Monday. We compare Thanksgiving week to the week prior, and also compare those time periods for the two previous years. (By including the weekends prior to Thanksgiving for each year we can distinguish between overall growth in app usage and use that is unique to Thanksgiving weekend.)
App Use Spikes on Thanksgiving
The chart below provides a high-level view of how overall app usage has changed from year to year, and changes between the week prior to Thanksgiving and the week of Thanksgiving. The gaps between the green, blue, and gray pairs of lines illustrate the extent to which app usage in the United States has grown year over year, with baseline usage the week before Thanksgiving up by about two-thirds compared to the same time last year, and about triple what it was during November of 2011.
The darker line in each color pair shows Thanksgiving weekend for a particular year while the lighter version shows the prior weekend for that year. The gap between the lighter and the darker lines in a color pair shows the difference in app usage by day between Thanksgiving weekend and the prior weekend. For the past three years, that difference has been greatest on Thanksgiving Day, diminishing to return to normal by Cyber Monday. In 2011 and 2012, U.S. app usage spiked by 20% on Thanksgiving Day compared to the previous Thursday. This year the Thanksgiving spike grew to 25%.
Given mobile devices are so personal, it’s no surprise that they tag along as we enjoy Thanksgiving weekend, but we wanted to understand the drivers behind the Thanksgiving spike in app usage. Is it just mobile shopping? Recipe apps aiding Thanksgiving cooks? Football fans checking stats from the couch? Social networking apps used to stay in touch with family far away? What accounts for the extra app sessions on Thanksgiving?
To find out, we compared app sessions by category for Thanksgiving weekend to the weekend before, focusing on the three days for which there is the biggest difference in overall app use.
The results are shown in the table below. The rows are app categories aggregated across operating systems. The columns show changes in app usage between Thanksgiving week and the same day the previous week for each category. Each cell shows that same comparison between Thanksgiving week and the previous week for the past three years.
Use of Shopping Apps Is Up Overall – Not Just On Black Friday
For obvious reasons, shopping apps consistently spike over the Thanksgiving period. Interestingly the bump this year wasn’t as large as the one last year; however that should not be interpreted to mean that enthusiasm for shopping apps is waning. On the contrary, between 2012 and 2013, overall use of shopping apps in the week before Thanksgiving grew by about 70%, so the 2013 Thanksgiving spike comes on top of a higher baseline usage level.
Media and Gaming Are Even More Popular On Thanksgiving
App categories beyond shopping that saw spikes in app usage during Thanksgiving week are fairly predictable. Media spiked on Thanksgiving Day in particular – most likely in part because it includes photo and video apps used to record family gatherings. It was also somewhat higher on Wednesday and Friday compared to the previous week.
More time for relaxation probably explains why game apps, which are always popular, enjoyed even more use over the Thanksgiving period than the week before.
Even Workaholics And Dedicated Calorie Counters Take Thanksgiving Off
What types of apps experience less use on Thanksgiving and the days immediately before and after than during the previous week? Mainly apps associated with things people take a break from over the Thanksgiving holiday: business and education, health, and news.
Travel apps also experience a decline in use Thanksgiving Day and the day after. That may seem counter-intuitive since Thanksgiving weekend is one of the busiest travel periods of the year; however it’s important to remember that people tend to be traveling to familiar places and often staying with relatives or friends so are less likely to need apps for booking hotels and rental cars or finding their way around an unfamiliar city.
Smartphones and Tablets Have Become Our Constant Companions
So shopping does contribute to the spike in app activity on Thanksgiving, but it’s only part of the explanation. This data shows the extent to which smartphones and tablets have become our constant companions, morphing their function to our whims and circumstances in ways even laptops never really did. They are truly our most personal computers. Sure, you can use them to shop without getting out of bed, let alone braving crowds at the mall, but your device can also entertain you with games, music, and movies. It can even tell you how far you need to walk to make up for that second helping of mashed potatoes – if you choose to look.
Santa is making his list and checking it twice, and with Black Friday and Cyber Monday coming soon it’s time to start making your lists too. As we’ve said before, apps and app usage reveal a lot about the interests and passions of device owners. With that in mind, we’ve compiled app-inspired gift ideas for the smartphone and tablet users on your holiday list.
To do that, we investigated app usage patterns for different Personas (psychographic segments) using a sample of 97,963 randomly selected iOS and Android devices. Our suggestions are based on the apps that members of each Persona use the most or to a greater extent than other smartphone and tablet users.
A Reminder On Giving Apps And Other Digital Content
Many apps are free, but if you want to give paid apps or other paid digital content note that you can give specific music or apps or set up a monthly allowance from Apple’s iTunes and App Store, whereas from Google Play you can only select a gift card. You can always add some suggestions for paid and free apps or other digital content you think a recipient might enjoy to make a gift card a little more personal or transform a holiday card into a gift.
Wishing You Joyful, Gadget-Lit, Holidays From Flurry
The best gifts tap into the recipient’s life and passions. Apps are a great way to gain insights about your loved one’s interests, so before you start shopping look at their phone or tablet for some ideas. We hope this has given you some inspiration for selecting holiday gifts, and wish you a joyful, gadget-lit, holiday season.
Links referenced in the infographic: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30
In October of 2013, Softbank Capital made a $1.5B USD investment in Supercell, the maker of two successful mobile games, giving Softbank a 51% ownership of the game maker. This investment caught the world’s attention. It wasn’t just the rich $3B USD valuation Softbank placed on Supercell that intrigued onlookers, but more the speculation that Softbank and its CEO Masayoshi Son were onto a bigger trend. Mr. Son has a solid track record in anticipating big shifts in worldwide markets in general and the tech industry in particular. In the mid-nineties Softbank was a publishing powerhouse. By the mid 2000s the company became an Internet powerhouse. Today, Softbank is the third largest wireless carrier in the world. So what does this investment signal about the big shift that Mr. Son is anticipating? At Flurry we believe he is placing bets on the Mobile Content Explosion that is taking place around us.
Early Indicators Signal the Content Explosion
At Flurry, we have always looked at the applications being started on our platform as a leading indicator of the app economy’s health. This is very similar to how U.S. economists treat housing starts as a leading indicator for the national economy. Typically, developers engage with Flurry and start applications on our platform a couple of months before they list them on App Stores. So if the activity on Flurry increases, it signals that more apps (and content) will be available on the stores within a couple of months.
Looking at application starts on the Flurry network since January 2012, we see an increase in the quarterly growth rate. This is in stark contrast to theories that the app ecosystem is congested. In fact, in just over 18 months, the rate of which new apps are being started on the Flurry network has nearly doubled as shown in the chart below.
While Flurry’s market share in analytics could have increased, we don’t believe it is the major factor in the acceleration of applications starts. Instead, we believe that we have entered a new phase of mobile content explosion, driven by rapidly changing consumer behavior. Over the last two years, application developers and media companies have seen the shift from personal computers to smart mobile devices including phones and tablets that are now in the hands of over 1.2 billion people worldwide. They have also seen the wild and global success of gaming, utility and messaging applications such as LINE, Kakao, Snapchat and WhatsApp. They are simply acting accordingly. With the hopes of reaching these 1.2 billion people with a press of a button, app developers and media companies are building mobile apps like never before.
There is an Audience for That
Pundits have criticized the increasing number of apps and have often claimed that while there are millions of apps out there, very few are being used. They also claim that a few app developers have the lion share of usage, especially in the United States and other mature markets such as Japan and South Korea. Earlier this year, a report from Comscore claimed that Facebook (and Instragram) accounted for 26% of all times spent on mobile. In its latest earnings reports, Facebook’s COO Sheryl Sandberg almost confirmed Comscore ‘s numbers and claimed that Facebook’s share of people's time is larger than that of YouTube, Twitter, Tumblr, Snapchat, LinkedIn, AOL and Yahoo combined. While Facebook’s reach and percentage of time spent are in a league of their own, there appears to be plenty of whitespace for others. In fact, just on the Flurry platform the number of independently owned app developers that have a worldwide audience of over 20 million Monthly Active Users (MAU) has jumped from 7 in Q1 2012 to 32 in Q3 2013. That is whopping 357% growth in 18 months.
In the same period, the number of app developers with an audience over one million MAU has risen from just under 400 to 875, a whopping 121% growth.
These numbers are simply unprecedented, especially because most of these app developers have risen organically, and not as a result of consolidation or through mergers and acquisitions. If anything, the market, its reach and the time spent on mobile is still with the “middle class”, or the mid-tail developers and content owners. Among the 1.2 billion device owners, app developers are finding millions of people to enjoy their apps and the content behind it.
Flurry’s numbers, which show the fast rise of app developers with large audiences, seem to indicate that worldwide, consumers with smart devices are still hungry for apps and mobile content, and app developers are building at increasing rates to feed this demand. We believe that once again, Softbank’s Masayoshi Son could once again be onto something really big.
According to the U.S. Census Bureau, more than 12% of people in the U.S. speak Spanish at home, and nearly fifty million Americans are of Hispanic origin. The collective purchasing power of Hispanics is expected to reach 1.5 trillion dollars by 2015. Brands are clearly aware of the importance of this demographic group, and there are advertising agencies dedicated toward helping brands communicate with Hispanic consumers. Nonetheless results of some recent research suggest that brands are not doing enough to create positive app experiences for Spanish speaking consumers in spite of the fact that Hispanics in general are enthusiastic users of smartphones.
Last week’s Advertising Research Foundation forum on the intersection between mobile and culture inspired us to investigate what the overall U.S. appscape looks like from the perspective of Spanish-speaking device users. Flurry is able to do that because we can detect what language a user’s device is set to each time they start an app session, and we record data from over a billion app sessions each day in the U.S.
Defining Spanish Interest Apps
We started by defining Spanish Interest apps as those for which there is at least one U.S.-based user with their device’s language set to Spanish for every twenty set to English. We used this as an indication that a non-trivial proportion of the app’s users speak Spanish, indicating some level of interest in the app among Spanish speakers. As shown below, 15% of apps with fifty or more daily users in the U.S. were in this category.
Android Speaks Spanish
The overall percentage of Spanish Interest apps is in line with what you might expect given the prevalence of Spanish speakers in the U.S., but what’s surprising is how those apps are distributed across mobile operating systems. As shown below, nearly half of Android apps are Spanish Interest apps compared to less than 5% of iOS apps. In other words, 47% of Android apps have one or more user with their device set to Spanish for every twenty who have it set to English. That is only true of 4% of iOS apps.
It’s important to note that this may not be so much a reflection of app availability as it is a reflection of device ownership patterns since the apps a user can run are determined by their device’s operating system.
Spanish Speaking Android Users Love Games Even More Than Everyone Else Does
The apps that have the greatest number of U.S. users with devices set to Spanish as compared to English are mainly Spanish language apps (some originating from the U.S. and some from countries for which Spanish is the dominant language). Other than that, Spanish Interest apps span the full range of app categories but are disproportionately likely to be in the Game, Live Wallpaper, Personalization, and Photography categories in Google Play. As shown below, Spanish Interest apps are also over-represented in some iOS categories, but because the base rate on iOS is so low (4%) the absolute percentage of Spanish Interest apps is still relatively low even in categories in which they are over-represented.
Fans Of Pets, Sports, And Bargains
By definition, the audience for Spanish Interest apps includes Spanish speakers, but we wanted to find out what else we could learn about the audience for those apps. We did that by identifying which Flurry Personas (psychographic segments) are over- and under-represented in Spanish Interest apps. We found that Pet Owners, Sports Fans, and Value Shoppers are among the over-represented Personas and Auto Enthusiasts, Home and Garden Pros, and Real Estate Followers are among the under-represented. A more complete list is provided below.
Brands Don’t Speak Android
Many brands have taken a while to embrace mobile at all, and to the extent that they have they have tended to start with iOS. Some haven’t ventured any further. In some cases that is due to concerns about fragmentation or for brand safety in the more open Android ecosystem, but it is also due to the demographics of the user bases for iOS and Android. A considerable body of evidence suggests that, on average, iOS users are more affluent than Android users and they tend to spend more money in a variety of product categories.
Combine that with the tendency of Spanish Speakers to use Android devices, and it’s easy to understand why few brands have created Spanish language apps. The fact that some of the Personas most coveted by advertisers are under-represented in the apps in which Spanish speakers are most over-represented reinforces this point.
Given the size of the Spanish speaking population, the shift of consumer attention toward mobile, and the advertising budgets following them, this situation needs to change. Our results show that if marketers want to reach Spanish speakers on mobile they are going to need to do so using Android. That creates an opportunity – or really a necessity -- for Hispanic-focused advertising agencies to lead in creating high-quality brand promotion on Android. While Android users in the U.S. may be less affluent than iOS users, collectively they still have massive purchasing power, and world-wide, Android is even more dominant than it is in the U.S. Given the size of the Android audience world-wide, teaching brands to speak Android could pay big dividends well beyond those that can be generated from U.S. Spanish speakers.
Many of us at Flurry love Mad Men, but we believe that Don Draper’s advertising industry is ancient history. Don would probably mistake smartphones for cigarette cases and tablets for coasters. More importantly, sophisticated buyers and sellers in today’s advertising market are making decisions in real time based on masses of data rather than months in advance based on charm and corporate hospitality. Advertising buying is being disrupted by efficiency gains from real time bidding (RTB) and effectiveness improvements achieved by using big data to inform mobile advertising transactions.
Data-Driven RTB In The Mobile Space
Recently Flurry launched an RTB Marketplace that enables advertisers to bid for the attention of smartphone and tablet users one at a time. We are betting big on the trend toward programmatic buying for two reasons. First, it enables precision targeting that was unimaginable in the pre-digital age and is still uncommon. Buying ad exposures one at a time enables advertisers to reach precisely-defined audiences wherever they are and whenever they use their devices. That level of precision would always have been useful, but it is especially important now that consumer interests, preferences, and lifestyles have become so varied.
Second, RTB brings a new level of efficiency to ad buying. The whole nature of an auction means that an advertiser who is willing to pay the most to reach a certain type of person will earn the opportunity to do so. The price advertisers are willing to pay provides a clear signal of the relative value they place on a customer or potential customer.
In this post we share initial results from our Marketplace to illustrate the power of combining the price signals provided through RTB auctions with the individualized targeting capabilities made possible by big data.
Building A Mobile Audience One Person(a) At A Time
The chart below illustrates three important results related to the value advertisers place on different types of mobile users and the available supply of those people’s in-app attention. The items being plotted are Flurry Personas. These are groups of devices whose owners access particular types of apps more frequently than people using other devices do.
The size of the bubble associated with each Persona represents supply, or the relative number of auctions for the right to serve an impression to a device in the Persona. Of the Personas shown here the greatest number of available impressions were for Casual and Social Gamers and the least were for Fashionistas and Food and Dining enthusiasts.
The x-axis, clearance rate, shows the percentage of auctions that had a winning bid, resulting in an advertiser displaying an ad on a device. As can be seen by looking at the right bound of the x-axis, less than half of the auctions had a winning bidder. This is a normal and expected result in RTB auctions. Reasons for auctions not clearing include price floors being set too high, bidding technologies used by advertisers’ representatives responding too slowly, some publishers being able to sell their ad inventory for higher prices elsewhere, and advertisers being uninterested in the inventory some publishers offer.
The y-axis shows the average effective cost per thousand impressions (eCPM). Even though these impressions are sold on an individual basis that is still the common pricing metric.
Fashionistas And Foodies Command A Premium, But Hipsters And Music Lovers Are Cheap
Examining supply, price, and clearance rate together reveals a lot about the state of play in the mobile advertising market. First, the fact that Fashionistas and Food and Dining Lovers are in the upper right corner implies that those Personas are of greatest interest to advertisers. It seems logical that those would be desirable psychographics, but the limited supply of ad inventory for those Personas also helps explain why prices and clearance rates are high. It means there are opportunities to generate mobile advertising revenue by publishing apps and content that attract Fashionistas and Food and Dining Lovers.
At the other extreme, Music Lovers and Hipsters have relatively low clearance rates and relatively low average eCPMs. While the supply of impressions for these groups within our Marketplace is not particularly large (as shown by the medium-sized bubbles), we hypothesize that people in these Personas are fairly easy to reach outside of our Marketplace because music fans spend a lot of time in music apps and many apps attract mobile-savvy Hipsters. It also makes sense that advertisers compete less aggressively for Music Lovers considering how inexpensive music is now compared to the pre-Napster era.
The overall diagonal pattern formed by the personas shows that the market is working efficiently, as expected. How do we know that? If a Persona had a high clearance rate but a low average eCPM we would expect advertisers to bid up the price to secure inventory. The fact that there are no Personas in the lower right corner shows that is exactly what has happened.
A position in the upper left corner of the chart means that auctions to advertise to that Persona have a high average eCPM given their rate of clearance. Here, we would expect publishers to drop their floor prices to achieve higher clearance rates. The fact that there are no Personas in the extreme upper left corner suggests that is also happening. There are some Personas with positions approaching that upper left corner: News and Magazine Readers are the most extreme example. We see two possible explanations for why those publishers didn’t drop their floors in search of higher clearance rates. One is that some of those publishers are able to sell impressions that don’t sell through our Marketplace direct or to use them themselves (i.e., to promote their own properties). The other is a policy of keeping prices above a certain level to maintain a premium image even if it means sacrificing short-term revenue opportunities.
Power Lunches Are Losing Out To The Power Of Data
RTB moves at lightning speed. A publisher can shop a single impression in the nanosecond before the winning ad appears. Compare that to the speculative, mass-market approach of the Upfronts, and it’s easy to see that advertising buying is likely to be completely disrupted by RTB.
The Persona-based targeting described in this post demonstrates the power of data to inform each bid. Advertisers no longer need to make buying decisions based on stereotypes about which types of people are interested in what type of products or content. They can define their target audience precisely, and aggregate that exact audience efficiently impression by impression. Mobile also contributes to this type of precision targeting since smartphones are highly personal devices loaded with apps that reveal much more about the person looking at the screen than standard demographics ever did. The long held promise of digital advertising is finally being realized on mobile.
All of this leads us to believe that advertisers or publishers who want to do things in the old way may be better off kicking back, pouring themselves a drink, and watching an episode of Mad Men instead of entering the fray in the mobile advertising space where data-fueled RTB is sure to win.
Just five years ago, PCs reigned supreme and so did the US software industry. In 2008, U.S. companies produced an estimated 65% of all PC software units sold on a worldwide basis.
In only half a decade, smartphones, tablets, and perhaps most importantly, apps, have changed the nature of the software industry. In this post we look at where apps are being developed and used and discuss the implications of that for the Post-PC Era software industry.
More Apps Are Now Being Created Outside The U.S. Than Inside The U.S.
Let’s start by considering where apps are being produced. The chart below shows the percentage of the apps that were recording data through Flurry Analytics as of the start of June in a given year broken down by whether they were created in the US or in another country. As shown in that chart, even in 2011, only a minority of apps were created in the US. By June of this year only 36% of the apps we measure were made in the U.S.A.
U.S. Made Apps Still Dominate App Engagement, But Their Share Is Slipping
Of course, some apps enjoy much greater use than others, so we next considered how the picture changes if apps are weighted by total time, which takes into account both user numbers and engagement. Once time is taken into account, things look considerably better for the U.S., suggesting that, on average, user numbers or engagement are greater for apps made in the U.S. than for apps created elsewhere. That makes sense given the size of the U.S. population, the fact that it was an app pioneer country, and the number of English speakers in other countries who might be able to use U.S.-made apps without any localization. Nonetheless, even the weighted percentage of apps made in the U.S.A. has dropped in the past year.
Use of Local Apps Is Strong In China
This should not lull U.S. app developers into a false sense of security however. That becomes evident from examining where the apps used by people in particular countries are made. That’s what the chart below does, starting with the United States. Nearly sixty percent (59%) of the time U.S. users spend in apps is spent in apps developed domestically, meaning that more than 40% of the app time of U.S. consumers is already spent in apps developed in other countries.
And while U.S. made apps are used elsewhere, unsurprisingly, people in many other countries spend a significant amount of their app time in apps developed in their home countries. For example, 13% of the time spent in apps in the UK is spent in apps made in the UK and 8% of the time spent in apps in Brazil is spent in apps made in Brazil. But as is so often the case, it’s China where things get really interesting. Nearly two-thirds of the time spent in apps in China is spent in apps made in China. U.S. made apps only account for 16% of total time spent in apps in China. The size and growth rate of the Chinese app market imply that the worldwide share of time spent in apps that are produced in the U.S. can be expected to contract further.
Translating apps and adapting apps to make them culturally appropriate is necessary in a country such as China to get most people to download and use most apps made elsewhere. Until recently, rapid growth in countries that didn’t require that type of effort meant that many developers based in the United States probably didn’t want to bother. With growth in the smartphone, tablet, and app markets in countries such as the U.S. slowing and a lot of remaining room for growth in countries such as China, some developers may now be reconsidering that position. It will be interesting to see if many can successfully adapt their apps for world markets.
While many U.S. app developers are just starting to think about globalizing their apps, it has been a near necessity for developers in some other countries from the beginning. Consider the situation facing a developer in a small country where the local language is not one of the world’s dominant languages. Unless they create an app with global appeal (e.g., a flashlight app), or that can be adapted to local markets relatively easily (e.g., translation of a weather app), they are likely to end up with very few users. That is a problem since the time required to develop an app for a small number of users is no different from that required to develop an app used by a large number of people.
App Developers In Other Countries Have A Head Start Globalizing
Creating global or localizable apps turns that problem into an opportunity. The chart below suggests that developers in countries such as Finland, Denmark, Bulgaria, and Slovenia are taking advantage of that opportunity. The numbers in the chart reflect the impact of app developers in a given country by taking the total percentage of time users worldwide spend in apps developed in that country and dividing it by the total percentage of apps developed in that country. Note that this is equivalent to taking the 2013 percentages in the second set of charts in this post and dividing them by the 2013 percentages in the first set of charts. For example, 70%/36% = an impact of 1.9 for the US. A metric of one or greater indicates that, on average, apps developed in a given country command a disproportionate share of time. The bigger the number, the greater the impact of apps developed in that country.
Given the dominance of English, the tendency of U.S. cultural products to spread internationally, and the fact that much of the app economy developed in the U.S., it’s not surprising that the metric for the U.S. is greater than one. China’s large population and the difficulties developers outside of China face developing for that market also make its high impact somewhat expected. What’s much more interesting is that the other five countries that have an impact metric of one or greater are all small, and four of the five speak relatively rare languages.
Globalization Of The App Industry Can Be Expected To Continue
This, and the low cost of app development imply that the app part of the software industry has the potential to become truly global. For example, as of June of this year, developers in twenty-three countries contributed at least 1000 apps to the more than 350,000 apps Flurry measures worldwide.
There are already examples of worldwide app hits that were developed outside the United States. For example, Angry Birds was developed by Finland based Rovio and has become a worldwide success with hundreds of millions of downloads. The same applies to Cut the Rope, which was developed by Russia-based Zepto Labs and Fruit Ninja which was developed by Australia-based Half Brick Studios.
Three key factors suggest that the markets for apps and app development will become increasingly global.
First, the App Store and Google Play take a lot of friction taken out of software distribution in the app world.
Second, the market is already very large and growing quickly in many parts of the world. In July of 2013, Flurry tracked 1.15 Billion monthly active devices. As our results have shown, there is still some ‘home field advantage’ for local developers, but in theory developers anywhere can create apps for users anywhere.
Third, the cost of development is still relatively inexpensive, especially if you factor in the average salary of software engineers in countries like the Philippines, The Ukraine, and Brazil. The cost of promotion is rising, but total costs are still a fraction of the cost of development, packaging, distribution, and marketing packaged PC software.
In short, geography is becoming increasingly irrelevant in the Post-PC Era.