*supercollider

“My iPhone app has had 200,000 downloads — is that good?”

By all accounts the long heralded “Year of Mobile” is finally upon us, and it’s raining apps. IDC is predicting the iPhone will have 300,000 apps by the end of 2010, and Android will rise to 50-75,000.

And we can count on brands to help fuel a lot of the growth and hype. Just in the last few months we’ve seen app releases from Van Cleef & Arpels, Piaget, Gucci, Hilton, Polo Rugby, Jack Daniels, Weber, Audabon, Jaeger-LeCoultre, Discovery, Time Out, Stella Artois and more. The iPhone bandwagon is picking up speed, and all the big brands are hopping on board.

This is not without merit either. Admob survey results have 70% of respondents say that they’ve downloaded a branded app already, and 60% said they’d be interested in doing so.

So for all this activity, how do we know what success looks like?

This post is a first shot at exploring the short and long answers to that question, by addressing the following:

• Context and mitigating factors: Like any other metric, measuring iPhone apps “success” is relative and depends on the context. What kind of app do you have? Who are you targeting? How have you promoted the app?
• Metrics: Just as microsites are not all about pageviews, iPhone apps are not all about downloads. Besides star ratings, sales and re-usage rates are going to become increasingly important metrics for brands.
• Benchmarks: Understanding that context makes benchmarking problematic, it’s still good to have some guidance. I’ve tried to find examples for each type of metric where possible to help give some idea of what to shoot for and what constitutes success for your effort.

To frame all that, let’s look through the eyes of the metrics, starting with the stat most brands are currently referencing when referring to success: downloads.

Downloads

People like a big, simple number, and one most akin to traditional forms of measurement, so it’s no surprise that downloads is what we’ve collectively defaulted to. Here’s the collected downloads to date for all the brand apps I could find who’ve publicly reported figures.

Now here’s the caveat — while you can use these types of numbers as guidance for what success looks like, when comparing with your own app make sure you keep the following in mind:

• Promotion
  Promotion makes a huge difference to the success of an app. You simply can’t compare downloads without factoring in the investment in promotional support and media.

  Nearly all of the most popular apps have received support, often through traditional channels including PR and brand websites, but the key is that this is usually supplemented with a significant mobile media buy through a network such as the Google-owned Admob.

  The effect of these media buys can’t be underestimated, due to the compound effect you receive from pushing an app into the top 50 listing, where over 40% of iPhone users find new apps from. Even just pushing it higher within a section listing has huge implications to your overall total.

  As an example, Sherwin-Williams was able to push their app from number 70 to number 18 in the utility section through advertising, resulting in a 500% increase in downloads.

• Availability
  Not every app is available worldwide. But the benefit of localizing and making the app available in global app stores is huge, as even the smaller countries will contribute downloads to your sum on a daily basis, and taken in aggregate over the course of the year it really adds up.

• Differing objectives
  Is a brand utility used every day worth more than a branded game played once or twice then deleted? You simply can’t compare the two directly, they have fundamentally different objectives.

  Despite that, many times marketers are still reporting downloads as the key success measure about the success of a brand utility instead of frequency of use, increases in brand loyalty or affinity, or other more relevant metrics (more on this later).

  As with any marketing activity, it’s key that metrics are selected and prioritized that are the closest aligned with our objectives.

• Counting updates in downloads figure
  I think most agencies and brands are using third-party reporting software, but if you are using Apple’s own report out of the box it’s not at all clear that Updates and Downloads need to be manually filtered. There’s a binary code in one of the columns that distinguishes between the two, if you don’t sort against that, you’ll count Updates in your Downloads figure, which could inflate your total downloads stat by a factor or 2 or 3, a huge difference.

• Duration
  Stats really accumulate over time for a popular app, and there is some advantage from having launched in the app store when there was 1k apps rather than 100k competing for attention. For example, Bank of America’s mobile banking and ATM finder app lanched in July 2008 as one of the first ever branded apps. By February they were up to 2 million downloads, and then in September they were reporting 3 million.
• Free apps are downloaded 7x more often than paid apps
  Most brand apps thus far have been free, notable exceptions being Kraft’s very successful 99 cent iFood app. According to Pinch Media’s research, the average number of downloads for a paid app is 9,300, compared to about 71,000 for the average free application.

Beyond the need to segment between brand utility and entertainment apps such as games, possibly the most important overarching point to make about downloads is that the long-tail is all important.

Stats vary, but it’s clear that like most media landscapes, there is a small number of blockbusters and a much larger set of apps who are ignored entirely. Appsfire has a nice graph that makes the point:

According to their data, only 20,000 of the 100,000 apps in the app store are being used at all, and 80% barely have any active users.

It’s very much a situation of “the rich get richer” in the App Store, as 60% of iPhone users find new apps by browsing the lists of Most Popular apps. So if your initial promotional blast gets you into the top 25 downloaded apps, you are much more likely to build massive compound downloads from that point on. If you never make it in that list, you will never see anywhere near comparable success.

In other data, Pinch Media reports that “the top 10% of paid applications average nearly 75,000 downloads. The second 10% of applications fell to a mere 9,232, slightly less than the overall average cited above. The third 10% fell by more than half that, to 3,849. A full 50% of all paid applications have an average download of less than 1,000.”

The average number of downloads for an app is a mere 25,000. Still, this is better than Facebook apps by a fair margin, where there are a few superstar apps with millions of users, but the majority of Facebook apps have less than 100 users.

Star rating

Star ratings are public, which means we have a much richer data set than downloads to play with here. And Johnny Makkar from Attention Digital has put together a comprehensive compilation of over 200 star ratings from branded iPhones application as a Google spreadsheet.

In Johnny’s summary of the star ratings, he notes that out of the 200 apps listed, only 15 have a rating of 4 stars or better.

That’s a pretty dismal stat. However if it’s any consolation, Neil Perkin and Seth Godin have talked recently about the problem with using starring systems as a reliable form of measurement, a lot of which comes down to the motivations people have for applying a star ranking in the first place, and the resulting polarization of results. Anecdotally I have a feeling that brands seem to be especially punished by this phenomenon, possibly because the expectations may be higher, or maybe people just don’t like the brands and see it as an opportunity to vent.

Re-usage + interaction rate

“Although novelty apps have had the most downloads, research shows, use of them fades quickly. “You play with it a few times, show it off in a bar, then you don’t use it much,” said Raven Zachary, president of Small Society, an app maker in Portland, Ore.”

This is the biggest problem with using downloads as the key stat. It’s up-weighted in favour of novelty apps, and gives no consideration to the biggest white elephant in the iPhone room: most apps that are downloaded are rarely used again.

“We think 100 engaged consumers are much more important than 1,000 downloads,” Ed Kaczmarek, director of innovation for Kraft Foods.

The North Face Snow Report app has been downloaded 300,000 times to date, but the more important stat may be that they report people are using it multiple times a day. Building a brand utility that becomes that valued by a customer has value outweighed the sheer impressions generated alone.

Achieving this type of usage is no mean feat though. AdMob usage stats from their universe of 2,239 apps indicates very few apps have a large active user base:

Pinch Media stats indicated that only 20% of people returned to a free app after the first day of download. Worse, by day 30 only 5% of people are still using the app.

My prediction is that re-usage becomes a benchmark at least equal in importance to downloads, especially in the case of brand utilities. This is stat is somewhat analogous to the “time spent” metric that has been more common in microsite reporting in recent years, but unlike that metric which can be skewed by people who have left their browser open or are simply lost in your site looking for something, whether someone comes back to your app time and time again is a pretty good indicator as to whether they actually like it and are having a good brand experience.

Though again, if you are developing a novelty app that’s promoting say a movie launch, maybe you don’t care about this metric very much at all, and would be happier with more downloads and less frequent usage than the reverse. It’s definitely case dependent.

Sales from the app

Mobile commerce is still far from commonplace, but remember that eCommerce was in that position a decade ago and is now a juggernaut.

eBay has reported that it’s app has resulted in $400 million of sales through the eBay marketplace. The question is whether these purchases would’ve happened anyway, and it would be interesting to see whether the app has resulted in an incremental uplift in eBay sales. Regardless, it’s clear eBay has opened up an important new channel for their customers.

The other reference of note is back in August Pizza Hut reported $1m in sales from 100,000 downloads of their iPhone app in just the first two weeks of release.

Sales of the app

Kraft was one of the first brands to test the idea of not just providing a brand utility as added value to the consumer, but to actually charge for it as well. Their $0.99 iFood Assistant app featuring recipes and tips and has spent most of the year in the top 100 apps chart, with Brandweek sources estimating the number of downloads to be in the seven-figure range. Weber’s seen the opportunity and is now upping the ante with their $4.99 On the Grill recipe app.

With the introduction of micropayments from within iPhone apps in 2010, in-application purchases of content is also a trend to look out for. In one recent example, in two weeks the vitaminwater Sound Lab application converted more than 64,000 downloads of 50 Cent’s latest single “Baby By Me.”

Other metrics

Although Downloads, Star Ratings, Frequency of Use, and Sales are the most common ways of measuring success of a branded app, there are other benefits to brands that can be achieved and measured.

A few examples:

• Audi claims their A4 Driving Challenge game has resulted in about half a million referrals to the A4’s iPhone Web site.
• Kraft says that 90% of people who use the app also go on to register at kraftfoods.com. The app is also helping Kraft reach men. “A strong percentage of iFood Assistant users are men,” Kraft spokesman Basel Maglaris says. “We’re appealing to a broader base of consumers than our traditional audience (of) women.”
• Lion Gate’s Stun-O-Matic app to promote its new action movie “Crank: High Voltage,” generated 800,000 trailer views from 2 million downloads of the iPhone app.
• REI has looked at their Snow Report app’s detailed usage and notes that “Over time this has translated into 750,000 minutes spent using the application and with the REI brand top of mind”
• The 9.8 million downloads of Barclaycard’s Waterslide game app have reportedly translated into 650,000 hours of brand engagement
• Of the nearly 300,000 who downloaded the vitaminwater appliation in the first two weeks, users uploaded 1 million remixes to the companion site at 50soundlab.com, a very high engagement figure.

What about sponsorship?

A new model is emerging in the iPhone landscape, which is the sponsorship of applications as an alternative to launching brand-owned apps.

Some examples of this have been Sprite and Zoozbeat, Smirnoff’s 6 month sponsorship of Time Out, and Swatch Distill.

Is a download of a sponsored app worth the same as a download of a brand-owned app? I’d argue it’s not. Just like Red Bull with their brand-owned events, it’s a different level of value for a brand to own something rather than to sponsor it. But how do we define the difference?

Last thoughts

A few points of summation to leave you with:

• Pick metrics that are aligned with your objectives
  Are you looking for awareness and campaign support? Downloads is probably your #1 objective. Are you looking to deepen your relationship with your customers and increase brand affinity? Re-usage of the app might be the best choice. Commerce? Sales.

  Other things to consider over and above sheer downloads…do your users like your app? Do they keep coming back to it? Is it shifting perceptions of your brand in a positive way? Is it driving sales? Figure out what’s important to you and measure that, don’t get too blinded by downloads alone.

• Downloads are important, and you need promotional support to get them
  Reach is still important, so approach it like anything else — don’t release and pray, promote it. Even viral films get their initial seeding boost, and iPhone apps are no different. Nearly every successful app has a solid promotional strategy behind it.
• Don’t take your ratings too personally
  Out of 200 brand apps, only 15 were rated four stars or more. Focus more on the actual qualitative feedback, look for ways to continually improve the value and UX of your app, and ultimately look at re-usage stats. If people keep coming back, there’s a good chance they like it.

• Downloads without re-usage is a missed opportunity
  95% of apps aren’t being used after 30 days. Getting an iPhone app on a customer’s phone is a major opt-in invitation to what could be a long-term relationship channel. Unless you are deliberately looking to only support a one-off campaign, think about your apps as comms and engagement platforms, which should evolve and develop over time, continuing to add value. Get that strategy right, and you can in the 5% of apps that stay active after thirty days, and you’ll have a rich portal into your customers lives for months or years to come.

That’s it for now, if you have any more benchmarks or stats to add to the collection, or any thoughts on how to best approach setting and achieving your metrics, I’d love to hear from you in the comments!