Blog Post

It’s Time To Stop Using Downloads As The Key Metric For Apps Success

Stay on Top of Enterprise Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

The mobile apps industry is suffering from download-milestone overkill. Developers and publishers boast about how many times their apps have been downloaded on the various app stores, while analysts and the store owners themselves use downloads as a key metric to gauge the success of those stores. While these figures aren’t meaningless, they’re only a guide to potential success.

There are two new pieces of research that illustrate the problem. First, research2guidance, which has published a report comparing the average daily downloads per app on various stores, taking Apple’s App Store as the benchmark. The report claims that “apps on Nokia’s Ovi Store had 2.5 times higher download numbers in Q2 2011 compared to apps on Apple (NSDQ: AAPL) App Store”: 160% more average daily downloads per app.

Microsoft’s Windows Marketplace was 80% up on the App Store by the same analysis, with RIM’s BlackBerry App World up 43%. Google’s Android Market was down slightly (5%), while stores from Samsung, LG (SEO: 066570), GetJar and Palm (NYSE: HPQ) were further down.

Here’s the company’s chart:

The calculations seem pretty transparent: divide the number of estimated daily downloads on each store by the number of apps. Yet this doesn’t even begin to tell the full story.

“Despite all the hype around the major platforms Android and iOS, publishers are still overlooking the hidden potentials of the niche players,” suggests research2guidance. Overlooking, or concentrating on an entirely different metric: how much money is being made on these stores? Not to mention how much it costs to develop apps for them.

Download numbers are no guide to how much money is being made. They don’t tell you how many apps are being sold, and they can only give an indication of the potential an app has for making money from in-app purchases and/or advertising.

This isn’t a pro-Apple anti-everyone-else viewpoint. There are plenty of companies trumpeting iOS download figures with no indication of how many people are actively using their apps, or how much money they’re making. Meanwhile, all the evidence suggests that while some apps can generate monster downloads on the App Store, many many more aren’t being downloaded at all.

The second current piece of research that filters into this discussion comes from analyst firm Ovum’s new report on mobile apps. It predicts that Android will overtake iPhone this year for the total number of app downloads – 8.1bn to 6bn – with the gap widening to 21.8bn and 11.6bn by 2016.

These are big, meaty figures that are likely to fuel many more “Android overtaking iPhone” headlines in the days ahead. Yet downloads only tell part of the story again: Ovum thinks that in 2016, iPhone will generate $2.86bn of paid app revenues compared to Android’s $1.5bn.

Or, if we’re doing the average-per-app thing: $0.25 per app on iPhone versus $0.07 per app on Android. A slightly nebulous calculation, since this is dividing paid app revenues by the number of total paid and free app downloads: and does not include sales of in-app purchases or advertising income.

The point is this: boasts and comparisons based on download figures alone don’t hold much weight. Perhaps it’s time to phase them out.

For developers and publishers, the key metrics should really be revenues, or at least active users, which provides a much better way of telling how well an app is actually doing.

The Facebook applications world is measured by daily active users (DAUs) and monthly active users MAUs), so why should mobile apps be any different – beyond mobile social games, which already use these metrics. Companies like comScore (NSDQ: SCOR) – through its partnership with mobile industry body the GSMA – are starting to track this, but there is just as big a need for developers and publishers to start talking in these terms.

And the App Stores? Developer payout figures are one useful stat, although so far it’s only Apple that has announced milestones like its first billion dollars paid to iOS developers. But the truth is more complicated, taking in the ratio of paid to free downloads, the success of advertising models, the ease and resulting adoption of in-app purchases, and a host of other stats.

This article originally appeared in MediaGuardian.

9 Responses to “It’s Time To Stop Using Downloads As The Key Metric For Apps Success”

  1. Downloads obviously translate to potential advertising revenues, just like how higher page hits on internet translate to higher rates on advertisements. Just like my finance Prof who just made fun of “hits” on internet, this article tries to make fun of downloads. Agreed, downloads by themselves do not mean anything, but those in business understand these metrics translate to $.

  2. RedMercury

    Wait.  These metrics were perfectly reasonable when Apple was ahead.  Now that other stores are catching up or surpassing Apple, these metrics aren’t useful?

    One of the things to also keep in mind is that the “paid download” argument isn’t quite fair as there are lots of apps–probably more on Android than on iOS–which are free yet receive revenue from in-app advertising.

    Quick!  We have to find a new model where Apple excels!  I know!  How about how much users get charged for the Apps?  That seems like a good one–since most Android apps are free/ad-supported, there’s no way to track that info for Android!  So we can compare revenue from sales against revenue from sales and Apple will still come out ahead!  Brilliant!

  3. Agreed, absolute download numbers are one dimensional, ignore other more relevant factors and lead to much useless cheerleading. Other factors, arguably more important, would include numbers of free/paid apps, app class ie. utility v time waster, revenue model, market penetration etc. etc.
    I do think it’s worth pointing out that, contrary to Ovums projections, Apple has already passed the 15B download figure(July ’11) after the 10B download countdown in Jan ’11(won by a British woman) and could easily achieve 20B by years end given a doubling rate year on year. So how do these figures square with Ovums figure of 6B for this year or that Apple has(probably) already exceeded the 2016 projected figure, half a year and 4 years ahead of their schedule?
    It certainly leads to accusations of misleading fudging and platform bias.

  4. I rather see the total downloads continue to be used as a measuring stick for apps rather than using profits made from their sale. One of the main advantages of android over iOS is how many great apps are free. Plus many great droid apps only ask for donations rather than making you buy an unknown app without trying it first. I bought a few apps for my itouch that really sucked, and had no way to return them for a refund.

  5. Well, I think we are missing a couple of key metric in this evaluation. 

    We know very well how many times an application has been downloaded, but we do not have any clue of the application life after that event. Life of the application on the phone, and usage are key metrics that are needed to evaluate your application success.

    True said you can get some of those informations from back end connectivity, if your application has some but it would be great to have those date coming from the App Store.