Appsfire rolls out a “PageRank for apps” to help separate the wheat from the chaff

We’ve covered app-recommendation service Appsfire on more than a few occasions in the past – for example when it launched new iOS and Android apps, and when it rolled out its App Booster SDK earlier this year. And this week, the company is introducing another new way of knowing whether an iOS app is worth your time or not.

Just to recap, the underlying concept behind Appsfire is a simple one – it helps user sift through the app rubble to find the golden nuggets. Its new App Score metric is the result of two years of work, analyzing hundreds of millions of data points about the App Store. “Some people will call it the PageRank for apps,” says Appsfire co-founder Ouriel Ohayon.

What’s wrong with the current set-up? Surely there are plenty of systems in place to separate the wheat from the chaff?

“Discovery in the App store is broken because it is mainly based on two factors that are not related enough to the quality of the app,” says Ohayon. “The rankings in the App Store, and the ratings of an app – both are far from enough. Rankings are made for two hundred apps, but what about the other 630,000? So many great apps never make it to the rankings. Should they be ignored? Every great app deserves to be discovered and appreciated.”

Whilst Ohayon says that rankings, ratings and reviews aren’t irrelevant, they don’t really tell the whole story. “They are good signals, but they are only some of the signals you need to consider when you evaluate an app,” he says.

The Appsfire App Score

“When we created Appsfire our mission was to help users find the greatest apps whether they are highly rated or not,” continues Ohayon.

“Over time, we built a rich database of every single app computing in near real-time, millions of data points, to eventually know whether an app is great or not. We call it AppGenome. And the foundation of our AppGenome is our quality score.”

Less pontificating: what exactly is App Score?

It is a single number between 1 (Cr-app) to 100 (masterpiece). Users don’t have to know all the details behind it – the same way they don’t have to know why search results appear top in Google.

It’s a dynamically computed quality score, drawing on dozens of parameters each day and presenting, via a single number, the relative quality of an App. In terms of the data points used, these can be grouped into three broad categories:

• The Ranking Score: Evaluates the ability of an app to stay visible and get consistent, quality visibility – the ‘air time’ of an app. So this is not an app getting in the top 10 or top 50, but its consistent presence in a visible ranking in all stores. This means up to 123 in-country stores globally.
• The Rating  Score: Evaluates the consistency, frequency and velocity of good and bad ratings an app receives across the board.
• The Developer Score: Evaluates the reputation and success of a developer across the apps they have created over time – “Call is KLOUT for developers, if you prefer”, says Ohayon. This helps to identify the potential of an app based, before it has gained any traction whatsoever.

The last point there is an interesting one for sure, in that developers are rewarded over time for consistently good work. And conversely, those with a reputation for building bad application will be marked down the way.

App Scores also factor in other non-App Store external parameters, such as mentions on Twitter and Facebook, references in key publications and on review sites. “As with Coca-Cola, the key here is not just the ingredients but how we learn to mix them to find the appropriate balance and how we render it for our users,” says Ohayon.

Ohayon cites one example to illustrate the potential for such a rating system. MoveTheEggs has no ratings in the Israel App Store. So…is it any good? It could be excellent, it’s just that nobody has bothered to play it yet. But with a score of 83, you can be reasonably certain that it won’t be complete garbage.

Gaming the system

All good systems can be gamed…right? Well, right. But Ohayon says it won’t be easy to get one up on it.

“A developer would have to game the rankings, ratings, Twitter mentions, Facebook Likes, review sites across all of his apps, in a consistent manner over time,” he says. “You can game a ranking for a day or two. It is really difficult, or near impossible, to game all those parameters at once – all the time. This is where App Score is powerful.”

He adds that they’ve also introduced special safeguards that highlight suspicious activity around an app, for example, a 1-day-old app suddenly receiving thousands of ratings. “We also spotlight apps that have a limited value, or that are clones,” says Ohayon.

Looking to the future, App Score will be opened up to third party services via a private API, but for now it will be limited to Appsfire itself.

Oh, and all you Android users out there are maybe thinking “what about us”? I know I was. We’re told that the system will be rolling out to include Android apps in the coming weeks, and a number of new parameters will be introduced too.

So, App Score is a great new feature for sure, and it has the potential to become one of the key barometers for the true value of an app.

➤ Appsfire