Home » , , , , , ,

App Discovery: Is There A Method To The Madness?


Prashant is Co Founder of Signals , a stealth mode start up in the mobile app space. Prior to starting Signals, He was part of  Spice Labs. A hard core mobile enthusiast (addict as he often calls it) Prashant is founder of the Delhi Chapter of Mobile Monday, a Global community of mobile developers. When He is not thinking about mobile apps he loves to travel, write poems,
cook and read.He lives in Noida with a Macbook, an iPad, an iPhone and Two Android devices. You can follow him on Twitter @pacificleo. In this post he writes about mobile app discovery.

A couple of weeks back my co-founders at Signals got a pleasant surprise. An app which they had developed some time back, somehow made it to the ‘top downloaded’ section of Google Play. They were happy and confused at the same time. The question nagging them was: How?

As I write, the top 10 apps on my HTC device include a shayari (a form of poetry/ verses in India ) app, which is not the best in class, both in terms of design and content but still it’s listed ahead of Angry Birds and Temple Run.  Looking at the order of the list one can’t help but wonder how.

The answer to this “How” is a billion dollar problem being faced by the app ecosystem today: The question of app discovery. As number of apps keep on increasing, this is only going to get worse. I think abundance of content is not a problem in itself, for desktop internet also has abundance of content. But content makers on internet have some methodical approach to deal with the issue of discovery. The problem is far from being solved for them but there seems to be a coherence, a method to the madness. Unfortunately the same can’t be said for much of the app ecosystem . Most of us are flying blind .

Advertisement

When it comes to the app ecosystem, I feel that part of the problem is due to the lack of structural understanding of app stores. In Internet we had things like Page Rank, Meta Data, Index algo, Robot.txt, Nofollow etc. When it comes to apps, our understanding of such structural aspects is very limited and we don’t have enough conversations around this. In this note I would like to share my thoughts on the discovery issue and would request readers to chime in their opinion.

Is app discovery same as web discovery ? :

Yes and no. This question needs to be examined from the point of view of user behavior. In the early days of internet all search engines were directory listing sites with varying levels of Meta information thrown into them. Yahoo was the most successful of them all. Outside of word of mouth, there were two  ways  of ensuring discovery. 1)  Buying banner ads on the YahooFrontpage   or 2) Ensuring that your site was in one of the top few listings in the category. Users were not searching, they were traversing along the directory structure and most of the time they didn’t traverse beyond the first few levels. So it was crucial that you were on right side of the sorting method. Many in the industry belived that since sorting was alphanumeric so sites like 123india.com ended up being an accidental benificiary of this. It was the advent of google that changed it. In a post google world, all (indexed) content was equal. Discovery was dependent on pagerank. User could optimize their content for page rank via SEO and SEM .There was some downsides to it due to SEO spam but on the whole, the outcome was net positive.  Google took the control of discovery out of the hands of editorial staff or fixed (static) algorithm and put it into a system which was a little more democratic.

For apps, which stage of evolution are we in? App stores are equivalent of portals of the 90s – users are “browsing” and not “searching” yet. More than 80% of downloads come  from featured/ top listed applications. Meta Information like app description and meta data affect the ranking but we are not sure how and how much. The whole thing is a black box. There are some strong trends but a Page Rank equivalent of the app world is not in sight. So for the time being  presence management is critical in the app world. Let’s talk about the elements of presence management .

Understanding and exploiting silos within app stores:

App stores are not monoliths. There are silos around geographies, devices, languages etc.Here are some of these silos and ways to exploit them to our advantage.

Geographic Silo: What you see in app store in India can be very different from what you get to see in France. If your app is specially relevant to French users then you should highlight that by things like providing description in French, putting keywords in the language, localizing the content and interface in French. Your chances of getting visibility become high. Here is an interesting observation: Most devices sold in the Middle East and China have a vernacular input support .These are big markets for all app stores. Yet less than 10% of developers choose to localize their app descriptions in Arabic or Mandarin. My anecdotal research suggests that a little less than 30% of search queries on app stores are in local language and most of them are underserved . Therein lies an opportunity. In addition to that you can also try using country  specific  promotion strategy ( Free apps on St. Patricks Day ! )  or content strategy ( Holi themed Bubble Breaker game in India ).

Device Silo:These silos are around devices too.For example  iPhone 5 might be different from iPhone 3GS. If your app is using a specific feature of a new device which can be used by the platform owner to demonstrate the capability then they might be interested in highlighting your app. Examples of such things are BBM integration in Blackberry, Pureview integration for Nokia, Retina display support in iPhone , NFC in Samsung Galaxy S3 or Heavy duty processing power in  Nexus4. Game Your Video , Viddy, Socialcam  are all examples of the same. All these apps showcase video capabilities of iOS. All these were featured by Apple at some point . What you should remember here is that you don’t need to build a new product from scratch .You need to incorporate the new technology in some feature of your product and do it well. A case in point is the use of iPhone 5 optimization by Temple Run to make it to the ‘Featured’ app list in multiple geographies, as shown in the table below (via App Annie).

 

Categories:  One should be very careful about which category one puts one’s app into, for that will decide the competition . A kids’ education app may be placed in games, education or kids apps section. Each will have its own level of competition . It’s always good to launch your app in a low competition but relevant section to get some traction and then expand to adjacent, more generic categories. One can always switch categories over time. Take a look at the following graph of the app Annie to know how Draw Something used category switching to its advantage by juggling between various categories of the app store to enhance visibility.

Ratings & Reviews: All app stores use reviews as a component while deciding/ranking the apps. Both the number of reviews and their star ratings play a part but in general higher number of average reviews (3 star) are far more helping than few instances of exceptionally positive (5 star ) or negative (1 star) ratings . The thing is that for the most part getting these 3 star ratings depend more on nudging the user at the right time than the inherent quality of the product . Put a policy in place to show the user a pop-up requesting a review whenever he starts playing the third session of game in a row. It might be a suboptimal user experience but do it anyway (unless you are Angry Birds or Facebook ). You might have the best product but if you don’t request for a review at the right time, you won’t benefit from the latent goodwill of your app. Most of the users are too lazy to promote your app. Quick Thought: How many of you have written a review of Angry Birds?

App size: Some app stores have this policy where apps beyond a certain download size are shown only when the user is connected to WiFi. In the case of a country like India where wifi connectivity is not that common this can be fatal .

Pricing: Imagine you are the one managing the iOS App store. You have to choose the top ten paid apps to showcase in the ‘Feature’ section in the upcoming week. You have got a tie for the last spot. There is one app with a price of $5 and it has sold 10 copies and another of $2 which has sold 25 copies.  Both of them contributed $50 each to your earnings. Which one will you showcase?  Well Most of us will assume that it will be the $2 one because of higher transaction volume. But in practice the $5 app might be the one to be selected. Most app stores believe (and perhaps rightly)  that power is not in the content, power is in the spot (Exhibit 1: I downloaded that shayari app ).

So whatever they put on showcase, it will get some minimum number of downloads. So if the last spot in paid listing is valued at 2000 downloads a day, putting the $5 app there will fetch $10K and the $2 app will fetch $4K . Which one will you choose ? Your guess is as good as mine. The point is that pricing aggressively and shooting for volume is a good strategy for a publisher in the start but that might not be the best one for a store owner who has enough volume coming in but who wants to increase transaction value. For example look at the chart of the top paid apps on iOS today. The only new entrant in the list is BloonsTD5 (the one at #4). It is priced at USD 2.99 while rest of them are priced at USD 0.99 . And Bloons was released for the first time only 12 days back. How did that happen? I think in addition to being a good product, the right pricing played a part too .

 

 

Advertisement and  Discoverability:

Most App publishers see advertisement as a user acquisition channel. But there is one little known tactical aspect of advertisement which if used wisely can help you enhance visibility. This requires a basic understanding of how app stores divide their users into geographic clusters. Let’s say it’s cheapest to buy cost-per-download based advertisement in El Salvador, it’s slightly costlier in Turkey and it’s most expensive in USA. This costing is mostly a reflection of how much  an advertiser is willing to pay to reach out to those audiences. Publishers tend to buy ads in these geographies so that users acquired by them can be monetized later via high yield ads. Seems a rational thing to do,  right? Well, not always. Let’s assume that you have a freemium app which relies on in-app purchase and your experience says that conversion rate of free to premium is not dependent on geography . So where do you go to buy ads now ? El Salvador, right ? Wait! Not so fast! Let’s say that in the Apple App Store El-Salvador is coupled with US and Turkey is coupled with the Middle East cluster. Featuring is decided in a cluster-wise fashion . So 20K download in El-Salvador might not help you get first of the pack in the US cluster but 15K download  in Turkey can make you the top app of the week, thereby giving you visibility in the top list and more downloads for free. If managed well this can create a snowball effect. The point is, align your advert spend in accordance to the featuring cluster of your app store . Milk every download in every possible way.

Search keyword : Though majority of app store users browse app stores along the directory structure but a non trivial section of them also indulge in keyword search. These users can be captured through Keyword Optimization. Services like Searchman can come in handy here .

Social alignment : This is slightly tricky and you need to use your own judgement while applying it. The critical point is how much importance ranking algos assign to the social integration in app. Which means that if your app integrates with Facebook, Twitter, Google+, BBM and you invite your friend via them then it sends a signal to the app store owner that you like the app enough to recommend it to friends. Now this is not always true, especially if your app workflow requires you to sign-up with your Facebook credentials . That might create some false positives. Many believe that this factor was very helpful to Draw Something. At that time the ranking algo of the Apple App Store used to give exceptionally high weightage to social signals like how many friends you invite. This was something that worked to their favor. I assume that they did a similar hack with FB News feed too. But their primary user acquisition channel was the App Store only.

I can list some more but I am not sure if they will be applicable or effective by the time I publish this. These are not silver bullets. App store dynamics are changing on a daily basis.These tricks can’t substitute for the need and hardship of making a great product . It might (hopefully)  help you to smartly promote your great app to more people . If you know of any such trick, please share in the comments.

*

(c) 2012 Prashant Singh. Disclaimer: The views expressed above are those of the author, and not necessarily representative of the views of MediaNama.com.  

If you have an opinion or details to share with our readers, please do send across your contribution to nikhil AT medianama DOT com. Do take a look at our guidelines for guest columns

  • http://www.beingpractical.com/ Pravin J

    Hey Prashant, great article. Absolutely loved this.. some of the points you have highlighted need to be known to larger audience. 

    I repeatedly say – building a product and marketing it is two different things. Know of a startup who had placed their app in directory that was 15+ / Adult only; They never got discovered and eventually shut down the product after 10 months and pivoted! Lets connect – @beingpractical:twitter :)

    • http://twitter.com/pacificleo Prashant Singh

      Thanks Pravin. I am glad you liked it. There need to be more conversations around it in our industry.  We should surely connect and carry the thread forward . 

    • http://www.praval.com/ Praval Singh

      Just when I was thinking of connecting you guys, I see you here, Pravin. :-)

      • http://www.beingpractical.com/ Pravin J

        Thanks Praval :)

  • http://twitter.com/deepansh01 deepansh agarwal

    Brilliant article. Really insightful.

  • http://www.facebook.com/sidharthrao Sidharth Rao

    Good stuff Prashant! Really helped.

  • http://www.facebook.com/profile.php?id=1597502943 Mayuresh Shreeram Soni

    Prashant awesome stuff! I loved the way you approached the whole problem. App Discovery i agree is yet to evolve. Your comparison of app content with web content is really helpful. And i agree with the fact that app content is still more browsable than searchable. 
    Yet here is a thought, how about converting app content to web content. Today web content related to apps are videos, screenshots giving a pseudo feel of the app. 
    If apps could be played on the web, on social networks, on all devices it would help a lot in the app discovery process. Apps would become more searchable on the web. 
    This is how we(AppSurfer) approached the problem. We allow apps to be played in the browser, on facebook and across all devices. In addition to that we give apps a single app link across all devices, and this would help the app get ranked higher if developers use us as the sharing platform on the web.
    We hope to continue make a positive impact on the app discovery process, and make it better for the developers as well as the users.
    App Discovery as a process a small diagram:

  • http://twitter.com/rohithbhat Rohith Bhat

    Interesting article, esp. on advertisement and discovery. The Turkey example was great.

  • Anonymous

    Hi Prashant.

    With regard to App Discovery, you are discounting a major factor, that is the “Pitch” all major publishers make to Apple before they even get down to developing the App. That is why on launch all apps by major publishers get good amount of featuring as well as visibility, both globally on the US store as well as local stores.

    In my experience, App launches are highly dependent on Featuring opportunities on the App store, as they can make or break the fortune of an App, as visibility on the App store increases the probability of adoption, atleast by early adopters.

    With regard to Paid User acquisition, this trend is mostly led by Apps which rely on In-App Advertising for revenue, as they need to show a big installed user base to get the impressions and clicks.

    Also, App discovery on the App Store is highly different from in behavior to the Google Play Store. On the Google Play Store, due to the wide diversity of devices, device compatibility becomes the biggest variable that is responsible for the visibility and therefore, success of an App.

    On a macro level, what we need to understand is that Apple to a large extent operates the App Store as a curated showcase managed by itself, with user based curation in the form of Top App lists. 

    Alternatively, Google, as is their wont, have structured the Play Store to be based upon search algorithms that determine App visibility, and therefore, eventually, the Rankings. The recent integration of Google+ profiles in the App review system is also a step in that direction, as also seen in Google’s web Search, where User Generated Content is increasingly getting more weight-age in Google’s search & ranking algorithms.