How to double your app's downloads with keyword optimization | Hacker News Reader

How to double your app's downloads with keyword optimization

blog.discofingers.com

66 pointsmrborgen38 comments

Is this an advert for Sensor Tower? Certainly reads like one.

Is there any way to get this information (competitor's keywords, keyword app usage, and popularity) without paying $78/month? How do "Sensor Tower" get it?

Hi there, I am the author of the article.

I have no connection to SensorTower, other than a couple of mails I have sent to them, to suggest features and improvements.

However, SensorTower has been the core of our optimisation strategies, so it's just how I had to write this article.

I do think they are to expensive, yes, but I use them since I know that they work. Appcodes is cheaper, only 15 USD per month, but I haven't tried them in a while. However the found is commenting this thread.

Maybe he has something to add about their new upgrade?

Wish someone from AppAnnie/SearchMan/MobileDevHQ would join the discussion to.

Thanks for a great article. Do you know if the keywords used in reviews have any impact on rating? I assume not since it's the first thing to be gamed.

Thanks!

I am not sure, but I don't think so, for the same reason as you.

However, I have heard your overall ratings (and I suspect comments too) will affect your rankings for keywords in general. One of the best way of climbing the keyword ranks is by getting great ratings.

I spent a ton of time testing the Apple's search algorithm for AppCodes.com - didn't find any relationship between ratings and keyword ranks.

Here's a summary of what I found out: http://www.slideshare.net/kolinko/new-rules-in-app-store-sea...

Basically - neither downloads, nor ratings, nor description counts.

Thats interesting. How do you think they rank them then?

No idea - whatever we did, we never managed to move apps' positions one way or another.

I've also seen apps that have few downloads, few rankings and lousy reviews, but get positioned higher than competitors that are more popular and have better rankings. Although popularity seems to have some influence.

What I'm telling people is to see if they rank high. If they do - keep the words/phrases. If they don't - forget them, and look for some other ones.

That last sentence pretty much sums up ASO :)

We did a writeup after a whole bunch of statistical analysis of various factors that are correlated with higher ranking.

http://blog.sensortower.com/blog/2014/09/07/why-162727-apps-...

Of course, some of these things (like # of ratings, how long ago the app was released, etc) are likely correlated with other factors that are more difficult to measure in the app store.

Appstore optimization makes sense when you have 50 daily downloads but want to have 100. Relying on search results only wont make you lot of money there.

Anyway in most cases you can start with buggy SensorTower to find out what keywords will bring you more relevant traffic at lower competition. After than just keep tracking things with a free AppAnnie ASO tool.

Partly true. You can only go so far with keywords at a specific time.

However, the approach isn't totally unscalable, as you can work your way up to targeting more and more trafficked keywords as your app grows.

An exptremely important part of ASO is to keep on doing it every time you update your app.

I have not looked into the economics of app stores myself, but I imagine they are following a similar trend that google did. I bet there are a lot of opportunities for developers selling cheap long-tail apps (recipes, quotes, etc, i.e. "resources") to niche customers. But I have no idea.

There are indeed a lot of opportunities there. But they aren't as drastically as in the early days of SEO. Partly because Apple gives you so few opportunities to affect your rankings.

But a lot developers just ignore ASO, which opens up possibilities for those who take it seriously.

SensorTower is buggy and annoying, I much prefer AppAnnie & MobileDevHq. I compared[1] them a while back and ended up concluding that the free tiers of AppAnnie/MobileDevHQ are enough to do some basic ASO.

[1] http://www.androidb.com/2014/07/10-tools-to-help-with-google...

Thanks for the tip. Interesting article. I have found that once you get used to SensorTower, it works quite well.

what I mostly need when doing ASO is:

1. Traffic estimation 2. A fast way of researching competition for a specific keyword

SensorTower gives us both and results from updates suggests that there are a connection between their estimations and true life.

How do you experience AppAnnie/MobileDevHQ traffic estimation tools?

A nice writeup. As for AppCodes.com, you missed some stuff we offer:

We've got competitor tracking, support for other languages and data export (downloading csv with the search phrases+results). We also support the pre-launch analysis :) There's also an AppStore prediction tool that is quite unique compared to the competition ( http://techcrunch.com/2012/03/12/appcod-es-launches-app-stor... )

I'll look at that again and update it, thanks for the heads up. Maybe these were features you've added later, as I tested all those tools.

We would love to fix the bugs. Would you mind sharing your experience and what you found annoying?

We would love to fix the bugs that you have seen in Sensor Tower. What were the issues that you ran into?

But what nobody tells you about ASO is that these users are much more unqualified than the other users you get. Downloads coming from friend referrals or reading an article about your app are much more likely to give you a good rating or convert into a paying user.

Hmmm.. What do you base that upon? My stats tell a different story. The downloads are usually of high quality.

This of course depends wether or not you manage to keep the keywords highly relevant. And the technique I am describing usually forces people to be more specific and add more relevant keywords. (Rather than what most app developers do, which is adding generic words like "fun,game,kids,book,music").

Depends on whether you spam the store, or use words/phrases that are relevant.

Hi everybody!

Thanks for all comments, if anybody need some tips / advice regarding their own App Store Optimisation, feel free to contact me at:

per@discofingers.com

How much does a keyword in the title actually matter in ASO?

It matters quite a lot. More than the regular keywords. Therefore you should squeeze a few of the most important keywords in the apps title. But it shouldn't get to spammy, as I have a feeling that that scares people away.

If you have an exact match (i.e. "music maker") between search and title, you chances of ranking high increases significantly.

Have you tried SearchMan for analysing keywords?

Just slightly. Might be that they are equally as good as SensorTower. I just got used to SensorTower early on, and have been using them since.

Also tried Appcod.es for a while, but the seemed to stop improving their service.

Anybody else who knows other potential good tools for keyword optimisation?

Hi, AppCodes founder here - we recently made a major design update.

As for adding new features: anything specific you were missing?

Hi! Good to hear, would love to try it out. I liked the service when I tested it a couple of years ago.

However, what made me not return was your paywall, which is placed right after the login. In other words, I could do anything without paying.

Given all the free tools out there, this approach made me skeptical.

I am not an expert on SaaS, but I suspect it would be better to give people some kind of freemium option.

It really depends - there's been some posts on HN criticising the freemium approach:

- http://www.tylernichols.com/web-development/i-am-done-with-t... - http://blog.rapportive.com/is-freemium-right-for-you

When building AppCodes I decided to sacrifice the learning curve for the ease of use once the tool is understood.

This kinds of products don't work well when published with freemimum plans - people don't try to understand the product, skip the tutorials, sign up for free, spend 5 seconds evaluating the tool, and unless the learning curve is very calm - leave.

What I prefer is for people to take a while and make sure they understand the ASO process ( http://www.slideshare.net/kolinko/appcodes-app-store-marketi... ), and then - once signed up, to spend a while learning the tool. The expectation is that the ones that go through this will get a better experience afterwards. The ones that won't - there's a no-questions asked refund :)

Another thing with freemium is that it works best with products that people use consistently for months and years. While this might be the case for web SEO or for app sales tracking, in case of App Store SEO it really isn't. You want to spend 3-5 months setting up your position, and once it's all set and you can forget about it.

If I was running a freemium service, I'd have to find some tricks to keep bringing people back. I'd also be more inclined to gear the service towards the app developers with more than one app - because only those devs would bring in the profit.

I don't track our competition too closely, but the last time I checked, that was what they were doing. Publishing charts of your position, and sending you e-mails of your daily rankings (even though rankings in the app store rarely ever change), and moving towards Enterprise customers. They also tend to deliver an interface that's easy to learn, but not that effective really (e.g. they don't bother distinguishing between iTunes keywords, and App Store search phrases). At least that was the case with the ones I checked...

If you disagree, I'd greatly appreciate a reply :)

I see your point. Especially about the retention part. It's not really a prouct which people use for years. Though I wonder; could it be? What if you included daily downloads stats too? I use AppFigures for that, but I really with they would include keywords intelligence to, so that I didn't have to use so many services.

The thing is though, when I don't even get to see the interface before I pay, I would really need a lot of good reviews from friends before I sign up.

You should clean up your front page to, it's a but messy. Don't mean to be rude it's just my honest opinion.

Love your Slideshows by the way! :)

Thanks for the feedback. As for the daily downloads - as you said - we'd have to compete with AppFigures.

As for seeing the interface - did you see the demo? ( http://www.appcodes.com/demologin/ ). Not trying to make a sale here, just curious.

Thanks for the feedback regarding the front page. Can you be a bit more specific here? What seems too messy for you? (btw. e-mail support@appcodes.com if you'd like a free account as a "thank you")

When are you launching Disco Fingers?

Thanks for asking! We might send it in to Apple today actually. Our dealine was two weeks ago, but you know how it its :)

Do you think it looks interesting?

In hindsight, we probably should have published this article after release... Didn't even dream of getting on the front page of HN.

How about localisation?

That's extremely important, yes, and we do that too. We'll write an article about that later on.

In general, you should localize both keywords and screenshots (and maybe even title, but that depends) for the following languages:

- Spanish - Portuguese - French - Italian - Russian

We also do it for Chinise and Japanese.

The key here is to find translators who understand the game, which can be difficult if you are talking to traditional translation agencies. I find that some people on Odesk and Elance with marketing and translation experience do a great do at this. And they are quite affordable :)

What's worth noting is that in U.S. Apple seems to position apps both for american and spanish keywords.

In other words: if you have spanish words in your keyword list, Apple will consider them in the U.S. search.

I know, thats a nice little "secret". What do you think would happen if you added a bunch of engligh words in the spanish localisation?

Reply on news.ycombinator.com