Facebook App Ad Experiment Results

http://txt.jazzychad.net/gist/7634378

by jazzychad • 13 years ago

20 13 13 years ago

txt.jazzychad.net

lalos• 13 years ago

any reason on why did you choose Women in the United States Aged 35 and older? would be nice (and more expensive) to compare results with other target audiences

kunaalarya• 13 years ago

Probably the reason why: He has a word puzzle game, and that's the typical demographic for those games.

monkey_slap2• 13 years ago

I've always heard that the CPI with Facebook Ads was ridiculously high. Thank you for posting the results of your experiment. 0.009% installs is amazingly low to me, but I'm not too familiar with ad-based installs for iOS apps.

I've weighed heavily on word-of-mouth and just "quality" (website, app, reviews, etc) instead of CPIs.

Could you elaborate on how you chose your demographic?

lnanek2• 13 years ago

The one guy I know with an app, who directly makes more money from ads than they cost, has an expensive outdoor GPS app, not IAP. One thing I've noticed is that web apps are ridiculously cheaper than mobile ads. So even if your conversion sending people to a web site instead of an app store was only 10% as much, it is still more cost effective. Personally, I've only bought ads whenever I'm trending in app stores and want to stay in the trending category, which nets a lot of users. Although I often do ad exchanges/trades/or market in other apps of mine.

monkey_slap2• 13 years ago

I've heard the ad exchange/trade/etc before, but I'm curious how others actually implement this. Do you make an iAd-esque banner and advertise your other apps? Do you use an existing system/framework?

If its your own implementation, how do you manage/track impressions?

austinl• 13 years ago

I would be interested to hear the results with the free/IAP app. I've read in a few places that even a $.99 app loses 90% of potential installs right off the bat, so I wouldn't expect much with a $30 campaign.

cclogg• 13 years ago

Yeah, running ads with a paid app is a losing battle it seems like. Even with a free app, it's going to be tough because your app has to then make more per user than you spend on ads (this is the whole Zynga, Clash of Clans, etc model). And they do mass ad buys with crazy optimization and marketing teams...

blackysky• 13 years ago

As your first campaign you should try to catch a large audience like people in US over 18.. that's it ! and test your ad copy first... then base on your data you can decide the perfect demography for your app.. Are you going to blow money on the first second and third iteration ... the answer is yes but at least you will know exactly what's work best for your app on facebook...Ad campaign optimization is not easy. My best advice is to use a great ad tracking tool... you must track everything... no guessing game here... Finally a small variation is you copy and/or app presentation could have drive better results with the same $30.....don't give up :)

hisyam• 13 years ago

Can you recommend which ad tracking tool to use for Facebook Ads?

nl• 13 years ago

$6/install is high, but could be made to work if your app had a viral component that encouraged more than 9 installs (of course, realistically you'll need a free version to have anything like that level of viral installs.)

Also, what happens if you raise your price? Some experiments have shown that the install rate doesn't decrease dramatically between $0.99 & $1.99

carbonatedmilk• 13 years ago

So now you know your effective CPA for the ad you ran. You can count on a 3.7% install rate. Your app costs $1 For every 100 people who click your ad, 3.7 of them buy it - Making you $3.70 You need to pay less than $3.70 per 100 clicks

So set your CPC for that ad at 0.036 and write another, better ad. Want to cut down on spurious clicks? Make sure it's obvious it's a paid app before they click, and make sure you're talking right at what they need/care about

Mitchella• 13 years ago

Here are the problems with only spending $30 on ads.

#1. Ads are experiments both is where they're located but also the text, target, etc. IE: you didn't spend nearly enough and try enough possible ads to get anywhere.

#2. You're only charging $.99 - This just flat out doesn't work as a model unless you're a novelty that goes viral. Even then though, the majority of apps that achieve this are running at a $.99 price point in order maximize downloads and not revenues because they capture an audience to their brand or via social networks and eventually market expansions, in app purchases, or other materials (the moron test is a great example of this).

Like another user said your purchase rate isn't going to drop horribly when doubling to $1.99. Once a user has made the decision to purchase an application, by which I mean spend any amount of money, the difference of a dollar in the mind is minimal. It's the hurdle between free and paid that creates a huge barrier in conversions.

Example: Currently you have ~4% conversion rate (with that demographic at least). This means that at $1 price point (and because of apple fees, $.70 profit) you'll need to spend less than $.027 per click to the page to break even.

Now under the assumption that you drop down to a 3% conversion rate when doubling your price. You'll make $1.39 per purchase, meaning you can spend $.042 per click to break even.

Lastly, while still on this note... you're unable to ever offer a 'sale' unless you'd like to go temporarily free. Sales, while not a huge deal for applications that no one has really heard of... they do still drive people to purchase through impulse if they were already thinking about it.

Warning: While I say this and it tends to be true... test it... always test it! Play with different prices and see what happens.

#3. Offer upgrades/in app purchases. If someone has already come to pay $1 for the application, they're likely to invest more into the game.

marojejian• 13 years ago

Flagged for broken link...

Facebook App Ad Experiment Results

Related Stories