Experiment with Mobile Advertising, AdMob.com
I did a little experiment with mobile advertising using AdMob.com. The result is rather dismal. I created an account at AdMob, and deposited the minimum amount, $50. I then composed a text ad (“Bitskis: Game for preschool kids!”), uploaded a little icon, entered the “Bid Range” of 4 cents, and hit “Run”. $50 was gone in 10 minutes, and here are the stats I got:
170,337 impressions
1,250 clicks
0.73% CTR
This comes out to be 4 cents a click. Out of 1,250 people who presumably clicked on the ad, only 1 person bought it. This is pretty hard to believe. The ad is specifically for selling iPhone apps, so anyone clicking on it should know what happens when he clicks (loads the app store page for Bitskis 123 with a “Buy” button), which means they must have some degree of interest in buying an app. There are many possible explanations for this dismal result:
- Technical problem: The people who clicked on it, didn’t actually get to the app store page. I read on line that some people experienced this problem with AdMob in the past, but I would imagine that they have fixed this problem by now. So, this is unlikely.
- Most of them might have clicked on the ad by mistake or through misunderstanding. Perhaps they appeared in contexts that were misleading. For instance, the users didn’t realize that the button was an ad, and assumed that it was part of the app they were using. So, as soon as their iPhones tried to launch the app store page, the users quit out of it. I think this is a pretty good possibility.
- Most of them were just curious about mobile ads in general, or weren’t even aware that they were ads, and clicked just out of curiosity (not curious about the content of the ad, but curious about the ad button itself.) In fact, I’ve done this myself. Since ads on mobile devices were still new to me, I clicked out of curiosity. I had no interest in buying or even learning about the app that it was advertising.
- My friend suggested that it’s also possible for some sort of automated bots to click AdMob’s ads. I’m not sure what AdMob is doing to prevent abuse. Some app developers might use such tactics to automatically click on the ads within their own apps in order to get paid for those clicks. And, AdMob might not be detecting such abuse.
- People simply hated our app even though they all saw the app page. Well, it’s hard to admit this, but it’s certainly a possibility.
I searched online for other people’s experience with AdMob and they all reported similar results.
- Is paying for Advertising worth it?
- Results from my Great Advertising Experiment
- Does AdMob or iPhone Ads Make You Money? Not In My Test
I’ve always wondered if advertising in general work or not. Now with the online advertising, the performance of any ads are easily track-able, and among the people I know, nobody has succeeded (particularly with Google AdWords), which is why I’ve never bothered to try it before. It appears to me that the business of online advertising is like lottery; they are selling hope. The vast majority of them cannot generate enough sales from the ads to pay for the ads.
But if you think about it, it’s common sense. If you can actually generate more sales from the ads than the cost of the ads, you can keep pumping money into the ads indefinitely without any risk. Hell, you might as well spend millions of dollars (borrow if you have to); you have nothing to lose because you will recoup the cost from the sales anyway. If it were that easy, everyone would be making millions.
So, the chances are; if the advertising worked for you, you probably could have done well even without it. Someone on the discussion board I mentioned above aptly said: “Go viral or go home.” I agree. It appears that the only way to make money with iPhone apps (or anything online) is for one of your apps to go viral. Sometimes, even a viral app needs to reach a tipping point before it actually goes viral. I would imagine that advertising can help push your viral app to the tipping point. But if your app has no viral potential, I don’t think advertising can help you.
It follows, therefore, that the online advertising business is essentially making money from people who are betting their money against all odds because they don’t know the odds. As I said above, it’s selling hope or dream, like lottery. For this reason, the accuracy of the tracking methods used in online advertising would eventually become a problem for the whole industry. Once the odds of this lottery becomes widely known, very few would bother advertising. Right now, they are making money from people like me who wanted to try it out. Once everyone wises up, it would be hard to make money from online advertising. This also means, it would be hard for the app developers and content creators to support their products through advertising. Right now, the money is coming from the inaccuracy or the inefficiency in the dissemination of the information about the odds of success. If everyone knew about the true statistics of online advertising, the advertising revenue would probably drop dramatically. But who knows? Despite the odds, lottery is still a big business (for the government).
One interesting exception to all this, is GOGII’s textPlus which is currently holding the #7 spot in Top Free Apps. textPlus offers SMS text-messaging for free. Most of us have to pay for SMS text-messaging, which is really annoying. So, textPlus actually saves us real money. GOGII must have a bulk discount deal for SMS messaging, and they just have to make sure that the revenue from AdMob is greater. If the cost of bulk discount goes up (which is unlikely), or if text-messaging becomes free for everyone, or if AdMob starts paying out less, this business model would not work, but for now, it’s an ingenious idea. Any app that can save real money for the users would probably do well no matter what, especially in this economy.
Update (6/17/09)
Be sure to check the comments; Ali from AdMob shared some interesting thoughts on this matter.
—posted by Dyske » Follow me on Twitter or on Facebook Page
Dyske,
I am disappointed that AdMob has not been as successful for you as either you or we would have liked. One of the reasons why we launched download tracking for iPhone apps earlier this year was to address precisely the kind of issues that it appears you faced when running your ad campaign — namely, quantifying the economic value that the AdMob advertising network provides to you as an advertiser. That transparency does come with a price, as your experiment illustrates, but the growth of our network to-date indicates that most advertisers who advertise on AdMob derive enough value from the advertising to continue to spend. And many, like you, closely scrutinize their ad spend to make sure that it is matched with an equivalent level of value or benefit to them as advertisers. We will continue to do everything we can to make the AdMob network as effective as possible in meeting all advertisers objectives, but, as you have discovered, there will be cases (and hopefully they will be limited) where the economics for whatever reason don’t seem to work. That may be due to seasonality, the offer in the ad itself or a myriad of other factors. One thing that you can also continue to expect from AdMob is even greater transparency on the ROI that advertisers are generating on the AdMob network and more powerful tools to help advertisers measure and make decisions on how much to spend on the AdMob network.
Please feel free to contact me directly at ali@admob.com if you have any further questions about how AdMob may be able to help you meet your mobile advertising needs and objectives.
Sincerely,
Ali Diab
VP of Product Management, AdMob
I tried this as well and can report similar results. I spend $100 over two days bidding at the lowest price I could. These were my results:
- Impressions: 422,315
- Clicks: 1,986
- CTR: 0.47%
- Sales: 4 over both days
My sales actually decreased- essentially I believe that ads had no effect though.
@Ali – I wonder if part of the problem is that these ads are not targeted with keywords like a Google ad would be. For instance, my app is a directed to the wine niche so i would expect better results if only wine drinkers would get to see the ad.
PS: part of the problem is that we have no stats from Apple. If you use AdMob you can get an idea of what your “conversion” is – how many people who read your app description buy it. But, there is no way to compare the results from AdMob to the natural results from people just visiting the Apple website. So, it could be that on a good day my app sells 8 copies, but that is out of a 1000, 5000 or even 1 million views of my app description.
Hi Matt,
Thanks for sharing your data.
Re: targeting
I think the logic is that AdMob does not charge for impressions; only for clicks. (Correct me if I’m wrong.) So, theoretically, there is no need to target because only those who are interested in your keywords would click. But that is only in theory. If the reason for the poor performance is #1, #2, or #3 in my list, then this idea does not help. We end up with people who have zero interest in purchasing our apps clicking on the ads.
Dyske, Matt –
Targeting actually does matter, especially to drive App Store conversion since a significant proportion of iPhone users on our network are overseas. So, if your app is registered in the US App Store, international users won’t be able to download your app. That is compounded by the fact that because our US iPhone traffic is so competitive if you select run-of-network targeting on iPhone, you will very likely only get a small amount of US iPhone traffic, which won’t convert. Which is why we always recommend selecting country targeting for iPhone app ad campaigns.
Ali
@Ali,
That’s interesting, but if your ad is in English, wouldn’t that automatically target the English-speaking audience?
But in my case, it shouldn’t have matter because my app is for small kids to learn how to count, and it’s available for all countries.
@Dyske,
Actually, no, since many apps, and particularly games, are written in English but played around the World. In the case of Bitskis, I suspect that since it is a (somewhat) language-centric learning game, the country (and, hence, language) targeting matters more than for other apps in the App Store, which is why you experienced a significantly lower-than-average conversion rate on our network.
Ali
Thanks Ali,
That makes sense. Basically what you are saying is that users who do not speak English are used to looking at banner ads in English, so they would click on them, but if the app requires that you speak English, they wouldn’t buy it.
That is an interesting point, and it is probably true with most apps, but I sort of doubt that it is true with Bitskis 123. We have 4 apps for preschool kids, and two of them are specifically for English. But, I picked Bitskis 123 for this experiment because learning numbers is universal.
But, your point brings up another question. In the ad, I used the word “preschool” which is not a commonly known word in non-English speaking world. So, it’s possible that many users thought “preschool” meant something else. (e.g. a preparatory school you attend before you go into collage.).
Well, perhaps I should give it another try, targeting US only.
You’re welcome, Dyske. I agree that a numbers-based learning game like Bitskis 1-2-3 is probably less sensitive to language-specific criteria, although Arabic numerals are not universally used (ironically, they are not used in the Arabic-speaking World and many parts of South and East Asia, places where there is a significant amount of iPhone run-of-network traffic). I would be happy to personally help you build a campaign if you would like. Please email me at the address above, and I can help you with that.
Kind regards,
Ali
We had identical results. $100, 115,00 impressions, 400 clicks, 0 sales.
Looks like google is about to squash admob by entering the mobile app advertising space (just announced today.) I can’t imagine they could possibly be any worse, so here’s hoping.