We decided to give admob advertisement a try and after a week of testing, we would like to share our experience with you. The advertised application was our secure notepad Enigma Notes and it was us who was paying for the ads (link to our app was showing in other users phones, we were not making money off the ads in our apps). There certainly is a lot of rant about how bad admob is but we wanted to have a first hand experience:-). Before we tell you what the result of the campaign was and provide some real world numbers you all are certainly looking for, we would like to comment on some of the questions every developer might be asking and the answers we have come with.
Question #1: How much does admob cost? You have to put a minimum of $50 into your account prior to starting a campaign. It is quite a lot of money for a very uncertain outcome but for the spite of future projects we decided to cough it up.
Question #2: What should my daily budget be? The minimum is 10 bucks. We wanted to keep the test running for as long as possible so we went with the minimum. Some blogs recommend at least $200 a day for a campaign to be profitable. We strongly doubt that daily budget has any influence apart from the amount of clicks/impressions you can get to before you run out of dough. Generally speaking, if you estimate 1 out of a 100 people who see your ad will click it and then 1 of the 100 people who clicked your ad will buy your app, you need 100 clicks, 10 000 impressions to make a sale. The cost of your app must be higher than that of 100 clicks (in our model case it is $1, but if you factor in the tax and what Google takes from you for selling the app it is more like $2) and therefore if you spend $1 or $200 a day does not change a thing in terms of profitability per click as paying more is not going to bring higher quality potential buyers. There might be some advantages from better penetration and people sharing their love for you app with their friends but you would need to get into hundreds of thousands of downloads (=hundreds of thousands of $$$) to make any difference.
Question #3: What max bid value should I choose? ROI is your mantra here and the minimum bid should be proportional to the price you sell your app for. Because the application is sold for just one buck, we could only go for very low bids in range from $0.01 to $0.05 to keep the campaign cost-effective (did not manage to keep it cost-effective nevertheless:-). From our experience we reckon that majority of advertisers go with the minimum bid because increasing the bid to $0.04 or $0.05 gives you a way too big advantage over the rest of the herd and summons hounds of hell that will siphon your daily budget in matter of seconds.
Question #4: More ads with different text or one simple and same for all? We have decided to have just one ad as multiple ads with different text do not fundamentally change the outcome of the test. We might try it in future with another apps though.
Outcome: So, after a week, heaps of impressions and clicks but no visible influence of downloads. Moreover, admob system acts in mysterious ways quite often. Even though they have the bet opportunity to suck all the money of your daily budget, on four out of five occasions campaign worked for just about an hour or two (getting tens of thousands of impressions) and then stopped for the rest of the day. It was frustrating and we have no idea why this is happening. Also, admob has to approve your ad, be ready to wait for a couple of days. This must be a real pleasure for devs who need to start their campaign ad-hoc on Friday afternoon. No way guys, expect Monday at the best.
Before we give you the figures, we would like to point out that we have managed to sell our apps through Google Play without any problems before and therefore the actual outcome is not negative because of the poor quality of the app but rather because of the poor overall quality of the advertisement system.
The numbers from last three days of testing:
Total impressions: 47 402
Total clicks: 453
Total cost: $5.67 (not that bad, it was worth the fun)
Average cost per click: 0.01
Total increase of downloads of our app caused by the admob campaign: NONE 🙂
Total impressions: 11 608
Total clicks: 114
Total impressions: 3 408
Total clicks: 32
Total impressions: 32 386
Total clicks: 307
Total impressions: 8 134
Total clicks: 66
As you can see, the number of impressions varies rapidly. Courtesy of the strange behavior of admob, on 18th of December the Campaign stopped after just an hour taking mere 30 cents from us and not doing anything for the rest of the day. Just lazy to do anything probably. Same with the 17th but the magical stop came later during the day. 20th December ridiculous again. The click-through rates were OK, but we suspect that vast majority of the people who click the ads do it by accident (thus our very unfavorable estimate of click per buy) and do not even wait for the page to load. This our revelation is consistent with many many many blogging developers around the world so if you want to go with the admob, expect very high click per buy rate. The concept of admob is very nice but from what we have seen so far it is not working the way it should work for your benefit. Read a nice article by our colleagues at Codepoke too, they have had exactly the same experience.
Despite the outcome, we will continue testing and we will be updating this post so stay tuned. We will eventually try to promote the free version of the app and try selling it through this channel.
Cheers from hedgecog!