On Tuesday, June 25, 2013 5:25:37 PM UTC-7, Nathan wrote:

> I told Adwords that they are wasting my time and not accepting my money, 
> and unless they can offer better suggestions, I'm not running any more 
> campaigns with them. Sad because this is the kind of thing you think they'd 
> be good for. 
>
> I feel duty bound to report, that after making this statement to Adwords, 
I did get a response from someone in Adwords support that said something 
other than "bid higher". 

They said that the campaign was screwed in some way on their end, and that 
I should create a new campaign with 'Display Network only - Mobile apps' as 
the setting, as before, and never ever use a mobile web site as a managed 
placement because it would screw it up. 

They also suggested adding image ads, 320x50 being the most popular size 
and the one I started with. I'll add more sizes as my graphics guy gets to 
them. 
So I was up and running again in July, and learned lots of great stuff. 
I've painstakingly selected 154 specific apps as managed placements.  
Adwords can track mobile app installs as a "conversion". It's not able, as 
far as I can tell, to distinguish which conversion it is, if you have both 
paid and free app installs as a conversion, for example. Be aware that it 
may take two days for such installs to be credited to their source click. 
Therefore, you would wait till August 2nd to evaluate the 

After a few weeks, I was eligible for using the Conversion (De)optimizer. I 
did one experiment with it, but did not let it run for long. It might 
stabilize after two weeks, but waste more money than I wanted in the 
meantime. 

So I've switched that off and just continued to micromanage the ad group, 
weeding out apps that gave me a lot of clicks without installs, and 
adjusting the bid up or down according to how some were performing. 
I am suspicious of CTR's that are too high, and I found that games were not 
a good place to advertise a serious app (YMMV). Even if the subject matter 
of the game and the useful app are similar, people are not in a mood to 
evaluate my app when in a game, and are probably just missing the fire 
button.
Overall, there was a 5% conversion rate - ie 5% of the clickers installed 
the demo app. 
I got a cost per install equivalent of $2 and could probably work it down 
closer to $1. Is that the right amount to pay per install? I'll have to see 
how the traffic performs. I've done an overhaul of my analytics, and see 
how much money and engagement this class of new users gives me. 

Meanwhile, I have run an ad group targeting keyword search. Most of the 
keywords are highlighted in red, because Google wants me to bid 
unreasonable amounts to be first page eligible. The fact is that some non 
app businesses stand to make a lot more money from those keywords, and they 
have discovered mobile advertising. In spite of this, I get some stray 
clicks once in a while. These convert at 18%, consistent with marketing 
studies that show intent based targeting (search) to convert much higher 
than contextual. The cost per install is about the same, since the clicks 
are so much more expensive. 

On the whole, 260 or so install of my demo app in a month isn't much when 
it is already getting 500 organic installs every day. Nonetheless, this is 
a laboratory of learning. It puts the marketing funnel that I mention in my 
classes in sharp perspective. I'll probably spend a chapter on paid 
advertising in my book, just because of all the learning it provides. 

I've found it is also a good way to evaluate potential ideas for ad 
supported apps. I can estimate how much certain kinds of apps are making 
using the Display Planner tool. 

Nathan

-- 
You received this message because you are subscribed to the Google Groups 
"Android Discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/android-discuss.
For more options, visit https://groups.google.com/groups/opt_out.


Reply via email to