: Paying for advertising on a cost per click basis - How to validate? We use various pay per click advertising agencies and are able to tune the traffic to our mobile sites according to various
We use various pay per click advertising agencies and are able to tune the traffic to our mobile sites according to various things.
The problem is, the number of clicks reported by the publisher never matches the clicks we see in our logs.
I'm told that it's natural to expect some drop off. Does anyone have any idea of what a reasonable drop off would be? 10%? 20%?
I do expect some drop off because being a mobile platform I guess people are more likely to experience timeouts, connection drops etc.?
More posts by @Jennifer507
You need to be measuring more than just clicks. You need to measure the effectiveness of clicks. Not all clicks are created equal. Which of these clicks would you like to get?
A click from a user that is ready to buy your product and just needs to find somewhere that sells it.
A click from a user that is looking for more information about what your product does.
A click from a user that made a mistake and clicked on an ad that looked like a navigation element on another site
A click from a user that got an incentive to click to win rewards on another site.
A click from a robot that is built to click on ads.
Each of those clicks will be counted as "1 click" by the sites on which you are advertising. Hopefully the ad network you are using has policies to prevent abuse (as in the last three cases).
In the end, what you care about is conversions. Measure how many people actually buy your product from having clicked on the ads. Evaluate how much you are willing to pay for ads in this particular ad network based on how the clicks perform. Clicks that you don't see can't convert, so measuring conversion will account for that.