Facebook click fraud? | Pranav Bhasin

Last week, we launched a couple of new features related to photo discovery on lifeblob and I decided to run a campaign on facebook to drive some traffic and see how people are responding to the new stuff.

Today, I looked at the reports and I was shocked to see the rampant click-fraud going on in facebook. And based on what I hear from others who have run similar campaigns, facebook seems to be least bothered about putting any checks on it.

Here is a report from the facebook dashboard for the campaign between 6th Nov to 14th Nov. As you can see, the total number of clicks that we got charged for are 4469.

However, my google analytics dashboard for the same campaign only accounts for 2372 visits from this campaign. Where did the rest of them go? And it is not an abberation on one of the days – the graph below will tell you that actual visits are consistently off by almost 50%.

Have you noticed similar abberations in facebook ads? Leave a note in the comments and lets come together and get facebook to do something about this.

Also, pass this link around so that if any of your friends are running campaigns on facebook, they also become aware that facebook is charging them double for what they offer.

pranav

@Venkatesh - Very true. The fact that facebook is a walled garden should make it easier for them to prevent this too.

share flag

spam
offensive
disagree
off topic

@Abhishek - That is correct. However a 50% drop for a site accessible only after login seems very high. I've run google ads in the past and seen ~10% drop or so. Plus google actually has a mechanism of detecting click fraud and also refunds some amount based on what it detects. This system is not perfect by any means either, but at least it shows that they care.

@Umang - We do have the webserver logs, but it would be a lot harder to parse through them for validation. I'll wait to see if someone around has an answer before diving into it. As per facebook, Clicks: The number of times users have clicked on the ads in this campaign. As per google, Visits represent the number of individual sessions initiated by all the visitors to your site. If a user is inactive on your site for 30 minutes or more, any future activity will be attributed to a new session. As per these definitions, the only way a click will not turn into a visit is when the GA javascript does not execute. This effectively means that either the click was by a bot ( which as you said is unlikely ) or if the user clicked on an Ad and closed the page before it fully loaded. A reasonable gap between clicks and visits is understandable, but I'd be surprised if 50% of the users that are clicking on ads are closing them before page load and that too consistently everyday.

Venkatesh

This is very bad. A company like Facebook should definitely pay attention to these things and correct it at the earliest.

Abhishek Goyal

I think facebook dashboard shows clicks, and Google Analytics shows if the browser loads the basic htmls and parses the analytics scripts and executes the js [also defined as visits by analytics]. There is a difference between the two and there is a possibility of big drop. if you run google ads too, you will notice this difference in clicks and visits.

Umang

A few comments: 1. Do you have webserver logs? A simple group by IP is a great first-step analysis. 2. The GA numbers are for visits. Are they comparable with the FB clicks? What does FB mean by 'unique clicks' - same user, same session, etc? 3. I wonder why the discrepancy in clicks. It is as if the user clicked on the ad but dropped off immediately after the redirect. FB ads are javascript-based (?), so I'm sure it is not that the GA script is not running. 4. Most click fraud is by bots and bot-nets. I doubt that there are a lot of bots on Facebook due to needing to sign up and log in. And FB does (used to, at least) some verification at sign up. Who, other than FB, would benefit from clicking on your ads? Hope you find a solution.