What is click fraud?

As I discussed in the last week’s introductory article on PPC, what is paid search and why do we need it, a company can bid on certain search terms on a search engine network, and therefore pay to appear higher in sponsored listings on the SERPs. 

Straightforward enough, but what if your costs are greatly exceeding your expectations or regularly coming from the same ISP and without a conversion? This could be click fraud.

The Econsultancy Paid Search Best Practice Guide describes click fraud as happening when: 

A third-party generates invalid clicks on a paid advertising link to drive up advertising costs with no intention of conducting business with the advertiser.

Who would do this to you?

  • Your competitor: The most obvious perpetrator. “I’ve started my own business, my competitor is ranked higher in sponsored listings, I’m going to nefariously click on its ads to drive up its costs for that search term and therefore push them out of the market, and then my own business can overtake or replace them in the sponsored listing”.
  • Site owners: Publishers themselves can click on their own website ads, and therefore make their own site more attractive for ad placement.
  • Customers:  Customers can accidentally add to click fraud when they regularly click through on paid search ads to regularly access a site, rather than using search or a bookmark.

How is click fraud committed?

There are two methods of click fraud. Automated and manual.

Automated click fraud uses software to repeatedly click on your ads. Manual click fraud is done by actual people, either from within a rival company or from outside the company hired to physically click through your ads.

What can you do to stop it?

Thankfully automatic click fraud is easy to trace, as it normally comes from the same IP address. Manual click fraud is harder to trace, especially as much manual click fraud could be construed as accidental.

Google has a click quality team, but is of course fairly clandestine about how they operate and monitor for click fraud, however Google has devised a three tiered system to deal with the threat.

  • Automatic filter. This filters out invalid clicks in real time. They don’t appear in your reports and you’re not charged for them.
  • Proactive Analysis. Google’s team investigates anomalies and fluctuations. If anything erroneous appears to have been carried out, your company will be credited.
  • Your own analysis. You can raise a query yourself with Google by filling in an online form and submitting information from your own investigation.

This third step is hugely important. The only way to have complete confidence in steering your costs away from the threat of click fraud, is to monitor the situation yourself, don’t just rely on Google or any of the other search engine networks you advertise on to police your traffic.

We recommend a five step best practice process for monitoring and reducing click fraud:

  • Identify metrics to audit.
  • Consider use of specialist software for tracking these metrics.
  • Audit metric values to create a benchmark.
  • Review current metric values against benchmark values.
  • When benchmark thresholds are exceeded present audited data to search engines.

There’s much greater detail and insight into the above processes in our Paid Search guide to help your business monitor and avoid click fraud.

Is Click Fraud really that much of a problem?

It’s hard to tell. Click fraud is a secret world of underhand tactics and unethical murk, so of course there’s no real hard data. How many pickpockets operate in the Wetherspoons next door? How many identity thieves rifle through your bin at night looking through your unshredded documents?

The answer to these questions depends somewhat on subjective analysis and how liberal or reactionary your politics are.

There are some statistics out there though, but do take these with a pinch of salt:

USA Today claim that in the first three months of 2010, between 17% and 29% of all online clicks are fraudulent. This is significantly up from Click Forensics’ claim that in 2009 the rate was 12.7%.

Microsoft in 2011 claimed that click fraud is ‘rampant‘ and has ‘incontrovertible evidence of dubious behaviour for around half of the search ad clicks and a third of the mobile ad clicks’.

Statistics seem to dry up post 2011.

The future?

If click fraud is to be eradicated completely then there is a solution. Forget cost-per-click (CPC) or cost-per-impressions (CPM), how about cost-per-action (CPA)? Where the advertiser only pays when a user has clicked through the ad and made a conversion, or completed a specific action, like filling out a form or signing up to a newsletter.

Seems fair, right?

Possibly. The downside is that revenue for the search engine networks would be significantly less, PPC represents the largest revenue stream for Google, so there would be understandable resistance, and goodness knows what the knock-on effect would be for ecommerce in general.

Perhaps there’s another solution? Perhaps we need some more concrete data on click fraud before we can even begin to gauge the depth of its affect and figure out ways to alter the current paid search system.

Has your business been affected with click fraud? Or do you have any more tips or research that may shed light on the topic? Then please let me know in the comments section below.

For more information on the future of paid search, read our post on what will Google’s paid search ads look like in 2014.

Further reading for beginners

During my first year at Econsultancy I’ve been making a point of writing beginner’s guides to any new terms or phrases I find particularly baffling, or that I might suspect other people may find baffling too. 

The following related articles should help clear up a few things…