AdWords Click Fraud Reduction - Competitor Clicks

Some low budget advertisers fear that if a competitor clicks on their adverts early in the morning, that the budget will be exhausted and they’ll have no spend for the rest of the day. We’ve not seen this pattern in any of our clients’ accounts, but it is often asserted in the AdWords Help forum that this is a serious concern. There’s a general buzz about click fraud at the moment.

We’ve had a think about this specific early-day-competitor-clicking problem and we have some ideas to avoid it. You’ll need some background information first, but we think this idea will work effectively, and doesn’t offer a simple response for a competitor who is behaving aggressively.

At midnight, Google starts a budget for a new day. Unless you’ve deliberately told AdWords to do it differently (accelerated advert serving), Google will ration out your adverts over the day. It divides your budget by what it thinks you could spend if you appeared every time, and then uses that ratio.

Example:

• You offer $10/day
• AdWords estimates $100/day
• Google serves your advert 1 in 10 times

OK, so that was easy. Now, your competitor comes along at 08:30 in the morning. Fires up a browser and does some searches, clicking madly on your adverts. He only gets to see you 1 in 10 times. However, the CTR you get is surprisingly large. Google recalculates… If you are getting this type of CTR, it’ll need to throttle the adverts even more… So after a little while, you’ve reached 1 impression in 100 opportunities and your competitor goes off, satisfied that you are dead for the day and will get no real visitors.

Chances are that your real targets for search aren’t searching in volume at that time of day. Most consumer markets heat up around lunch time. You can run an “hourly” breakdown to see what volume of impressions you get exposed to. With a limited budget, it isn’t a perfect representation of the daily profile, but it’ll give a feel for when most users are online.

Here’s the trick. Make your midnight, after the start of the day, before the real users come online. That is, if you know that the search peak starts at 11:30 in the morning, local time, make your accounts’ midnight, 11:00 local time. That way, the best that a competitor can do first thing in the morning, is to attack your budget after most of it is spent. If you make the accounts’ midnight, your noon, then you are well into the customer search day, before you start spending your daily budget. Your competitor can’t get access to your unspent budget, but waits until the budget is nearly exhausted, towards the end of the accounts day.

This needs a diagram… or two:

Those diagrams should make it clear. By offering the budget to real users, first, the competitor only gets to see the tail end of the budget. You can spend more on your clients and less on competitors. And your competitor will find it hard to predict when your budget day starts, making it difficult to target a time when the maximum damage can be caused.

Following a suggestion by John Krystynak on Andrew Goodman’s SEM2 Forum, can you do better? Maybe. If you use Google’s “Accelerated delivery”, you no longer have the adverts dribbled out, but fed as fast as people click. The risk is that many people do the same search repeatedly, so you may have the advert clicked on a few times, outside the limits set by Google’s Invalid Click Detection for double clicking and below the rates needed to trigger any other invalid click criterion (hypothetical speculation here - Google does not publish how their invalid click detection works). However, if your audience is conservative with multiple clicking, then you may be able to exhaust your budget when likely buyers are online and before the competitor comes online to attack.

Caution: You can only set your timezone for your account - the “midnight” that you want - ONCE for an account. Get it wrong and you’ll need to set up another account. Adjusting the “accelerated delivery” can be done multiple times in a day (do it once every other day or every other week to try and measure the effectiveness, though).