How to poison your competitor’s PPC campaigns

by Andrew Miller on 08/23/10

How to poison your competitor's PPC campaigns

For the record, I DO NOT advocate this method of poisoning competitors’ PPC campaigns but it exists and you should be aware of it. I’ve seen it on a few client accounts and have taken steps to mitigate the damage. It’s an ongoing, evolving problem so you should check regularly to see if you are being targeted.

Here’s the bottom line. Your competitors have a vested interest in seeing your campaigns fail by becoming wildly inefficient and unprofitable. They can get more customers for less money while you pay more for fewer.

How it works

A savvy competitor knows you are bidding on some of the same keywords. In this example, I’ll use “dining room furniture” (phrase match) as the keyword in question for two companies that sell dining room furniture.

You, like a good PPC marketer, set up your campaigns with keywords (i.e. “dining room furniture”) and negative keywords (i.e. “antique”, “plastic” or “zebra print”) to filter out keywords for products that you don’t carry. Normally this would be enough to keep your ads showing for only the most relevant search queries and minimize wasted clicks but it may not go far enough.

Your competitor could use automated search tools (rank checkers, bot nets, or simple scripts) that search for long queries that trigger your AdWords ads hundreds or thousands of times and rarely, if ever, click on them.

Why it’s a problem for you

Eventually, your keywords and ads suffer from unusually low Click-Through-Rate (CTR). The lower CTR results in lower Quality Scores and higher Costs Per Click (CPC).

In plain English: your ads are clicked less frequently than Google expects. Since the big G are not going to make much money that way, they will lower your ads’ Quality Score, which makes your ad rank lower and costs you more per click.

Meanwhile, your competitor has already taken steps to make sure their keywords are not affected. See “How to fix it” below.

How to detect it

Use the Google AdWords Search Query Performance report to identify the search queries that resulted in your PPC ads being triggered and clicked.

In my client’s case, I saw many variations like these:

  • 3 piece dining sets kitchen dining room furniture furniture (136 impressions, 1 click)
  • wood dining room chairs kitchen dining room furniture furniture (92/1)
  • for the home furniture kitchen & dining furniture kitchen & dining sets (111/1)
  • for the home furniture kitchen & dining furniture kitchen & dining tables (192/1)

These phrases look odd, and there were dozens or hundreds of variations each receiving a few hundred to several thousand searches per month. They even seem to fall into templates and patterns.

Typically, they have some very similar characteristics:

  • Repeated keywords such as “furniture furniture”
  • Extremely long-tail phrases that don’t make sense
  • Queries that would not normally receive hundreds or thousands of impressions and very few clicks

Obviously, not all high-volume keyword are going to fall into this category. It’s really only the weird ones that you need to watch out for.

How to fix it

Once you’ve run the Search Query Performance report for your campaigns, comb the list carefully looking for query variations that meet the criteria above. Once you figure out the patterns for these queries, they become easier to find.

Extract snippets of those queries and add them to your negative keyword list for each campaign. As an example I added the following:

  • [furniture furniture] (exact match)
  • [dining furniture kitchen & dining] (exact match)

This will prevent my client’s ads from appearing when the automated search tools repeat these queries over and over. Our Quality Scores rise and our CPC falls. Every time.

My hunch is that your competitors know the keyword phrases and have already added them to their negative keyword list so their ads don’t show up when their bots are let loose. Sneaky, eh?

Taking 30 minutes to tweak your accounts can save you thousands of dollars each month and result in more conversions from your PPC campaigns. Do this once a month to stay on top of changes to their methodologies.

Where to go from here

Just some fun ideas if you are still being affected by these attacks:

  • Some simple log file analysis can be used to capture the IP addresses of your website visitors that click on these ads. If you can trace those IP addresses to a competitor, great. More than likely though, they will be anonymized or run through a bot net or proxy server somewhere. After all, if they are devious enough to pull off these schenanigans, they are probably clever enough to hide their tracks.
  • Use the AdWords IP Exclusion tool to prevent certain IP addresses from seeing and clicking on your ads.
  • Report the activity through AdWords support or contact your AdWords rep for help.

Anybody else out there seeing this? What have you done to fix it? Any other tips?

(CC photo credit: gato gato gato on Flickr)


You May Also Like:

  1. Bored With PPC? Try These!
  2. Landing Page Quality Matters More Than Ever
  3. Steph Stalks Google’s Ann Arbor Staff

11 Comments on » How to poison your competitor’s PPC campaigns

{ 11 comments… read them below or add one }

Kenny August 23, 2010 at 11:53 am

Nice post dude – Google has told us in the past that some of this activity is from a certain search partner and due to the way that the queries look in the search query report when they are from that particular partner. They, of course, will not disclose who that partner is. I’m not saying at all that this activity isn’t largely malicious or competitor-related, but search partners could be having an impact too. One way to know for sure would be to separate out your campaigns to Google.com only versus Search Partners only, and see what the data still shows in search query reporting.

Kenny

Andrew Miller August 23, 2010 at 5:09 pm

Good point, Kenny. We are using the Google.com and search network targeting so the inflated impressions may not be coming from Google.com. Regardless, the negative impact costs a lot of money and head-scratching for a lot of people. If it continues I’ll turn off the search network to see what happens.

Steve Vail August 24, 2010 at 12:38 pm

Good post Kenny,

The problem is in Search Query reports, you only see those keyword phrases that were clicked on. Now if I’m clever enough to employ these tactics, you’d also think I’d be smart enough to not click on suck keyword phrases to avoid detection. No impression reports, so how do we tell if we’re being cheated on the keywords and phrases that aren’t getting clicks?

Andrew Miller August 24, 2010 at 12:58 pm

Good point, but I think that scenario is just as likely to be detected by a PPC marketer when reviewing keyword-level data. Lots of impressions and no clicks (or a very low CTR if there happen to be a few legitimate clicks thrown in) should trigger an analysis of that keyword or phrase anyway.

Kenny August 24, 2010 at 1:06 pm

Steve,

That’s something we’ve complained to Google a lot about – show me the impressions for which I DIDN’T get a click. I may care more about those than the ones that did receive a click on. They’ve continued to expand how much is shown in the search query report (you used to get the big volume clicks stuff and then “and 8,000,000 other queries” garbage). I suppose Google isn’t exactly driven to make use 100% efficient though. That would just make our CTRs better and drive our CPCs lower. :)

William Vicary August 25, 2010 at 7:21 am

Kenny, On top of allowing you to make your CTRs better and drive your CPCs lower it would also open a huge amount of keyword data that should be invaluable for your SEO efforts.

Andrew when you say “should trigger an analysis of that keyword or phrase anyway.” is there any tips you could give as to how to detect these additional keywords if one believes they have been subject to impression fraud?

Andrew Miller August 25, 2010 at 8:16 am

Willliam, by “analysis of that keyword” I simply mean digging into the reasons why that particular KW has such low CTR. It could include Quality Score, not enough negative keywords, irrelevant ad text, or any other number of things.

Also, look into the conversion rates of those keywords to see how those visitors interact with your site. That may provide some insights into whether or not you are targeting the right phrases.

Kenny September 2, 2010 at 1:51 pm

Andrew, as a follow-up to this, I am 99% sure a lot of this activity is from sponsored links on eBay. They became a Google Search Partner a while back, and based on our search query data and Google’s previous guidance, I’m pretty confident. Separating out the search partners should further prove this. Haven’t done that yet, but I feel pretty confident based on the impression levels and way the queries look coming in that that’s what’s going on for us, and probably for some of your account as well. Shame you can’t separate the day by partner.

Kenny September 2, 2010 at 1:52 pm

the data by partner, that is.

PPC Wizard November 12, 2010 at 4:00 pm

That’s dirty, but would it really be worth all the work to a competitor?

Tony Nguyen January 18, 2011 at 12:14 pm

Really glad to have found your blog. Great stuff. Didn’t even know that was possible, someone inflating the search impression automatically.

Leave a Comment

Previous post:

Next post: