The Internet Advertising Pro

Monday, July 03, 2006

Keyword Click Fraud - How Concerned Should Search Marketers Be?

When I ask some large search marketers if they are concerned with click fraud, surprisingly many will answer “We know it exists but we just consider that as part of the cost of doing business…”.

Let’s consider this statement. For many online merchants, search engine advertising represents 50% or more of their entire monthly expenses.  For example, a small business generating $60,000 a month in online sales might spend as much as $10,000 to $20,000 on keyword advertising without thinking twice. After all, it’s what’s driving the sales. But if I were to ask that same marketer “Would you be concerned if someone were entering your office every single night and taking $150 out of the cash drawer?”, or “defrauding your credit card account for thousands of dollars each month?”…. Now, would that be considered a cost of doing business?

There are a few main reasons that advertisers are willing to accept this problem:

  • They are not aware of the scope. An average marketer can throw away from 15% to as much as 70% of their ad budget, amounting to hundreds or thousands of dollars each month.

  • What they really mean is that “I don’t have time to deal with this now and`keyword advertising is still cost effective despite the problem”. They might be aware that uncovering and proving invalid clicks can be a time consuming process, or they have no idea how to do this.

  • They are under the assumption that nothing can be done about it, so if you need to be on the search engines, you have to live with it.

This logic was fine, two or three years ago when invalid clicks represented a much smaller percentage of an advertiser’s monthly bill. In recent months, however, the same marketers are scratching their heads and thinking to themselves “Gee, the results from my search campaigns keep declining, while at the same time, keyword prices are higher than ever…”.

With the problem of invalid clicks ballooning over a relatively short period of time, many marketers are assuming that most of the decline in performance is the result of other factors, like increased competition, saturation, frequency, and normal business cycles. While all of these do play a role in performance changes, most advertisers may underestimate the scope of the largest single factor affecting their PPC performance.

I have been performing click traffic audits for clients and my own businesses since 1998. What many people don’t realize is that this is not a new problem at all; it’s a rebirth of a problem that actually peaked back in 2000 and then went away for a while, while internet advertising on mid-to-low tier sites moved from the CPC (Cost-per-click) to CPM (Cost per 1000 ad views) and CPA (Cost per Action). As founder and CIO of a CPC based ad network, launched in 1998, I witnessed the most sophisticated click fraud schemes I have ever seen in 1999-2000, dwarfing the complexity of today’s new robot and botnet systems, and effective enough to elude the current measures taken by the search engines to detect suspicious activity.

In 2001, Commission Junction and other affiliate networks quietly phased out all CPC pricing in favor of CPA, citing other reasons, such as decreased advertiser demand for CPC, but not disclosing that one of the main reasons was the epidemic of click fraud that had become impossible for them to control. At the same time, ad networks (serving primarily graphical image ads) moved to impressions based campaigns. So the decline of click fraud was in direct relation to the decline of PPC advertising. Also at this same time, even Google was only offering CPM based ad positions, at the top of their search results.

Then, in 2002, with the overhaul of Google’s AdWords program to what we know today, PPC was reborn. While advertisers were reluctant to pay for impressions of text links ads (which offer little to know branding value), direct marketers were eager to pay only for clicks, and only when a prospective customer was searching for their keywords. This was the most effective (in terms of direct response) form of online advertising ever; hence, its mass popularity. Over the past year it has become so popular in fact; that the same affiliate networks who vacated the CPC model years ago have recently announced a return to CPC, and phasing out of their CPA programs.

Unfortunately, this new frenzy of spending on PPC eventually woke up the old, dormant click fraudsters as well as the new opportunists that appear wherever there is money to be taken. Additionally, new types of click fraud emerged as otherwise legitimate businesses discovered they could quickly deplete a competitor’s ad budget by repeated clicking on the PPC links. There are several other varieties of click fraud, as well as other types of inadvertent invalid clicks that also waste advertiser’s money, but these are beyond the scope of this article, which is focused on deliberate affiliate fraud, as a means of directly profiting from PPC advertising with no intention of doing business with the advertiser.

In response to the growing scale and increased awareness of the problem, a cottage industry of click auditing services and click fraud prevention software was born. While the search engines and ad networks increased their in-house measures to prevent and discount invalid clicks, growing distrust and an apparent conflict of interest has more and more advertisers seeking an unbiased, third party opinion, and one that is backed up by evidence not provided in the ad networks’ reports.

Tips to reduce your exposure to click fraud:

  • Our recent studies have indicated a direct relationship between the average price-per-click and the ratio of invalid clicks vs. high quality clicks. The higher the price paid per click, the higher the percentage of invalid clicks. One of several reasons for this is that the higher number of clicks required to profit at a low CPC is much easier to detect than the low number of clicks needed when the rate is high. Another possible problem associated with high CPC is that competitors can deplete your ad budget which much less effort. Unfortunately, many advertisers have no choice but to pay high CPC if they want traffic to their campaign. We do not recommend reducing bids or ended high priced keywords, however, it is essential to realize the increased risk and allocate your scrutiny of paid traffic relative to the price you are paying per click.

  • Risk increases in direct relation to the number of affiliate sites you are running on through a particular ad program. The ratio of invalid clicks is significantly greater on search “networks” as compared to the search engine itself, and on content “networks” as opposed to individual content sites. For this reason, if a campaigns run on both the search engine itself and search traffic from affiliated sites, attention needs to be paid to the percentage of traffic derived the affiliates as compared to direct search on the search engine.

  • Risk increases proportionate to the quality of the network’s affiliate sites. If you are studying your referring pages and are finding a large number of pages with little or no content other than the paid listings, this traffic should be more closely analyzed. The same applies to unprofessionally designed pages, pages that redirect to another page or refresh automatically, pages that open invisible or blank windows, pages with controversial content, pages with keyword spam, and pages that seem to be of little relevance to your campaign or keywords. If you do not know what pages your traffic is coming from, it’s time to start using a good tracking software or service, which will give tons of information on your traffic that the search engines and networks do not provide in their reports.

  • Unfortunately, risk increases significantly with traffic from certain geographic locations. If you are running your campaigns without geographic targeting, or targeting one of these regions, additional precautions should be taken to monitor this traffic. We have had some clients completely block their (otherwise global) campaigns from entire countries for the sole reason that the majority of traffic they were receiving from these countries was found to be invalid. (Note: As many international affiliate fraudsters become more sophisticated, they are developing methods to hide behind US proxy servers and use US based slave computers.)

  • Monitor, Track, Record, and Compare your traffic with the clicks you are being billed for. High quality click tracking and auditing services can cost as little as one penny per click and save you hundreds or thousands of wasted ad dollars every month.

About the Author
Jeffrey Howes is founder and CEO of Internet Advertising Technologies, Inc., and co-developer of, and its click tracking software product ClickHawk™. Prior to work on the TrafficSentry™ product, Jeff, a Computer Science major and entrepreneur, with 15 years of business marketing experience, has been in ad network industry for 8 years, beginning with the launch of a leading ad network in 1998.

Additional Click Fraud Resources

PPC Fraud, Google Stands to Lose Millions
How Does Click Fraud Affect You?
Click Fraud - An Overview
Click Fraud - Is it happening to you?
Google Agrees To $90 Million Settlement In Class Action Lawsuit Over Click Fraud
Yahoo Settles 'Click Fraud' Lawsuit


Thursday, February 10, 2005

Are Super Bowl Ads Really Worth The Money?

“Are Super Bowl Ads worth the money?”

Well, that is the question of the day. The overall impact of a Super Bowl spot, or any television ad for that matter, is impossible to quantify, and there is, of course, no single answer that applies to all advertisers. What we really want to no is whether we are getting the best return on the investment, as compared to allocating the spend to alternative media options.

Granted, the Super Bowl is a unique media opportunity, unlike any other in it’s ability to reach tens of millions of Americans, all at once, and all paying full attention to the advertisers message. Nothing else can touch that. Then there is all of the (free, so to speak) additional press pre-game press and post-game “buzz” as a result of the commercial. However, the results being to vary, and become less clear, when attempting to measure how this massive spike in brand exposure translates into long term brand memorization, product awareness, and ultimately, sales.

Several studies, like this one by Clickin Research Inc., have concluded that lesser known brands may benefit more than well-known names, in terms of increases in brand familiarity after airing Super Bowl commercials. In fact, the studies demonstrate there is little or no measurable increase in brand familiarity or likeability among well-established brands.

Most of last year’s Super Bowl advertisers, including America Online and, did not buy time this year, which may be an indication of the lack of tangible return from last year's ads. Mitsubishi’s ad last year, for the Galant sedan, directed large volumes of traffic to the company’s web site in the time immediately following the airing of the ad. However, sales of the Galant plummeted 33% in 2004, while the company’s overall US sales fell 37%.

We are all aware of the list of dot coms who burned their cash stockpiles on Super Bowl spots in the late 90’s. A Super Bowl presence brings glamour and prestige, but a Super Bowl ad is not a replacement for good market research and a solid marketing plan. Advertisers must be able to deliver a product that consumers want.

Super Bowl advertising is not for a company who will be spending all or most of their entire ad budget on the spot. While the Super Bowl ad may be a good way to immediately boost awareness of a lesser-known brand, the message must be repeated in ordered to sustain any increase. The studies are based on pre-game and post-game surveys, all taken within a 24-hour period, and do not attempt to measure longer-term impact. Let’s face it, the American public is quick to forget, and that’s not taking into account the average blood alcohol content of over 0.10% on this day.

Looking at CareerBuilder’s “Monkey” campaign as an example, an estimated $5 million was spent on 1 minute of commercial time, plus an assortment of sponsorship positions. The ads are not ranking well in post-game surveys, and alone, they may not have produced any substantial gain in brand familiarity or likeability. However, these ads were only a small part of a $200 million cross channel campaign that will include other TV spots, radio, online (their user are all on the Internet after all), and print media. The continual repeating of the message will keep the brand in the public eye for months longer than the initial ads alone would have been retained.

This then brings up the question “Did they really need the Super Bowl component of this campaign?”. Well, to that there is no definitive answer. There is no question they got the one-of-a-kind exposure, and brand association, that only a Super Bowl ad can create. Assuming that two thirds of the audience watched the commercials, they reached 57 million consumers, all at once, and had their complete attention for 1 minute. On the other hand, they did pay $84 CPM for the privilege, and there is some question as to the audience retention, past the short-term. Now, if we take that same $4.8 million spend and allocate it as 50% regular prime time television, and 50% online, we increase exposure to over 1.3 Billion total impressions, over a several month period. The shear reach and frequency of a campaign this size would guarantee virtually complete brand recognition and product awareness of the advertiser. Not to mention, we have only spent 2.4% of the total budget.