Internet Marketing Blog

The Marketing Experiments Blog recently posted the results of a click fraud study. The study itself is interesting but somewhat inconclusive due to its short duration and limited scope. However, they make some excellent suggestions for preventing or at least reducing click fraud, some of which include:

1. Watch for significant variations in ad campaign performance.
2. Be conservative with content ads.
3. Don’t display ads where you don’t sell products and services (e.g. if you only sell in the U.S., don’t display ads in Europe).
4. Use analytics tools or click-fraud-detection tools. (I’m skeptical about the economics of the latter for small campaigns.)
5. Click fraud from search engines is quite low, not so for contextual ads (3rd party websites).
6. Measure your results and ROI so that you can bid to your budget, not because you would like to see your ad in the #1 position for branding purposes.
7. Find alternatives to expensive, high traffic keywords as these tend to attract more fraud.

You think you’re paying a lot?

A recent post on CyberWyre identifies some of the most expensive terms upon which to bid in AdWords. Not surprisingly, “mesothelioma lawyers” tops the list at over $50 per click. AdWords’ use of CTR in calculating bid position means that most advertisers aren’t paying this much. Top position on Overture (Yahoo Sponsored Search) for “mesothelioma” cost $24.75 at the time of this writing.

Most of us aren’t in the lung cancer lawsuit business, so these bids may seem out of sight. However, there are plenty of other terms that require substantial investment. The top bid for “business intelligence” on Overture is over $12, almost double that for AdWords. Terms including “real estate” can be bid up over $20 on AdWords and $10 on Overture.

Want to reduce costs? Remember:

• Use negative keywords
• Use keywords matching options
• Bid based upon ROI
• Separate search and content network ads campaigns
• Target the right geography (or other segmentation options when available)

According to a study by WebSideStory, Google dominated search in the United Kingdom over the month of February. Their data comes from their HBX Analytics customer base. Even if their customers could theoretically be skewed towards Google (why? I don’t know, but they could be), the size of the study and the huge differences are probably indicative of the broader market:

The one other complicating factor is that the study includes paid advertising which could skew the results in a bit more in Google’s favor.

The Internet Advertising Bureau (IAB) recently proposed Lead Generation Best Practices for advertisers and publisher/vendors. Advertiser best practices focus on transparency toward the consumer, and publisher/vendor guidelines promote data integrity.

These best practices are intended for online direct marketing efforts like opt-in email campaigns and newsletter subscriptions, but their principles are equally valid for offline direct marketing. Application of the advertiser best practices may instill a level of trust that makes your target audience more receptive to all of your direct marketing messages.

A new Google-sponsored comScore study found that searchers still make most search-related purchases offline. The study also shows that the minority of online purchases occur in the initial Internet session; the majority occur in subsequent sessions.

Consumer Vs. Business
The new study looked at conversion rates for various consumer categories over the 2005 shopping season. I believe that it is safe to infer similar if not more pronounced such behavior for business purchases because the latter typically have longer, more involved sales cycles, often involving more people.

Offline Conversion and Search ROI
Ignoring offline purchases can grossly underestimate the effectiveness of online marketing efforts. The ROI of SEO and keyword advertising can easily be double what online purchases suggest where there were as many offline purchases as online. If an online marketing program is close to breaking even solely from online purchases, it would take only a small offline conversion rate to make it profitable overall.

Purchase Latency
The other key issue raised by the research is the critical delay between awareness and purchase. Except for simple and/or impulse purchases, it’s very unlikely that your customers buy from your website when they first find it. If your online conversion measurements don’t take this into account, you will underestimate search marketing ROI. Measuring delayed purchases is further complicated when the searcher is not the buyer, a common occurrence in business-to-business transactions.

High Click-Through-Rate (CTR) is a good indication that people find your ad compelling. AdWords even rewards higher CTR by letting you bid less for the same position than someone who has a less effective ad. So the name of the game is maximizing CTR, right?

Not always. Aside from having a very effective and compelling ad, there are two other reasons you may have a high CTR: Poor selectivity and click fraud.

Poor Selectivity
People are clicking on your ad that are poor prospects for an actual sale. Your ads aren’t prequalifying, instead they attract clicks from people looking for a product or service different from those that you offer. Ads that are in the top position(s) are also much more likely to receive clicks from people who either don’t know the difference between paid and natural search results or who don’t care.

Click Fraud
Suspiciously high CTR may be the result of competitors trying to drive up your advertising costs by clicking on your ads. While for lower positions this has a partially compensating benefit of improving your CTR and hence your bid, at the highest ad positions it simply costs you more money. Plus, at the top positions you are usually paying top dollar (hgih CTR notwithstanding) and are more visible to the competition that may otherwise not click on you if you were further down the list.

The best way to identify these two problems is through examination of website statistics. If many visitors that come from your high CTR campaigns are only staying for one or two pages and/or a very short period of time, it is very likely that you are experiencing one of the two problems above. This assumes that your average visitor stays longer, otherwise the main concern is that your website may not be effectively converting good prospects into buyers.

I admit it: I’m Google fan. One really big reason is Google Maps, which rewrote the book on mapping (although ironically it’s now falling a little behind competitors on the feature front). Similarly Google Finance demonstrates how to combine DHTML and fabulous interface design to create a very powerful yet easy-to-use visualization and analysis tool.

Let’s look at IBM. Move your mouse over the graph without pressing any buttons, and it will show you the date and time under your pointer as well as the volume and share price. Slick. Now drag the chart with your mouse. Cool. Want to view a longer period of time? You can click on the standard 1d, 5d, 1m, etc. links, or you can use the graphic timeline above the main ticker. Drag the bars at either end of the range to adjust how much time is shown in the main chart. Want to move your arbitrary time slice back or foward in time? Use the scroll bar between the timeline and stock chart. It updates dynamically. Very cool. As you scroll, you might notice the list of news on the right scrolls up or down to keep the items matching the visible portion of the stock chart plainly in view. See the letters next to each event? They’re in the chart as well at the corresponding date. Click the letter in the chart to highlight the news item in the list, or vice-versa. One more little detail: scroll down the page until you see the Management section. Now hover your mouse over an officer’s name and you will see more details, links, and a picture (if available).

I like some razzle-dazzle and novel techniques, but extra features need to be more than fluff. In other words, fancy “gadgets” are only justified if they make a site significantly more useful than it would be without them. Google has consistently excercised good design judgement on the tools it makes public. I suspect that the gatekeepers that control what gets out of the labs and into beta employ significant restraint over numerous less successful features that we don’t get to see. Tools like this take a lot inspiration, and perspiration; they must go through many design and usability test iterations before they work so well.

Don’t fall in love with features that add few new capabilities or that compromise ease of use. Most of us don’t have the resources of Google to exhaustively test new ideas. This doesn’t mean we shouldn’t try, but recognize the effort it takes to get it right, and test ideas before committing to them.

Usability guru Jakob Nielsen recently summarized common website mistakes that result in lost business, both online and offline. These include:

• Aggressive home page promotions that annoy users and diminish trust.
• Lack of consistent design and navigation throughout the site.
• Poor information presentation/navigation and lack of good search.
• Over-emphasis on being new and different at the expense of speaking the language that customers understand.
• Assuming that our opinions reflect user opinions/perceptions.

I describe the web as being the “ficklest” medium: users can leave your website with ridiculous ease. There is nothing keeping them there but your ability to convince them that you have what they want, and that it is more worthwhile to look at your offering than to look for someone else’s. The errors described above are among the most common causes for visitors not becoming buyers.

These are easy traps to fall into, but they aren’t terribly hard to avoid if appropriate actions are taken early on in the website design process. Awareness of potential problems is often all that’s needed to avoid them, so keep this list handy for your next web project.

I just gave 3D Seek’s sketching search engine a try. It’s pretty impressive: sketch a mechanical part and it will provide images that it thinks may match your sketch. I sketched a simple triangular bracket, and while there wasn’t a perfect match, the search results were compelling enough that if I were looking for that part, the listed vendors would probably get a call from me to find out if they had more stock not represented here or could make a custom part. Note that the tool is presently in Beta and does not appear to support Firefox.

There’s been a lot of talk in recent years about searching through multimedia content: pictures, sound files, video, etc. Current popular resources like Google Images don’t use automated mechanisms to classify the actual content. Instead, they either look at the text around the media files or a real live human being provides the “indexing”. 3D Seek is, at least for now, targeting mechanical parts which greatly simplifies the image recognition problem in numerous interesting ways. These simplifications make it little more than a novelty for anyone not looking for a relatively simple mechanical device, but its success in doing so is quite encouraging.

Should you start picking your website images for ease of sketching to attract users of these kinds of search services? No, we’re still a very long way from that.

by Christine Slocumb, Principal, The onQuest Group

As a public relations consultant, I’m amazed that large public relations firms do not craft releases that rank well in search engines. They still view releases as items that only garner media coverage. Distributing releases over PR Web allows the use of anchor-text linking resulting in quality reciprocal links for chosen keywords. I am also finding that freelancer writers placing pieces in quality publications are picking up releases more often off of PRWeb than the traditional PR Newswire. I still advise clients to post their releases on both traditional and online wires, but for $100-$300 you can’t beat the investment in releasing on an online wire service with traffic-generating features such as anchor-text links and guaranteed placement in Google News.

I believe the large PR firms have not adopted these new strategies for press releases, because optimizing press release content for search engines demands a different writing style that’s a cross between an ad and news release. For example, a main keyword or phrase should be in the title of the release and the first paragraph. But it’s just a matter of time until every PR firm adopts this style, so jump on the opportunity to optimize your releases while it still gives your firm a competitive advantage.