Archive for November, 2006

I have some additional information regarding the CPA Rule in AtlasOnePoint. First, you will NOT want to add only 1 CPA rule to all the keywords in the Keyword Analysis Report that have high visitors but no conversions. The CPA Rule will not work on these keywords because there are no acquisitions.

The CPA rule is typically used for fine-tuning the CPA on each individual keyword, and will not work effectively across your entire Keyword list. It also won’t work effectively if the target CPA is $20, and the current CPA is $300. That’s too large of a swing for the CPA rule to work. In that scenerio, it’s best to set the target CPA only 10% lower than the actual CPA. Then, wait until a few hundred clicks on the keyword…look at the new actual CPA. And, setup a new target CPA that is 10% lower again. In my example, keep going thru this routine until the real target CPA of $20 is achieved.

Remember, you will want to create a CPA rule for each of the keywords that you need to apply this rule to. This rule works incrementally, meaning that the target CPA for keywords should be no more than 25% lower than the current CPA. Typically speaking. But, that’s not the case for large differences. In the large differences, lower the CPA incrementally by 10% as a rule of thumb.

So, to recap, create a CPA rule for each high performing keyword. Create a realistic CPA goal that can be achieved in realistic decreasing increments. Do not use this rule with keywords that have high visitor rate but no conversions.

I just heard about this company last week, and I’ve been meaning to check them out. SpotRunner
is a brand new company that offers an interface to choose and customize an already produced TV ad. Think of it like a Google Adwords for Television Advertising. You can target your ad by geographic location, right down to the zipcode. Choose your ad from an existing library of TV commercials. While browsing their ad library, I saw ads as low as $499.

Here’s an excerpt from their About Us page:

Spot Runner is the world’s first television advertising agency that’s Internet-based and designed to be used by local businesses. Now local businesses can conveniently and cost-effectively harness the power of high-quality, targeted, television advertising to reach their customers and sell their products and services. With Spot Runner’s revolutionary approach to commercial production, media planning and media buying services, local businesses now have access to a powerful marketing tool that was previously out of their reach.

And, we already learned how offline media affects Search. So, this is a great way for the little guy to compete with the big players. Really, leveling the playing field.

Finally, it seems like an affordable solution for even the small local business owner, to advertise on television with a high quality professional TV ad.

This could be old news….Is BlogHijinks the newest human-edited blog directory, or just another Link Farm?

I’m gonna rule out the link farm. In my interview with Greg from BOTW he describes how to identify the difference between a web directory & a link farm:

A link farm is no more than a collection of links, typically uncategorized, and displayed in a haphazard way. Link farms are designed for the sole purpose of manipulating search results, with no benefit for a user, and little, if any benefit to the site owner.

BlogHijinks is definitely not a link farm. The site looks great. Personally, I like the design. It’s easy to navigate. And, it’s in an industry that’s in the early stages. There aren’t a whole lot of other web directories that are dedicated to blogs. Although, it has some stiff competition from BOTW blog directory.

Earlier this year, Umdum, another human-edited web directory popped on the scene. Although, it’s not solely dedicated to blogs.

BlogHijinks seems like a great place to get a free text link. And, it also has a Social aspect to it, in that you can review each blog. That’s pretty cool, and unique. It also has some popular blogs like Lifehacker already listed.

So, I’m gonna extend a hearty welcome to BlogHijinks, and submit my blog for a human editor to review for submission so I get another inbound link.

I took a brief break before posting this, to submit my blog, and did so successfully. However, there was a page layout problem. It might’ve just been my browser, not sure though.

Just by accident, when doing some research for a client with my colleague Angela ( i’d link to her, but she doesn’t have a personal blog…yet) I noticed that there isn’t ANY search results for most of the Unicode symbols. Here is a screenshot of the Google SERP when I search for an asterisk on Google. Pretty interesting, huh? So, what if an advertiser decides to brand themselves using an exclamation point like E! Online? How about a brand like Yahoo which has an exclamation point to bid on : ! Does it make sense to bid on the unicode?

In the screenshot, you can see the Google one-box. However, I didn’t even get the one-box results for most of the others. And, most surprising of all…nobody is bidding on the dollar sign ($). I would think the large banks, and traders would at least bid on the $.

Can it be that the Google algorithm ignores punctuation? I bounced the idea off of my bud Chris. Now…let’s try a search for: html ( but put the <> around it). I got 3,730,000,000 results. However, you’ll notice the results don’t take into consideration the <>. It strips the <> from the Search results. To see this, just look at the Title of each of the Search results. Similarly, “dog-cat” is equivalent to a search for “dog cat”. So, it appears as if the Google algorithm ignores punctuation.

It seems like trying to optimize for any of the unicode symbols is impossible. However, the question still remains if it makes sense for an advertiser to bid on any specific symbol.

Where I can see it making a lot of sense is, if a brand decided to introduce a new product or service and represented it with a specific unicode symbol, then it would make a lot of sense to bid on the symbol.

What is Click Fraud?
Definition: Clicks for any other reason than to visit the advertiser’s website.

How does it happen? What tactics are being used?
1. Manual clicking. Usually from a competitor trying to exhaust the budget.
2. Automated clickbots.
3. Email distribution.
4. Spyware: Placing ads into Spyware programs that automatically click on ads, and result in a charge to the advertiser.
5. Search engine syndication affiliates attempting to generate more commissions for themselves.

Click Fraud Management Methods:
Search Engine Auditing: Search engines have their own proprietary systems to track & measure data. If fraudulent clicks are identified, a refund is made in the form of advertising credits, not an actual refund of dollars. 2nd tier SE’s tend to have a higher amount of invalid click activity. This can be attributed to either not robust enough click monitoring activity or to an increased dependency by their network partners and not policing themselves more aggressively. The SE’s have stated that they can’t fully audit all clicks for potentially fraudulent click activity.

Web Analytics: It is possible to identify Using web analytics to monitor key metrics such as: recurring IP addresses, Referrers, # of Pages Viewed per visit, Date & Time, volume & timing of incoming traffic, conversion rates ( eg: bots filling in forms spike CR’s), time on site, and click cadence.

Risk Factors:
The percentage of fraud depends on Advertisers and their individual campaigns. Risk Factors include: Geographic location, length of keyword phrase, CPC, high competition within a specific category.

The accepted industry average for click fraud is 13.7% which has been extrapolated across multiple industry categories (Source: Click Forensics). However, it is important to note that 13% of clicks does not equate to 13% of spend. Most advertisers use a higher % of spend on a small % of the keywords.

Strategies to minimize click fraud:
1. Monitor highly competitive terms & categories which could be at a higher risk.
2. Monitor smaller 2nd tier engines.
3. Review regularly scheduled web analytic reports.
4. Increase SEO efforts to reduce dependency on PPC.

Advertiser vs. Publisher Search Engine Agreements: Contract terms between Advertisers and the Search Engine network partners are worded differently. The SE contracts state that advertisers are responsible for “All” clicks. Interestingly, the SE contract with the publisher’s state that they will only be paid for “valid” clicks.

Agency perspective:
Since the Search Engines can’t see data across multiple engines and syndication partners, the agencies should not rely on the audited data provided solely by the SE’s to manage click fraud. The agency needs to monitor & gather data for clients, and work with SE’s to resolve any disputed click activity. Agencies need to take a holistic view at multiple click fraud solutions.

Dispute Resolution:
The Agency should be actively engaged in managing click fraud, and will handle any dispute resolution involving the search engines. The agency needs to resolve the dispute quickly, and ensure that it does not interfere with the client’s campaign objectives. When backend technology is made available, the agency should schedule regular web analytic reports, and review it with the Search Engines if and when necessary.

Research conducted by Atlas Solutions studied the impact of search engine rank to conversion rates. The study consisted of 41,460,000 clicks, 408,000 keywords, and thousands of ads across multiple industry segments that produced an industry agnostic study relevant across all industry segments.

The research focused on answering this question: How does the change of rank of a Paid Search ad, affect the conversion rate for that ad?

General finding regarding High volume vs. Low volume Search Ads / Keywords:
- Conversion rates for keywords generally fall when rank is lowered. This is mostly caused by the decrease in click potential for a lower ranked search ad. However, the study concluded that conversion rates for low volume keywords may actually increase at lower ranks.

Key takeaways from the study:
1) Advertisers must do a balancing act between efficiency and volume. Just by getting the best possible ROAS at rank #10, an advertiser could be missing out on a much higher volume of conversions if they increased the ad to a slightly higher rank. By decreasing rank, an advertiser saves on the CPC. However, that most likely will cause a reduction in site traffic. Thus, lowering CPC’s could do more damage than good by potentially lowering the conversion rates.

2) It has been found that conversion rates typically drop by rank. So, if the cost per click is a very small difference between rank, it probably makes sense to raise the bid to gain a higher rank, which corresponds to a higher conversion rate.

3) Conversion rates for low volume keywords have been shown to increase as rank is lowered. This is an opportunity to squeak out more efficiency and lower the total cost of the SEM campaign

The Atlas Solutions study is a useful benchmarking guide that provides a deeper understanding on how rank affects conversions. By continually testing & measuring, advertisers will discover the most cost efficient way to drive traffic & sales from their search campaigns.

For years I’ve never really known how to answer the question, ” What do you do for a living?” My typical answer is, marketing, internet marketing, Search engine marketing, or SEO Consultant, but the typical person has no idea what SEM is. When I say Search Engine Marketing, the typical response is: “Oh, what search engine do you work with?” Then, I have to explain what Search Engine Marketing is in a sentence or two. I’m getting used to the response.

But I finally have it figured out. It took me awhile, but I think I finally got it. When someone asks me what I do for a living, I will now say “Advertising” or, “Internet Advertising”. It is so much simpler, and I think people understand what advertising is. So, my thought is that they’ll think Internet advertising must be advertising on the internet. Sounds logical, right? Hopefully, that will make the conversation go much more smoothly. Although, I haven’t had the opportunity to test my new theory out yet.