I have blogged on Google’s Quality Score before.  It was the usual insight about why its important to get a good quality score.  I went into how getting a bad quality score may cause you to pay more to be in position number 4 than your competition paid to be in position number 2.  I even talked about some things to do that may help to improve a quality score.  I’m sure the insiders at Google had a good laugh because no matter how much I hypothesized, I had no varifiable evidence on what will increase a quality score.

Earlier today Google may have had a gitch that made public some of the variables used to make up the quality score.  This is the buzz anyway.  Eric Landers of Search Engine Journal published in his blog that immediately below each of the sponsored search results (AdWords) were three separate variable names and values.  Eric has screen shots and an opinion on what these may mean.  Since this gitch has since been corrected by Google I wasn’t fortunate enough to have seen this myself to allow me to share my insight.  I will at least point you to Eric’s Blog on this so you can get it directly from him.

Technorati Tags: Paid Search, PPC, Analytics, Campaign Tracking,  google content network, google site targeting

siteEDGE News, Pay Per Click, Local Search | No Comments »

Google Content Network Defined

The Google content network comprises millions of websites, news pages, and blogs that partner with Google to display targeted AdWords ads.  Ads to the content network are targeted based on content themes rather than specific keywords as is done on the search network. Unlike a keyword search which is well defined, a content match may not display your ad on the site you had in mind resulting in clicks you prefer not to have.

Identifying the Sites that Don’t Convert

On June 1st last year, Google made available a report that tells you which sites are not converting.  You can run this report at the ad group or the campaign level for a domain or a url.  Once you run the report you can feed the results into the Site Exclusion option in the Tools section.  This will cause the domains or url-s you designate to not show your ad resulting in a better use of your ad dollars.

Further Analysis

This on the surface appeared to be an excellent opportunity for one of our clients to save a few dollars.  After running the report for all the ad groups we learned there was about $1,075 that could have been saved in June if the sites with no conversions had been excluded.  Upon further analysis we determined that some sites did not convert for some ad groups but did convert for others.  Since the Site Exclusion option does not allow exclusions at the ad group level we had to revise our plan to exclude the non performing sites at the campaign level.

Still Not Bad

The initial excitement over the potention savings abated somewhat.  Instead of the $1,075 a month savings that we had anticipated from excluding non performing content network sites, we will save about $521 a month. We have to factor in that this is just one month’s worth of data but it is still encouraging when you consider a year’s savings would equal $6,252.  Is it worth a try or is there still more analysis we should do?

One Additional Thought

The thought that one month’s worth a data is not enough to make a decision stayed on my mind. On one hand we may be paying more than we have to each day.  On the other hand we may be making decisions on incomplete data.  Thinking it through it seems logical to conclude that if a domain or url doesn’t convert then don’t spend money on it.  The only down side would be if it converts later we would miss a conversion.  But what impact would 1 or 2 conversions have when we are talking about 100 conversions a day.

An Aha Moment

Then it occured to me.  Most of the conversions orginate from a few domains or URLs, but the bulk of the conversions come from many domains or URLs.  It is the old numbers game.  In the reports there were thousands of domains and URL-s that had only 1 or 2 conversions for June.  The number of sites that had just 1 or 2 conversions were not the same domains and URL-s that had only 1 or 2 conversions in July. If I had excluded the non converting domains and URL-s that had no conversions in June we would have eliminated 100-s of conversions in July.

My Conclusion

Based upon this data I concluded that sites should not be excluded simply because they don’t have conversions.  Another criterion should be placed on top of this that is simular to the cost per acquisition (CPA) measurement.  With the CPA you assign a dollar amount to what a conversion is worth.  If it is worth $40 is get a new customer,  $40 worth of clicks to get it is acceptable. Simularly if the non converting sites are costing you a few pennies a month to keep them then it may be wise to do so.  If they are beyond a thresshold you determine to be unacceptable, then by all means exclude them.  As with a CPA, some conversions are not worth the price.

Technorati Tags: Paid Search, PPC, Analytics, Campaign Tracking,  google content network, google site targeting

Pay Per Click, Analytics | No Comments »

There is danger in limiting landing page design and implementation to just five rules but I like Billy Blog’s presentation.  According to Billy you should:

Keep it focused: Force your page to have only one goal. If you are required to have two, then choose one as a priority and emphasize that one. Remove excess baggage like advertisements and navigation bars so that visitors have only two choices: convert or leave.

Give a good second impression: People don’t see the landing page first, they see your PPC or banner ad, e-mail or even search result. Build relevancy between them. Use the same messaging from the first point of contact to your landing page. I try to use the same words or exact title from advertisement to landing page. Keep creative consistent too. One thing that is often overlooked are offers, if you put an offer on your page, put it on your ad and vice versa.

Target your biggest (middle) audience: There’s 3 types of visitors. Ones that’ll convert no matter what, ones that might convert and ones that are just looking. You only need to grab that middle audience. A landing page is not meant to please everyone, it is meant to drive conversions, meaning pleasing only those that will convert! For example, putting less information on a page will drive away people who only looking to learn more, but help push along those that are looking to buy.

Stop talking about yourself: Customers come to your page to read about the product, not your entire company history. Talk about yourself to the extent that it will calm visitor’s fears about your legitimacy and quality, else you’ll clutter the page and intimidate the visitor with blobs of text. Third party validation logos (BBB, Hacker Safe) and quotes from happy customers are often enough.

Use a product shot: So a cheetah might be a great symbolic way to show how fast the computers you’re selling are, but really you should be showing your computer. Customers come in and will only spend a few seconds to see if they’re in the right place before hitting back and so you need to communicate what you’re selling fast. Why distract them with symbolic images, when your product is what you want them to buy? If you’re service oriented, then people probably are a good idea, but make sure they directly represent what you’re doing.

For more from Billy go to Billy’s Blog.

Technorati Tags: Landing Page Optimization, Landing Page Testing , PPC, SEO, Paid Search

Pay Per Click, Blogs, Landing Page Optimization | No Comments »

Online Activity
Tracking clicks to online sales is simple. Simply place the appropriate tracking code on your Thank You page and you know which search terms are and are not contributing to your sales. The same holds true for the submission of online forms, appointments, appications, chats, e-mails–pretty much anything that occurs online.

Offline Activity
Tracking clicks to offline sales is a bit more complicated. You can instruct the customer to mention a code they obtain online to claim a offline discount. Or you can have them print out a certificate or coupon with a bar code that they take to the store. You can even associate each search term with the the code or bar code on the coupon and certificate to track which keywords are performing. This could be very tedious if you are running thousands of search terms but is probably worth the effort.

In Between Online and Offline
Phone calls are not necessarily online or offline, but something in between. It is not as easy to associate a search term with a phone call in this “in between” world. The way to accomplish this is to assign an unique phone number to each search term. This could get very cost prohibitive if, like with most campaigns, you are running 1,000 to 10,000 terms. Imagine securing 1,000 or more unique phone numbers and placing them on 1,000 or more unique landing pages (actually you can avoid multiple landing pages with dynamic content) and setting up the campaign so each search term is associated with its unique phone number. The information learned from this will probably save thousands of dollars in the long term but still may not be worth the effort.

Need for a Click to Phone Call Tracking Technology
Ideally there would be a technology that automatically assigns a phone number to a search term, dynamically places that phone number on the landing page, forwards the calls to the real phone number, and tracks everything for you. Hundreds of phone numbers may still be required but let the technology handle the assignment and minimize the number of phone lines required to the absolute minimum. It seem like as much as we have done with computers we should be able to fulfil a simple requirement like this.

Technology Delivered
siteEDGE Agency clients have access to this technology. We still have to determine if the potential cost to be saved outweighs the cost of integrating the technology, but once we pass this due diligence we will not hesitate to use every tool at our disposal to minimize money wasting search terms and to maximixe money making terms.

If you would like more information on this technology feel free to call Steve Thompson at 816-587-8880 x102 and we will help you determine if it is right for you.

Technorati Tags: Paid Search, PPC, Analytics, Campaign Tracking, Phone Call Tracking

siteEDGE News, Pay Per Click, Tracking, Analytics | No Comments »

Google just launched a tool that makes it easier for anyone to figure out what they can do to improve the performance of their landing page(s). This is good news for PPC advertisers. As click costs continue to rise, it’s really important that you convert as much of the traffic that you are paying for as possible. But before you jump head first into launching an experiment, here are a couple of tips to help you get started:

1. If you’ve never conducted any type of testing before, don’t create a really complicated experiment. Limit your first test to two to three elements on your page (e.g. Title, Call to Action Button, Image).
2. Make sure you allow enough time for statistically significant results. If your test page gets 10,000 visitors per day you’ll see real results much sooner than someone whose site only gets a couple hundred visitors per day.
3. Don’t stop with landing page testing. There’s a lot of other “stuff” that isn’t directly on the landing page that has a huge impact on your bottom line as an advertiser. Be sure to evaluate other contributing factors such as campaigns, keyword grouping, offers, and your visitors’ intent.

It may take you several months to become a good strategic tester, but if you can get a highly tuned campaign showing performance improvements measured in multiples (this isn’t uncommon), its well worth your time.

Pay Per Click | No Comments »

Tim Daly recent article in MediaPost raised this very question. Is Impression Fraud More Dangerous Than Click Fraud? If you would have asked me this a month ago I would have said no. At that time not very many people were talking about it. The more we talk about this the more people will start recognizing this as an option that can be used to attack the competition.

As defined at Wikipedia, “Impression fraud is an insidious variant of click fraud in which the advertiser is penalized for having an unacceptably low click-through rate for a given keyword. This involves making numerous searches for a keyword but without clicking of the ad.”

This devious senario involves two competitors competing head to head. Even if one is a big company and the other is small the playing field is pretty level because this is the internet. Lets say that the smaller company has been writing better ads and has better landing pages and is spending the same amount each month as the large company. The larger company can spend ten times the budget of the smaller company but the landscape just doesn’t allow it. The playing field is leveled, or is it?

Now lets say that the larger company knows a bad month can wipe out the budget for the smaller company and decides to play dirty. The larger company pauses all campaigns for his company and artificially inflates the impressions with some bogus automation tool. The larger company is safe because they, temporarily, pull out of the campaign. Because of Google’s, and now Yahoo Search Panama’s, quality score penalty they will start paying more for a click than they did prior to this event. This will increase their cost per acquisition and possibly cause them to stop the campaign.

You might ask why am I providing all this information. People who have never thought about this have now been provided with an option that they didn’t previously consider. I pondered this very point but decided the best defense is education. Although not fool proof there is at least awareness about click fraud. Enough people are paying attention to click fraud and complaining when they see something out of the ordinary. This is forcing the engines to make appropriate refunds.

For impression fraud awareness can bring this same kind of reaction. If you see a unusual increase in impressions and notice an unusual decrease in click through rates for some keywords then you will know to at least consider this. If your cost per click rises for these keywords you might consider contacting the publisher to file a complaint. If enough of us are doing this the engines will start paying attention and, at least, give this the attention now provided to Click Fraud.

As always these are my thoughts but I look forward to your point of view.

Technorati Tags: Click Fraud, Impression Fraud, Search Marketing, Pay-Per-Click

Pay Per Click | No Comments »

Google’s Black Box

Advertisers are one step closer to having the information from Google required to determine how to improve their click cost on Ad Words. Up to this point we knew only of the black box that determined the Quality Score for the Ads. We know that the price we pay for a click is based upon a good Quality Score. Previously we could only infer that our ad and/or landing page was bad or good when our cost per click went up or down, respectively.

Why a Quality Score?

The enforcement of Google’s Quality Score makes sense. Its ultimate purpose is to nudge the advertiser toward more relevant ads and landing pages. As advertisers we see that an ad is doing good or bad by the click through rate. Since, reportedly, the Quality Score factors in the relevancy of the landing page, among other things, the introduction of a gauge other than the click through rate is welcomed.

Quality Score Grades

Google is now assigning a Quality Score grade (Great, OK, Poor) to each keyword. We don’t know where Google draws the lines but I think it is safe to assume that a grade of “Great” indicates you are getting a better cost per click that the guy with a “OK” grade. Also a Quality Score of “Poor” is likely to yield a higher click cost than “OK”.

Improving the Grade

The first thing I will do with the Quality Score grade is to compare it with what I thought I knew. If I have a “Great” score does it mean I will always have a good click through rate? Can I detect any variations and trends in this? Is there any correlation between a “Great” score and how many times I repeat a search term on the landing page? Can I overdo it and cause my score to drop? Can I detect other patterns related to a poor grade? Some may look at this new keyword column as a simple monitoring attribute. I see it as a little more information that can help perfect an ad word campaign.

Check back here periodically to see what I learned. Feel free to share your thoughts on this new addition.

Technorati Tags: Pay per click, Google Quality Score, Search Engine Marketing

Pay Per Click | No Comments »

If you’re an AdWords advertiser, make sure you read the post on the AdWords blog, Quality Score Updates.  As you may have noticed, your campaign management dashboard now contains a new Quality Score column.  (To see this, click on the Keywords tab in your Ad Group, then Show/Hide columns, then select Show Quality Score).  According to Google, this is a precursor to a change that will occur next week in the AdWords algorithm.  So if you see an “OK” or “Poor”, you should either select a keyword that is more relevant to the product or service that you’re advertising or make your ads more closely related to the keywords.  If you make the changes now, you’ll be better positioned to benefit from the algorithm changes which will reward those with “Great” quality scores with lower minimum bids.

For another good overview of the pending changes, check out what Andy Beal at Marketing Pilgrim has to say.

Technorati Tags: google, adwords, quality score, PPC

Pay Per Click | No Comments »

We run a lot of paid search campaigns for companies with service area constraints. In some cases, the client has already been running the campaign before we take over. So we get the opportunity to see the good, the bad and the ugly when it comes to local ppc marketing. Here’s a list of the top mistakes:

1. Running a National Campaign Without Using Location Specific Keywords

We’ve all seen the results of this one. You type in a term like “carpet cleaning” on Google. You click on an ad and then find out that the advertiser only services the greater Atlanta area and you’re in Kansas City. The only winner in this equation is Google. If you’re running a national campaign and you only service a certain area, make sure you add the location(s) served to all you keywords (e.g. carpet cleaning Atlanta).

2. Running a Local Campaign and Using Only Location Specific Keywords

This happens when an advertiser selects a specific city when setting up where the ad should be displayed and then includes that city name as part of every keyword (e.g. carpet cleaning kansas city, duct cleaning kansas city). People who live in the service area may not always include the location in the search query, so you’ll miss out on legitimate traffic.

3. Not Mentioning That You’re a Local Merchant

If you are advertising to a specific area and you have a physical location in the area, flaunt it! People like doing business with members of their community.

4. Not Testing Ads and Landing Pages for Market Differences

If your business services 5 metropolitan areas, don’t make the mistake of assuming that each area will respond identically to the same offer. It’s very easy and cheap to run separate campaigns and offers for each area. It’s also inexpensive to develop specific landing pages (where the person ends up after they click the ad) for each market. In some markets, it may also make sense to run campaigns in more than one language (e.g. a merchant in Houston may want to market directly to the areas’ Hispanic population). If you choose to do this, make sure that you can execute seamlessly from the click on the ad to the phone call or visit to your location.

5. Not Using a Local Tracking Phone Number on Your Landing Page

This reasoning here is very similar to what was discussed under # 3. The local tracking number is important if, like so many local merchants, your business thrives on phone calls. You’ll be able to gauge the performance of your ad campaign.

6. Not Examining the Click Through Rate (CTR) and the Conversion Rate (CR) at the Market Level

If you fail to examine the CTRs and corresponding CRs at the market level (e.g. region or city), you will end up investing too much in a campaign that gives you very little bang for your buck, and not enough in a campaign that’s a real winner. Here’s a guideline for determining what to do: High CTR + High CR = Invest More Money; Low CTR + High CR = Improve Ads then Invest More Money; Low CTR + Low CR = Try Improving the Ad and Landing Page before making a decision; High CTR + Low CR = Improve the Landing page then Invest More Money. For a local campaign that’s not getting a ton of impressions and clicks, it may take several months of testing before you get definitive results.

7. Not Taking Advantage of Yahoo Panama

We have seen significant improvements in available clicks and overall campaign costs for our local clients post Panama. The key is going in and properly setting up a local campaign on Yahoo for all the new targeting features. If you were running a campaign on Yahoo/Overture, not Yahoo Local, you need to go in and create new campaigns and ad groups that target your service area(s).

8. Running a 24/7 Campaign

This is particularly wasteful unless you are selling a product online (your site has a shopping cart). If not, make sure you set up Google and MSN Ad Center so that your campaign only shows during your business hours. (This function is not currently available on Yahoo).

9. Not Tying in Your PPC Campaign with Local Offline Advertising

Research shows that there is a strong connection between the airing of national TV commecials and online searches related to the product. While the connection may not be as strong for local merchants, it still makes sense to buy all the keywords related to ads you’re running on local radio or TV. This can be especially powerful if you use a local celebrity endorser in your radio or TV spot. People often remember the celebrity’s name and the product or service (e.g. Buck O’Neil mortgage) even when they forget your brand name (e.g. James B. Nutter mortgage).

10. Not Bidding on Your Company Name

Even if your site is number one in the organic rankings for your brand name, you should still include your brand name in your list of keywords. At least this will keep your competition from occupying one available paid search listing space.

If you’re not a local marketer, here are some other great tips.

Technorati Tags: ppc, local ppc

Pay Per Click, Local Search | No Comments »

Like many I have purposely delayed converting client accounts over to Yahoo Search’s Project Panama. I had a sneak peek months ago and knew about the new fresh look but was not convinced our agency should be the first to expose our clients to this unproven platform.

I learned one year ago at the 2006 SES Conference in New York directly from Yahoo that Panama would be using a quality score to determine the cost per click opposed to the traditional Overture bidding that we have grown accustomed to. There was a small uproar with some attendees when this was announced. To be honest I was kind of relieved. I still cringe when I think of the inflated bid prices that have been artificial set by hyperactive bidders who didn’t take time to romance the campaign instead of trying to seduce it. Now these individuals will have to spend time in writing good ads and landing pages, or pay more than the ones who do.

I have seen the blog postings of some who are unimpressed with Panama. A reoccurring theme is that they are happy with the new user interface but will not be satisfied until Yahoo Search increases the volume of available searches.

I actually agree with both these points but will point out one thing I have not yet seen discussed in a blog posting. Yahoo’s implementation of local search has paid off well for us and our clients. Our firm has a heavy emphasis on geo-targeted searches and have campaigns that target up to 20 DMAs for one client. The goal, of course, is to generate traffic only within these geographic pockets. Up to this point Google and MSN have allowed this but Yahoo Search was a non performer. And no, Yahoo Local was not the answer.

We didn’t want to miss the coverage that Yahoo offered so we made location a part of each and every search term. This is very tedious and made for a lot of search terms but we have done this for years with many clients and thought the results were worth the effort.

The initial results of our Panama geo targeted campaigns are extremely promising. The volume of searches has increased of course, but with the added benefit of multiple ad groups and the ability to test them we have also increased the click through rate. Since we have a good quality score, we are paying less for more.

What is truly note worthy is the increase in conversions. We have to study this over a longer period but first indications are, in proportion to the clicks, we are getting more conversions from Yahoo than with Google. For the client in this particular study we track conversions by telephone appointments with each call recorded. We are seeing cases where the phone appointments are up while the overall click costs are down. We contribute this to Yahoo having, for this industry, a better conversion rate than Google.

Again these are early results but look promising. I will give it another month and report back with further details. Let me know what you have seen.

Pay Per Click, Blogs | No Comments »