Let’s say that you attend a church in Charlotte, NC. and your pastor asks for your help in putting together a PPC campaign to boost awareness of the church and increase their membership. So you go ahead and set up a campaign, you do little bit of keyword research, and notice that there are some fairly well trafficked keywords that you can use. So you go ahead and put them in the campaign, and come up with your ad text.

A couple of days later the pastor asks to see how well the campaign is going, so you pull up the AdWords campaign, and you both see that there’s a heck of a lot of traffic coming in, in fact it’s a good job that you set a daily budget. Proud of the job you’ve done, you decide to show off, and show the pastor around the analytics package that you set up on the site. It’s when you click on the PPC keywords searched that you realize that there’s a problem…

Your broad, un-geotargeted keywords “Charlotte Church” has been showing up whenever anyone has done a search for the Welsh singer of the same name, and the vague copy that you wrote has her fans clicking through, costing your church money.

So what could you have done? Well, for one you could have done a much better job of targeting your audience. But you should have also specifically not targeted the wrong audience. Negative keywords are designed for exactly this purpose. Examples that you could have used for this would have been -welsh, - singer, -pregnant (if you’re up on your pop culture news). Using these negative keywords would ensure that you’re not paying for traffic that you don’t want. Simply enter the negative keywords into the engine, and if they’re detected in the query, your ad doesn’t display.

If you think the example is too contrived, here’s another one. Let’s say you’re an attorney in Lincoln, NE. What negative keywords would you possibly want to use for your PPC campaign? How about -abraham? That’ll stop your ads showing up for all of those history buffs / report writing school children. Maybe even -town car, unless you work with auto accidents…

The clock struck thirteen…

No, I’m not saying that MSN is watching everything that you do, I’m referring to the part of the book where they happily announce that the chocolate ration has been increased to 20 grams, despite it previously being 30 grams. A week or so ago I received an email from MSN letting me know how excited they were to let me know that I’d been selected to participate in their new beta program, where all of my ads would now automatically be displayed on their content network. How sweet of them to be excited for me, I felt special… but wait a minute, what if I didn’t want to be on their content network? Well, there’s a way to opt out they say. So here’s the way to opt out. Go to your account, click on a campaign, click on an ad group. Then look for the text to “Select ad distribution”.

Clicking on the actual text expands your options.

…and after a few seconds there it is. All you then have to do is hit save, and wait for the save dialog to pop up. Then repeat for each of the other 500 ad groups that you have live in their system. Or alternatively you can do as we did, and email MSN support, giving them your account details, and they’ll take care of it for you in 2-3 days (fingers crossed).

So having had an account migrate to the new MSN adCenter Beta, I decided to take a look at what they’d improved. After my last post on their geo-targeting solution I’d hoped that they’d made improvements, then again it couldn’t get much worse unless they’d decided to only display lat & long co-ordinates and have you guess on the city names. So how have they improved it? Have they leap-frogged the competition, or at least pulled level with the best solution - Yahoo Panama? Nope, but they have at least pulled level with the Google solution. They now group the cities by state, sounds simple, but it obviously wasn’t as they’ve taken this long to actually implement it.

So why do I like the Yahoo solution better? Is it just because it’s got pretty pictures? Heck, yes. When I’m geotargeting a location that I’m not familiar with, I want to be damn sure that I’m hitting my target markets. With Yahoo, you pick the DMA and you get the instant gratification of seeing the area of the market on the image shade in. With this MSN solution, despite me working and living in the area, I have no idea what the heck the difference is between selecting Alexandria and Arlington… which one covers Fairfax? Vienna? Manassas? You can’t tell. Come on Google & MSN, get with the program and steal borrow from Yahoo.

As I alluded to the other day, I’ve written an article on the Ways to improve the Conversion Rate for your Website, this is my entry for the Marketing Pilgrim Scholarship Contest. The article talks about using Dynamic Keyword Insertion, Geotargeting, and making sure that you send people to the Right Landing Page. Take a look, and feel free to leave comments here or there.

One of the first ways to improve your click through rate is to make sure that your ad is compelling (I’m not going to write too much on this topic for reasons that I’ll reveal later on this week). I’ve recently added AdSense to this blog, and I noticed this entry pop up.

The ad on the right is absolutely fine, although I’m sure Google wouldn’t like you to click it (or at least click it and actually use the service). The one on the left though… what’s it about? The title is Search Engine #. I have no idea what that means, are they talking about increasing your traffic? Your SE exposure? There’s no information, and there’s no call to action. Maybe in the 2 lines of text? Top 6 websites for Search Engine # is the first line, and the second line isn’t even used. Are they hoping for people to click through to clear the confusion? If so, one would think that that would lead to a very poor conversion rate, which makes the campaign rather costly. I would have loved to have clicked on it to find out exactly who the company was, but unfortunately under the AdSense TOS I can’t click on ads on my own site. However, searching for the term in Google - Search Engine # - showed them at #1, so I clicked there. It’s one of those spammy arbitrage site, which means that they’re most likely auto-generating the ads. So actually I’m kind of happy that their copy sucks, hopefully it’ll not make them much money…

A while ago I posted about the number of campaigns you could have per account on Yahoo and Google. While I’ve come to accept these figures, I have found a problem with Google that could cause you a problem. AdWords allows you to set a campaign to one of three states - on, paused and deleted. The issue here is that any account that’s not in a state of on doesn’t count towards the 25 campaign max. So if you have 25 campaigns that are on, the system prevents you from adding another one. If you have 24 campaigns that are on and one paused, well the system will allow you to add one more. However, should you decide to restart the paused account… you can’t as it’ll make you exceed your max, so you would then have to move one of the campaigns to another account, thereby losing the history. Since there are 3 states, why not have paused count towards the account count? The idea behind pausing an account rather than deleting it, is that you intend to restart it one day, the way the system is currently set up, you may be prevented from doing that.

Update: I’ve deleted the rest of this post as anon points out in the comments that there is indeed a delete function available from the campaign settings tab in Yahoo - Panama. I just hadn’t noticed it. Thanks anon, whoever you are.

Piper Jaffrey recently put out a comprehensive paper on many things search and internet related entitled “The User Revolution: The New Advertising Ecosystem and The Rise of the Internet as a Mass Medium“. This 425 page report covers a wide variety of topics, but I’m going to concentrate here on those related to local search, and IYP.

Firstly the basics, 70% of Americans are now using the internet, with usage growing 18% per annum. The internet is now the second leading medium at home, behind TV, and is easily the leading medium at work. 82% of users are over 18, and they spend 88% of the total online minutes.

How pervasive is local search? Well, in response to how often users used different services, email came in at 100%, local search actually came in 2nd at 57.6%, ahead of Instant Messaging, online purchases, music downloads, blog reading / publishing, social networking, and the rest. Quite an amazing number really if it’s accurate, but does that mean that there isn’t that much room for growth? If the potential usage increase that can be had is slightly less than double, then is the market fairly saturated? Not by long shot, unlike a service such as IM, there is a large offline equivalent to local search, that can be migrated online and monetized (once a winning monetization strategy has been arrived at). So what do they consider to be the potential market for local search? Well, they have the current total local ad market (local search, IYP, and classifieds) at $4.489 billion dollars, which makes it 12% of the current offline spend of $37 billion. In their opinion online dollars in this area has the potential to reach 70% of offline spend, so the total potential that they show stands at around 2.59 bilion dollars. This shows that there’s clearly a lot of room for growth in this sector. Of course the next question is that of canibalization. Many offline publishers expressed concerns that moving online would chip away at their core offline business. According to this graphic, that isn’t happening. Instead the IYPs are expanding the overall market, without negatively affecting the print market which has remained fairly static.

So who are the top players in local search? Well, this paper has the answer to that too (and unfortunately none of the sites I work with are listed here… yet…). The interesting thing in this graphic is how far ahead of the rest Mapquest still is. When people talk about local search it tends to be forgotten, but it’s the clear leader right now. Idearc’s Superpages.com occupies third place, but they’ve been expanding their reach through deals with Google Maps, and more recently local.com.



What about different monetization methods for local? What does the report have to say about that?

The pay-per-call feature also offers a more familiar way for some local businesses to pay for search listings. We believe call-based interaction and tracking could ultimately account for 30%-50% of total local search listings.

Well, that’s good news for the company that I work for, given our current business model.

Now for the part of the report that I don’t agree with. They show a table that displays the effect of localization on search keyword process. In virtually every case, they show the price for the general nationwide term being less than that of the ‘local term’ version of the same keyword. They don’t define exactly what this means. This could be a national campaign that used “Chicago Dentist”, a campaign targeted to Chicago that targets “Dentist”, or a campaign targeted to Chicago that targets “Chicago Dentist”. To make the claim that this shows that localization is going to drive more revenue for the search engines because of this actually contradicts a statement made on the same page that the longer the tail of the search term, the lower the CPC. Surely “Chicago Dentist” is a longer tail phrase than “Dentist”? Plus, I can guarantee that once you get out of the top MSA’s, the CPC is going to be less than that for the ‘general term’, so to make a blanket statement based on data for 5 of the top MSA’s is misleading.

The data for the above was provided by Yahoo Search Marketing on January 26th, 2007, which means that it wasn’t too long before it was invalid, as one week later YSM implemented quality scores on their Panama engine, which changed the pricing scale on their system from a straight auction style to the Google-like system. Also if the data is to be believed, there’s a fantastic opportunity in that the top position for one of the most searched on terms on the internet for the last few years, as well as that for one of the most searched on terms so far this year, can apparently be purchased exceedingly cheaply…

(thanks to Greg Sterling for pointing the Piper Jaffrey paper out).

I’m trying to verify some click charges on one of my accounts, so I tried to log in to Yahoo’s Panama. Here’s what I, along with everyone else in the office, have been seeing for the past 15 minutes.

Fantastic… Hopefully it’ll be up again soon…

Update: as is typical, it was up by the time this post was published…

Looking through one of my Google AdWords account I saw the new quality score displayed. As Yahoo had put in a little shaded in bar showing your quality score on a 5 point scale, Google decided they’d get in on the act. Instead of 5 non-defined points, Google went for 3 vague descriptive words:- “Poor, OK, and Great”. One interesting this is that they show you what the minimum bid for a particular keyword is. As you can see in this example culled from one of my accounts, the “Great” keyword has a min. bid of 4 cents, the “OK” keywords have a range of 6 cents to 20 cents, and the “Poor” keywords all have a 30 cent min. bid.

Quite a difference there. If all of the keywords manage to somehow come in at their minimum bid, or that multiple holds at higher prices, that’s 7.5 great keywords for the same price of 1 poor keyword. Now in reality that multiple doesn’t hold, but there is a differential. Looking at the average CPC for these keywords, the “Poor” keyword is costing us about 33% more than the “OK” keywords, now as to whether the “Poor” keyword is worth that from a conversion perspective, that’s a different story, but if you see this in your AdWords, that’s something that you do want to look at.

Now, I’ve looked through several campaigns, and it appears that “Great” always has a minimum bid of 4 cents. “OK” ranges from 5 cents to 20 cents, and “Poor” is 30 cents and up (I saw some as high as a dollar - naturally these were ones that had been labeled “Inactive for Search”).

Since we hit the end of the month we’ve been playing around with the reports in Yahoo’s Panama. A rather nice one that isn’t available in Google is the “Performance by Geographic Location” report. This report allows you to see a breakdown of the spend in your account by location. If you are targeting your campaigns in different geographical locations this is an easy way to pull that data and see exactly where your campaigns are working.