Ken McGaffin is Chief Marketing Officer with Wordtracker.com. He writes regularly on link building and online public relations. He is the author of the highly acclaimed Linking Matters Report.
Our discussion focused on recent changes and upcoming plans for Wordtracker, and tips and techniques for keyword selection.
Eric Enge: Tell us about recent updates to Wordtracker.
Ken McGaffin: We have been really working on the quality of the data that we are using, and also the statistical viability of it. One of the things that we will be doing in this coming year is introducing new data sources. At the moment we get our information from the metacrawlers; and so we’ve had that independently reviewed in terms of how valid that information is.
We were talking before the interview started about the importance of market research, and marketers making decisions based on market research that they do. Keyword research itself can be seen as a very important piece of the jigsaw and very important in terms of giving marketeers insight into what the customers are looking for.
One of the things that we’ve been doing is sampling our data on a daily basis. We do that ninety times to then put together our database which covers a ninety day period. Looking at how valid that is as a statistical inference. The good news for us and for all our customers is that that’s a very significant sample that we take. Particularly, for high volume keywords the sample is very, very useful in terms of predicting the relative strength of particular keywords.
Obviously, as you get deeper and deeper and deeper into the long tail, search volume for particular words is quite little, and you are prone to a greater degree of error. But, that’s really straightforward and simple mathematics, and something that you would expect.
Eric Enge: Indeed. I also see that you recently launched a UK specific version.
Ken McGaffin: Yes. While there is a great mutuality of the words that are used in the US and the UK, there are also some very specific differences. And so, we’ve been very carefully looking at that. One of the problems that we then get is coming up with data sources. We’ve been getting information sources from both search engines directly and from ISPs. When you have that you then need to figure out how to merge that data together.
It’s not simply a matter of putting it all into one tub, because that wouldn’t be accurate. There is duplication in the searches, so the search is done on a particular engine that may well be picked up by one of the ISPs that we are also monitoring. If we just through all of that together and said right, there you go, then that would overestimate some of the numbers. We’ll try to build systems that can track that and can pick out those multiple searches. That’s been a bigger technical problem then we expected. However we are on top of that. We’ve been collection UK keywords from a number of sources over the last sixteen months. We are very close to that being a very dependable source.
Eric Enge: Right. So what percentage of total search is your data based on at that point?
Ken McGaffin: It’s about 0.7% of all searches done per day across the Internet. That’s a significant number, because once we checkout all of the multiple queries from that, we are left with between three million and four million keywords that we add to our database everyday. Now, we’ll then take that sample ninety times; and obviously if you are taking a sample ninety times, it’s much more reliable if you are taking that one sample on it’s own in isolation.
Eric Enge: Right. With the ISPs that you are adding how much larger will that sample get?
Ken McGaffin: That sample can get huge; one of the things that we would say from that is if we were getting something like from 4% to 5% of the daily traffic, then that would be a very significant figure. So, that would be very robust basis on which we’ll build predictions of actual traffic.
Eric Enge: Right. So, the range you are targeting is 4% to 5%?
Ken McGaffin: Yes.
Eric Enge: Any idea when the ISP enhanced data would be what’s available to the public?
Ken McGaffin: We are really testing this out at the moment with the UK version. It is available now for people doing specific researches in UK. We are now going to be adding that experience to our US database, and we’ve already got agreements in place for additional sources in the US. We can’t say exactly when that’s coming on board, because we want to ensure the quality of our keywords. We would much rather have high quality then inflated high numbers. We are actively assessing that now.
Eric Enge: Do you have any advice as to how people using the tool should view high volume keywords versus long tail keywords, in terms of how they apply the Wordtracker data to their websites?
Ken McGaffin: As you do get deeper and deeper into the long tail, the actual predictions, obviously are not as robust as the ones for more popular keywords because the numbers are small. People have bear that in mind. One of the things about going deep into the long tail, is it gives you more information about your marketplace; it’s giving you the types of searches that the people are doing.
That in itself is a useful way of understanding your market, and so what I would say to people is yes, do the research and go as deep as you can. We provide up to a thousand results from a specific phrase. Do that research, but look carefully at the long tail and add your own marketing knowledge to it. So, what Wordtracker is doing there is helping you to find the keywords, helping you in the creative process, and giving you an indication of what the likely volume is.
People can then try it in their PPC campaigns. They can try it in their content, and they can do different things with it; but that’ll give them more information about the actual customer behavior. As you get deeper and deeper into the long tail, it’s a good idea to create content using that data, but it’s also important to monitor your results even more closely.
Eric Enge: Right. There are also ways to smooth out the data. So, if you are selling green widgets and blue widgets and yellow widgets, and you type green widgets into Wordtracker, and then you see that sometimes they type cheap green widgets, or buy green widgets, and you see all of these things that start to be more long tail terms; and you could see what some of the modifiers are, right?
Then, you can go and see what happens with blue widgets, and you see oh, the modifiers are a little different. Then, you look at yellow widgets and the modifiers are a little different still. If no one ever typed in cheap yellow widgets, but people did search on cheap blue widgets and cheap green widgets, you probably want to bid on cheap yellow widgets.
Ken McGaffin: You’ve hit it spot on. Look at cheap flights for instance. If we wanted to go into the long tail of cheap flights, what you would find is that people nearly always use a destination as a qualifier or as a modifier as you call it. However, if you looked at cheap travel, that’s not the case. People don’t put in destinations when they are looking for cheap travel.
One of the great things about doing keyword research, and particularly from a marketing point of view is you get to see those patterns. Because, language itself is not random; there are patterns and we are all creatures of habits. And so, we can pick out little nuggets like that, and then use that information to make assumptions.
Eric Enge: Right. Let’s talk about other ways marketer should be thinking about keyword research, and why that’s relevant to SEOs.
Ken McGaffin: People get exposed to all sorts of advertising messages, we get a huge number everyday. If an ad catches their attention; one of the things that they do is to search online after seeing an ad, such as an ad on a billboard. One of the things that advertisers often forget is that people do searches after being exposed to advertising. If you can bear that in mind and do the optimization before you launch a campaign such as a billboard campaign, then you’ve got a good chance of increasing the ROI on your overall advertising campaign.
I will give you an example, the Danish Butter Company, Lurpak; they had a billboard campaign here at Christmas. And, it was a very simple message; it was “leftovers we call them ingredients”. The message was simply that if you use butter like Lurpak, then you could make those leftovers into really great ingredients for some great food. The interesting thing about that is what are people likely to search for after being exposed to that ad? The answer is they are going to look for leftover recipes, such as leftover Thanksgiving recipes.
We did a whole host of research, and we found something like four hundred and twenty very significant keywords all around the phrase leftovers. The problem as far as Lurpak was concerned was that they didn’t rank for any of those searches. We did the searches up to the first fifty pages, and they just weren’t there. That’s a real shame, because they are absolutely missing out on the interest that their ad generated for leftover recipes.
The simple thing would have been of course to optimize for those popular leftover phrases before they launched the campaign. And, if they had done that, they would have picked up many of those subsequent searches that people would have done. That could be a tremendous source of revenue, and of course the real bonus on top of that is they would have picked up all of those people who were seeing leftover or searching for leftovers without having seen the campaign in the first place.
There is a double win in there, in terms of the benefit that marketers can get. One of the things that we have to be very aware of in the past, particularly within large companies, we silo the activities of search engine optimization, public relations, Direct mail, and advertising. The lesson from campaigns like that is, let’s not keep them separate; let’s think about things right across the border, across all our marketing disciplines. Then, we can all work together and produce tremendous synergies and really increase the ROI on our campaigns by doing that.
Eric Enge: Yes, indeed. There is a really significant interaction between display advertising campaigns, whether they are offline or online oriented and your pay per click and organic search campaigns. Those things are going to get harder and harder to keep treating like they belong in this complete vacuum.
If your display advertising is basically about reach and eyeballs, and you are successful in that, but at least some of the conversion is happening online, you are going to want to use words that are trigger phrases related to your business in your display advertising, that are also ones that you rank for. I would imagine it would have a very dramatic impact on your results.
Ken McGaffin: Yes, and for very little cost, really just the realization that that is the dynamic that is going on. Advertising campaigns are incredibly competitive. We used to be able to do the big ad and be certain that we were going to reach the mass audience that we wanted to.
Now, we’ve got to change our thinking in advertising, and we’ve got to just give ourselves that little bit on an extra edge. Keywords can be a source of creative ideas for offline advertising; and may just be that extra edge that an advertising company is looking for.
Eric Enge: Right. How important is it for SEOs to develop their keywords skills, and how can they do that?
Ken McGaffin: Once you get a passion for keyword research it never stops, because you learn more and more about your customers. To help with training, we launched Wordtracker Academy some years ago, and that has now proved to be extremely popular. We share a lot of our thinking on how to use keywords. The whole idea of that is to stimulate people to think, and also to attract contributions from experts in the area.
I think one of the things that search engine marketers must really do is be aware of what’s going on, read what other people are thinking and what they are doing, and constantly challenge assumptions. The response to our academy has been tremendous. Often search people are self-taught and sometimes you get into an area you don’t really know much about.
That’s where you’ve really got to start digging and listening to other people and reading, so that you can develop the skills. After that it’s a matter of testing and trying out and doing. One of the things that we are going to be introducing in about two weeks time, and you would be the first to know about it Eric, is we are bringing out a series of seven videos which show not so much how to use say the Wordtracker tool, but how to use keywords in different ways, each of which will help a business achieve it’s objective.
We are going to be distributing those free, and we want as many people as possible to benefit from them. In addition, this is something an SEO can show their clients, to help them understand part of what is going on in search engine optimization.
An educated client is a dream. They don’t need to know it as in depth as the SEO themselves, but if they know the principles, then of course they are easier to work with, and are more enthusiastic in terms of spending the money that’s required. The seven videos will help people who are starting out in their career in SEO to think about things they should be thinking about.
For the more experienced SEOs, it’s something that they can certainly pass on to the client and say these are the principles of keyword research. I think the more education and the more understanding that there is about SEO, the better for all of us.
Eric Enge: When is the right time during web site development to begin using a keyword tool?
Ken McGaffin: As we’ve said before, keyword research is just really market research. It’s about knowing about your customers, and so that’s the first thing an entrepreneur of any kind does, they get to know their customers, and asks their customers what they want. That’s how a successful entrepreneur avoids making big mistakes and learns how to concentrate on a real idea that has money making potential.
For anybody setting up a website, that’s the very first thing that they should do. I was a traditional marketing consultant for 20 years before I joined Wordtracker. I think that sort of research is something many small businesses don’t think about doing, yet I am convinced that’s the key to success. No matter what venture it is; do your keyword research first of all. You’ll get an idea of the size of the market, but you will also get an idea of the niches that exist within that market, and you never know you may well be lucky enough to be the person that spots the niche that hasn’t yet been exploited.
Eric Enge: Right. Basically, the knowledge of the keywords customers are using bring strategic value.
Ken McGaffin: Yes.
Eric Enge: This particular aspect of learning about your customers can advise you about basic site structure, because when you build the page, there are a few elements to each page that it is going to become known for. A good page may get traffic on hundreds of terms, but content wise it’s going to become known for one broad brush type concept. That’s what’s likely to appear in the anchor text of its in bound links, and the title of the page. Those are two of the biggest influencers in search engine rankings.
Ken McGaffin: I talked at length with Bryan Eisenberg about this. Bryan talks about the intent behind a keyword. Keywords not only tell you about what people want, but they tell you what they intend to do with that particular search. An example I use is the difference between digital camera and digital photography. Someone searching using the work digital camera is probably likely to want to buy a digital camera.
If they search on digital photography probably want to improve their skills or learn about digital photography. They may well of course be prime targets for selling digital cameras at some point in the future. But, the way to get them on side is to provide them another bit of decent content round about how to do photography rather then trying to sell them a camera straight off. I think what Brian was saying in terms of the intent behind different keywords is a very important lesson for marketers to learn.
Eric Enge: Yes, it is. If you are lucky enough to have a dynamic website you can personalize what people see based on the incoming keyword. But, even if you don’t have a dynamic website, you can map out the kinds of intents that people have when they come to the webpage on these search terms, and make it evident and visible so that the person who really wants to do research can find what they want and go down their track. And, the person who is more ready to buy can go down their track, and they don’t have to hunt for this stuff. If you do this, you’ll reduce your bounce rate, and increase your conversion.
I want to tell you a little bit about one notion that we use here; and I call it spreading the field. I wrote about this recently, but I’d like to just get your comment on it. When we are helping people with a web strategy and we have decided to emphasize certain things on their website, one of the things we do is we look at the competitiveness of various keywords, and everybody wants to win on the mega keyword and that’s fine.
What we do is we always suggest something like they pick 10 pages to link to from the homepage, because we are going to really jack these pages up. We are likely to take two very competitive pages competing for very competitive terms or maybe three.
We will take a handful that are a little less competitive, and then another handful that are less competitive still, and then yet another group that’s even less competitive. The purpose of this exercise is to have a range of things, so you can see which of those things moves the needle.
Your long tail terms are down at the bottom of the list, and will probably move up in rankings first; but if you are lucky you will find out that hey, everything moved in a way that was significant to my business except for the two most competitive ones. Then you can make decisions by analyzing how competitive keywords are as to what kinds of pages you can increase your revenue on immediately by taking some actions.
Ken McGaffin: That makes perfect sense, and I think that’s a really good way of putting it; spreading the field. The term appeals to me, but I think you are absolutely right. If you are giving yourself as many chances as possible to succeed, then that’s a good thing. But, what you then have to do is you have to focus in on what is actually working for you.
Once you see an impact like that, then you start building on it. That’s why one of the things that is so important in this whole keyword business is your web analytics program, because of the insight that it can give you about what are the real money making terms. What you are described sounds to me like a very good way of doing that.
Eric Enge: I’d be interested to have you talk a little bit about how your thesaurus tool and how it works.
Ken McGaffin: When I go and give talks at conferences people come up to me afterwards and say, I never thought of doing that related search. We often use chocolate as an example. If you are building a chocolate gifts site, of course you would be interested in the term truffle, which doesn’t contain chocolate, but it’s a related term.
People can get really into that mindset of no, I have got to have chocolate terms, chocolate terms, chocolate terms; whereas what you’ve got to do is you’ve got to think around that what are the terms around chocolate that maybe don’t contain the word chocolate at all.
That’s one of the things that what we do with our related terms tool. One is a straightforward and simple thesaurus, which is right down at the bottom of your results. Some people often miss that, but then what we also do is yes, we do search. We put a phrase in, then we search top sites for that phrase and we look at the keywords that they are using.
What you find when you do that is lots and lots of hidden gems, because you’ve got a whole army of people thinking about this particular subject; bringing their own specialty to it. And so, for someone who really wants to dominate a market, thinking in that way rather then going down a single path, they can win in the end. It’s just one of those really good brainstorming tools to use.
Eric Enge: Right. It gets you to think laterally or horizontally and really can give you some quick insight, and I would bet that it’s an underutilized tool at this point.
Ken McGaffin: Yes. One of the concepts we’ve come up round about this is a concept called keyword creativity. Keyword creativity is about looking at keywords as a very creative way of thinking about things. We can change our ideas and then have some fun playing with words, putting the words together or concepts together in strange ways, and we can really come up with some very interesting concepts. I think this is one area where SEOs have an awful lot to learn from the advertising industry.
In my office here, I’ve got a whole shelf dedicated to creative thinking and developing those ideas that are going to be winners. So, I would really recommend SEOs use some of the tools that have been established in the marketing business for many, many decades. I think there are some wonderful creative resources there that can really help.
Eric Enge: Right. Do you have any specific tips and tricks for people who want to expand their PPC campaigns.
Ken McGaffin: Let me give you three ways that we use PPC. These are the basic concepts; I am not an expert in this area. The very first thing is to give us ideas right at the start of a campaign. If we are doing PPC, and again if you use only the research tools the PPC providers will give you, then you are getting the same information as everybody else. One of the things with Wordtracker is we have no interest whatsoever in selling advertising based on keywords. We give you the keywords, and so just by using Wordtracker you are already ahead of the game, because then you have an additional source.
That’s one of the things that we would always say to PPC people, it’s a great way of thinking outside of the box, and it’s a great way of getting information that your competitors don’t have; so that’s no. #1. No. #2 is when we get those gems, the ones that start bringing you in money then you’ll want more of them. Wordtracker is a very good tool to help you find more of those words that you know are working. So, again it can expand the number of words that are actually bringing you money.
Then, the third major area that I would be looking at is when a campaign isn’t working. When we have done something, and gosh, you know what, it hasn’t produced the result that we want.
Eric Enge: Right, how about some organic specific tips and tricks?
Ken McGaffin: Right, okay. I think my biggest tip on that is, always challenge your assumptions. I think that fundamental is really important. The other thing that I would think about in terms of keyword research is that your rankings for any particular term are obviously improved by your links. But, the whole drill is to have keyword rich inbound links from quality sites. Well, if you are launching a publication or a new product, you use keywords and the name of the publication of a product.
What you will find is if you when you do your PR release, then people will link to you using the product name as the linking text. If that includes your keywords, then you are going to get the extra and additional benefit in SEO terms. I think again that’s one of the things that people often miss. When I am doing a publication, I think how can I include a keyword phrase in the title of the publication?
Let me give you an example, we published a keyword research guide back in 2005. It includes one of our top keyword phrases from keyword research. The last time I looked in Google this morning, we were #1 for that phrase.
Part of the reason for that is we’ve got about 20,000 links, which refer to that publication and recommend it. So, hey presto, we have a whole a load of keyword rich inbound links. I would advise marketers to forget about clever names, just think about keyword rich names.
If I get keyword rich names, then you are going to give yourself such a benefit in terms of the inbound links that come there, then that could be a very, very useful thing to do. So, that would be my top advice.
Eric Enge: That’s great. Thanks Ken, I think it’s been a great discussion.
Ken McGaffin: Brilliant, thank you.