This post was prompted by many discussions with clients in which we attempt to summarize SEO and PPC performance in practical, easily digestible forms. There is frequently a tendency for clients to oversimplify evaluations to “it works” or “it doesn’t” based upon summary data.
For SEO, we try to summarize performance in terms of increases in traffic and away from tracking only a handful of keywords. For PPC, we provide average results from multiple keywords, ad groups, and even campaigns so that we don’t overwhelm our audience with vast details and get into prolonged discussions of SEM minutiae. This summarization presents a problem: it paints performance with a very broad brush, one that may hide important details and diminish the significant contributions of different parts of advertising and SEO efforts. To mitigate this, we have begun to describe keywords in a different way that helps us better communicate to clients the intricacies of performance and reporting, and it avoids a common tendency to look at success in a binary fashion.
Separability of Keywords
Keywords for your product or service are separable, meaning that the users of one keyword are basically unique and separate from the group of people using every other keyword. This is pretty obvious for very different keywords: we would not expect the people searching for “used cars” to be the same as those looking for “new Honda Accords.” It’s also usually true for very similar queries, like groups of people searching for “Seattle restaurants” and “restaurants in Seattle.” These groups are still separable because someone who searches for one keyword variation is very likely to do one of the following: find what they’re looking for (and thus not need to search again), try a substantially different keyword if the first’s results weren’t helpful, or they will find your site and will be unlikely to return to it again if they felt the need to search again on a different keyword (because, presumably, changing their search means your site wasn’t what they were looking for).
Separability means that each keyword represent a different group of people.
Keywords as Markets
If each keyword represents a different group of people, then we should think of each keyword as an opportunity to reach different markets. Each market has a size, which is the number of people searching for that keyword. Each market responds best to ad language that is typically very similar to that keyword. Each market may have a different goal or product of interest (e.g. used versus new car buyers). Each market may be most responsive to different calls-to-action and conversion paths.
Target Market Expansion
With the groups of users being different for each keyword, these markets are additive: you can expand your target market by going after additional keywords. If you’re only advertising on or optimizing for “Seattle restaurants”, you are unlikely to be reaching a lot of people who are searching for “restaurants in Seattle”. Even if you do target both options by the “blunt-instrument” approach of using broad match in PPC advertising, you won’t be able to differentiate between the two to take advantage of differences in receptivity to ad language and effectiveness of landing pages.
Now we get back to where we started: reporting. If we have educated our clients about the market-nature of keywords, it is much easier for them to understand how the overall performance is truly an average that can incorporate widely divergent parts, ones that may have very different CTRs, CPCs, conversion rates, and ultimately degrees of success. This makes it much easier to describe relative degrees of success for the different parts of PPC and SEO efforts, and then to further pursue those markets that are working or allocate resources to improve those that don’t. For PPC advertising, this may mean moving budget from under-performing keywords to those that do well, or adjusting bids down on underperformers until they are profitable. For SEO, it places more emphasis on all search engine traffic, including “long tail” search terms that are often under-appreciated with a focus on rankings.