I used to think our book, Audience, Relevance and Search was the only book that emphasized basing your whole marketing and communications enterprise upon keyword data. That was until I read the second edition of Vanessa Fox’s Marketing in the Age of Google. Ms. Fox, who ran Google Webmaster Central before becoming an independent author and speaker, emphasizes using keyword data in every phase of the marketing, advertising, and communications lifecycle.
Here’s an excerpt:
[T]he term search engine optimization (SEO) is a bit of a misnomer because it implies optimizing your site for search engines when in reality you’re optimizing your site for your audience (who often searches) and ensuring your site is built to be successful in an online environment. (page 139)
While it’s great to find a kindred strategic spirit, I found myself wanting more tactical advice out of her book (outside of the excellent Google Webmaster Tools advice, which is to be expected from the founder and former editor of Webmaster Central). Mainly, I found it wanting in solutions to the problem of using keyword research across marketing and communications. Believe me, I scoured the book for what I was looking for and was somewhat disappointed not to find it. (Coming soon: a complete review at my other blog.)
In fairness to Fox, it was not in the scope of her project to get into the tactical challenges involved in collaborative keyword research. She repeatedly says she limits her scope to strategy: promoting the pervasive use of keyword research without delving into how to do it. This blog post will begin to fill this gap in her book. In particular, I will begin with the challenges involved in sharing keyword research and governance across multiple related websites. I’ll delve into challenges related other digital assets (press releases, advertising, social, etc.) in the end-to-end marketing and communications workflow in future posts.
In large enterprises, teams that create sites for individual brands will naturally seek many of the same words. Left to their own devices with nothing more than AdWords, they will find the words with the highest demand and start building pages around them. If no one is watching, the result over time is a lot of internal competition in Google. Google will rank the page it thinks is most relevant to those keywords from your domain, and it might not be the page that is most strategic to your company.
There are legitimate reasons to build multiple pages for the same keyword cloud. For example, users who just want to learn about some concepts related to your product categories have a different purpose than users intent on purchase. The trick is to use slightly different variations in the keywords for the two experiences. Users who want to learn about something use different keyword phrasing than those who are looking to purchase. To take a simple case, Learn user intents often are phrased in the form of questions. Purchase user intents often contain brand names.
In theory, it should be relatively straightforward to allocate keywords to related marketing efforts based on the kind of user intent modeling I describe in the previous paragraph. In practice, it’s not. The main cultural challenge is what I call the keyword land grab. Keyword research is a rare ability among brand marketers. Those who learn the skill first will build pages around keywords to stake out a claim on that keyword landscape. This first-come, first-served approach leaves brand marketers who are late to the land grab with slim pickings. Sometimes, they choose to compete, nudged by executives who want some of the action. More often than not, they don’t even know their efforts are doomed to fail because they are trying to compete with a page that already ranks from the same domain.
So how do you help competing efforts to share keywords in mutually beneficial ways? The most effective approach is to give the teams visibility to who owns what and which pages are performing well for the keywords in use. I’m talking about building an enterprise-wide keyword database that everyone has access to. Ideally, the tool will allow marketers to do keyword research more effectively than AdWords, discover which words are reserved and which words are available, and know that they will be most successful if they choose available words.
You will likely still need a board or council to break ties and settle disputes over the top keywords for your company. But a tool like the one I describe will reduce competition drastically over time. More importantly, it will help your organic and paid search activities capture more of the keyword landscape related to your products and services.
Natural Language Complexity
When I said it should be relatively straightforward to build an enterprise keyword database, I glossed over a big issue. User intent modeling is a developing science. There are clear-cut cases like those I mentioned. But there’s a vast middle ground between just beginning to learn about a product category and ultimately purchasing a particular model number. This middle ground is fuzzy. There is no one-to-one mapping of keyword phrases to user intents for certain audience segments. At best, you can make associations between audience segments, keyword phrases, and user intents. When I start to build out possible variations in these complex triples, my brain starts to hurt.
Besides calling in Watson to do the text analytics and take the load off our brains, there are things we can do to make sense of the complex set of user paths between learning and purchasing. The main thing is to focus on improvement rather than on trying to build the perfect database in one shot. The antidote to complexity is analytics. If you look at what keywords are performing well (i.e. converting) for which audiences and user intents, you can gain focus on parts of the picture. Performance analytics can help you tune your text analytics for those parts of the keyword database where you have good performance data. Those can be like settlements on the keyword landscape around which to build your marketing strategy over time.
Building an enterprise keyword database takes time. But if you are committed to the proposition that the vast majority of your customers use search to purchase your products and services, it is necessary. Vanessa Fox made the case for basing your whole marketing strategy on search data. If you are not convinced of this proposition, I urge you to read her book. It is chock full of facts and figures that collectively make a great case for putting search at the center of your marketing strategy. The keystone of that strategy is an enterprise keyword database.