Change Your Perspective on Keyword Research, Win Big
We’re gearing up for SES SF and chatting with some of the speakers who’ll be presenting next week. I caught up with Bill Hunt, president of Back Azimuth Consulting, international speaker and co-author of “Search Engine Marketing, Inc.: Driving Traffic to Your Company’s Web Site” to talk about keyword research.
In this short, three-question interview, Bill gives a ton of insight on how to grab hold of all those missed opportunities in your keyword set. He reveals that keyword research is not about volume of keywords or even volume of traffic for those terms (at least not right away); it’s about prioritizing and maximizing opportunities you already have.
Bill also talks about successful keyword research in a world of Google Instant and Bing’s intent-based results – something most companies aren’t even thinking about yet.
Jessica Lee: Do you have a quick-and-dirty methodology for keyword research you can share?
Bill Hunt: Yes. Answer a few questions about your business and target audience and focus on where you make your money, not getting immediately caught up in the search volumes.
Step 1: Ask, where do we make the most money? I just worked with a small retailer that was trying to target words for more than 25,000 products in their database. We sorted the products by profit margin, then expanded the relevant terms for that product.
Once we had the full cluster or terms, we looked at search volume across the purchase cycle. From there, they just worked their way down the list. This took the pressure off to deal with that scale, and allowed them to focus content and budgets in paid and organic around those words with greatest overall opportunity.
Step 2: Identify your purchase cycle descriptor terms. Once you have done a few of Step 1, you will see the pattern, the obvious “buy,” “purchase” and “discount” can be appended to all of your products. Surprisingly, these terms often have low keyword volume and are sometimes ignored in a traditional research.
Once you have a large basket of terms from these steps as seed words, you can expand them using Google’s Keyword Suggestion tool or any of the commercial tools available.
What’s your advice to those who are new to keyword research?
Follow the method above and don’t get caught up in all of the process that most people suggest. I think too many people starting out with research cast too wide of a net and miss a lot of the logical opportunities. They just start dumping terms into tools hoping for miracles and silver bullets.
For people who are new, I suggest they ask customers what they searched for when they call your customer service line. Most ask, “How did you find us?” but then also ask, “Who did you search?” and this often find some pretty interesting results.
Second, review the keywords that visitors used when they searched on your site. This tells you what they wanted once they were already on your site. It might be a product or service that they think you offer, but they cannot locate it. If you don’t offer it, this demand may help justify the need to add it.
Most commonly, a large quantity of site search activity for a specific phrase means your site navigation is not working. If they have to search for something on the site, it typically means they can’t find what they want in the header or other site navigation. While not directly a SEO function, it does help with usability.
Both asking a visitor how they found you and what they look for on your site are key signals that can help identify new products and navigational challenges, but most importantly, help you understand brand loyalty and affinity since they came to or are on your site looking for something specific.
My big warning for those performing keyword research is going crazy with keyword expansion and finding “more words” when you have not maxed out the ones you have. I just had a conversation with a company that wanted to hire me to help them expand their words. They had an Excel file of nearly 80,000 terms and felt they needed more to increase their revenue.
After talking to them, they realized it was not more words they needed, but to get greater yield from the ones they have. I am a big fan of putting words into importance tiers. If Tier 1 is your highest, you should have no more than 100 words for an average site and no more than 500 on a large site.
These Tier 1 words are your mission critical words – the essence of what you do and your business goals. You want to make sure you have 80 percent share of voice for these words in paid and organic campaigns. If you are a market leader with a greater than 80% share of voice with decent creative, your search efforts will be off the charts.
How has keyword research evolved since Google Instant and Bing intent-based results?
Unfortunately, for many companies, it has not evolved – which is why for the first time we are seeing more clicks on paid search than organic. I like to point people to the simple query of “cloud computing” – it’s a very expensive word. If you look at the context of the term and the variations of “cloud computing,” you will see that 87 percent of the related searches are for some form of “what is cloud computing”.
Google understands this, which is why all 10 of the organic results are related to “what is it” and not “solutions” as most of the paid listings represent. A few of the companies ranking well like IBM and Rackspace understand this, and have tuned their content from their solution to an education to match the intent of the searcher resulting in top positions.
Companies need to understand the full cluster of words around a topic, and tune their messages to relate to them. We need to understand any “qualifiers” people add to the query to filter results. The days of non-relevant broad matching and getting any page to rank well are over. We need to start thinking about what is the best content, and offer content that matches the interest and context of the searcher.
For the cloud service providers, people looking for the topic don’t yet understand the value of your sophisticated hybrid cloud offer; we need to think about how we can more effectively attract and engage, then retain those people looking for the solutions you offer at whatever phase of need and understanding they are in.
This is also becoming very important with key commercial intent searches. For example, this past weekend, I was looking for a new wine cooler. I wanted a “50 bottle duel zone wine cooler” and only one organic listing out of 10 (for Costco of all places) matched, and none of the paid matched. In fact, none of the paid results – except one – gave mention of the number of bottles.
This is the major problem with how we do keyword research. Google did not show any demand for this query, so I would bet that few people would add it to their keyword list. However, if you look at how people buy wine coolers, it is nearly always by either dimensions (to fit in a spot in the home) or by bottle count, followed by additional features such as duel zone, noise and color.
In my upcoming session at SES, I’m talking about not only how to find more words, but how companies can successfully mine the data they have to increase their overall performance yield.
If you’re headed to SES, don’t miss Bill’s presentation on Keyword Modeling Analysis on August 14 at 11:45 a.m. You can follow him on Twitter @BillHunt or on his blog WHunt.com. To learn more about his consulting company, check out Back-Azimuth.com.
3 Replies to “Change Your Perspective on Keyword Research, Win Big”
Hi, thanks for the article. This is a question for Bill, but I wonder if in the Wine Cooler example there the problem was not a lack of existing pages matching the query but Google’s omission of those pages from the first page of results.
Fwiw, doing the same search about 5 mins ago yields more matches than Bill points out, so I think he may be disregarding many synonyms that match the query just fine. Even looking at the screenshot there I can see at least one example where “refrigerator” and “dual temperature zone” were not highlighted as matching his request but are obviously the same or functionally similar products. Whether or not shoppers ignore certain synonyms or not is another discussion, but they seem to be available.
Not to take anything away from the article at all, as I agree a lot with his point about simplifying the process and using some basic logic and “real-world” research.
In my experience you can’t have to many term/keywords. As long as you are able to overview them and the handling cost don’t get to big.
To have success with only a few keywords you most be extremely good at convincing.
Thanks Jessica, learn a new thing from this blog post “cloud computing”. When a lot of people search for what is cloud computing, we should provide the content with the right keywords to rank high.