Friday, May 6, 2016

A Step-by-Step Process for Discovering and Prioritizing the Best Keywords - Whiteboard Friday

Posted by randfish

Keyword research, when done right, is a fairly complex process. Uncovering new keywords and appraising their value should involve a robust toolkit, a multitude of different sources, and a great deal of thoughtfulness.

In today's Whiteboard Friday, Rand shares a strategic and straightforward 4-step process (including a passel of tools to check out) for discovering and prioritizing the best keywords for your SEO campaigns.



Step by step process for discovering and prioritizing the best keywords whiteboard

Click on the whiteboard image above to open a high resolution version in a new tab!

Video Transcription

Howdy, Moz fans, and welcome to another edition of Whiteboard Friday. This week we're chatting about keyword research and a step-by-step process to choose and prioritize the best possible keywords that we can for your SEO campaigns.

So let's get started with...

Step one: Use multiple sources to get keyword suggestions

The first thing that a lot of folks do is they only use a single source. They go to AdWords for example, or maybe they'll go to Suggest. Or possibly they'll start with SEMrush, which has an awesome corpus and database, but it's sort of based on a single source. My strong suggestion is a lot of the sources have only one type of data in them and you want to combine them.

The five or six that I really like are, first off, AdWords is a great source. They're, generally speaking, commercially focused terms. AdWords knows that people want to buy those keywords for pay-per-click search, and so they try and include commercial terms that people are actually going to convert on. They hide a lot of stuff that frankly Google feels like is not going to get people the conversions they're looking for, because the problem is if you buy the wrong keywords, you don't blame yourself for poor keyword targeting, you blame Google for sending you bad traffic. So AdWords has hidden some of those things. They'll show them to you if you type them directly in, but not otherwise.

Suggest, you can go to Google Suggest and in fact, Google related searches - which are at the bottom of the search versus the top in the bar as you type - those both give variance and/or searches that people who search for this also performed.

Then you'll see there are a lot of tools out there. SEMrush is by far the most popular one - and, in my opinion, a really, really good one, too - for a keyword to rankings graph. Essentially what this is saying is, "Here are keywords that the pages that rank for the keyword you gave us also rank for," or same thing at the domain level. It's creating and mapping those things so that you can get broader terms than you ordinarily would have with just these other methods. That's pretty cool.

Another one that's very, very cool and very sophisticated, that some SEOs are doing, is topic modeling-based keywords. This is essentially saying, "Hey, show me terms and phrases that co-occur on lots of documents, high quality documents hopefully, with the term or phrase that I'm targeting." You can find those through tools like AlchemyAPI. It's a little challenging to use, but there you go.

Bunch of tools, SEMrush and AdWords. You can use Google Search for a bunch of these. Ubersuggest to get some of those suggestions. KeywordTool.io actually has a number of these inside it. SpyFu is similar to SEMrush. AlchemyAPI helps you with topic modeling.

Then - somewhat self-promotionally, and I apologize for that - but Keyword Explorer, which Moz just launched this week and which we're pretty excited about. I was actually the product architect for that. So I'm feeling quite excited and very proud of my team. Keyword Explorer, shamelessly, has all of these in there. I think our topic modeling is actually a little better than AlchemyAPI's. I think our keyword to rankings graph is almost as good, maybe in some cases better, maybe in some cases not as good as what SEMrush has. We also get suggestion-related, real time, and then we obviously have a big corpus that we've got from AdWords too.

Step two: Select keywords that match multiple types of search intent

I've done my keyword suggestions. I've got all of these. Now, I need to pick which keywords from these suggestion lists am I actually going to try targeting. To do that, it's not really a tool-based thing, although you certainly could use something like a Google Doc or Excel, or a Google Spreadsheet, or you could do it right inside some of these tools. KeywordTool.io and Keyword Explorer both have like an "Add this to my list" type of feature. AdWords does too.


But what I want to do is match multiple types of search or intent based on your content and keyword goals. So this has little to do with what the keywords actually say and more to do with, "What am I trying to accomplish with my SEO and with my content?"

For example, let's say I'm an online coffee bean roaster and seller. Maybe I'm independent. I have a location, but I also want to sell my beans and my grounds and accessories online, which is awesome. There are some keywords that are going to match with my goal of direct conversion. People are likely looking for this because they want to buy it, and I want to be in front of them when they're looking to buy it. Those are the keywords like "coffee beans online," "buy coffee beans," "coffee accessories," "stovetop espresso machine," getting more specific.

Then, I'm also looking at doing some strategic content to target folks early in the buying stage, like before they actually think, "Oh, I'm going to buy from them." I just want them to have an association with us. I want anyone who's interested in coffee - coffee aficionados, researchers, people who are passionate about the topic - to find me and have an association with my brand. In order to do that, I might target keywords like "flavors of coffee beans," "best independent coffee roasters in the US," "home barista resources."

People aren't going to convert on these keywords. "Best home coffee techniques." I'm trying to learn. I'm trying to get help. I'm trying to get content, not necessarily buy directly.

Then I might in my content strategy have some idea that, "Hey, I also want to target coffee influencers." People who are influential in this world. It could be journalists and bloggers, people who write for magazines, and folks who are very popular on Twitter or Facebook or have popular Instagram accounts. I want them to be aware of us.

So I might go after things like "barista competitions." Barista competitions, if I have a big list of those, well, lots of baristas and folks who run coffee shops are going to be looking for that. I can influence them, get in their head, get them to know my brand. "Coffee shop awards," same thing.

"Worldwide coffee bean wholesale cost comparison," aha, this is going to the suppliers and the coffee bean buyers around the world and looking at price trends and tracking. That's some of that data that a lot of those folks might have. Probably a small audience, but very influential people.

"How to open a new coffee shop," aha, now I'm targeting coffee entrepreneurs who are also potential influencers for me.

These lists won't apply to all of your efforts. Your efforts are going to be determined by your specific strategic goals, but you should make keyword lists and match those up to all the keywords you see here so that you know what types of things you're trying to do with those keywords. I would encourage you, if you're doing this, to make a different list for each one of those.

Step three: Collect keyword metrics and sort/filter/prioritize them based on goals

This is where we have to get very, very data-driven, because I want to do is I want to take all the keywords in each of the lists that I have and I want to get the metrics for them so I can prioritize properly. So what I've done here is I've taken a list of keywords. I have: my volume, how much are they searched for; difficulty, how hard will it be to rank in the organic results; click-through rate opportunity, meaning what other features are in the search results - images, news boxes, ads, videos up at the top, instant answers, knowledge graph on the right-hand side that's going to draw clicks away from my potential to get searchers to click on my result.

I need click-through rate opportunity in my scoring. Otherwise, I might be biased to keywords that look great but in fact get very little click-through rate.

Importance, this is essentially my personal priority, and it's something where it's not a metric that comes from anywhere else. I use it internally. I know, for example, that "coffee beans online" is a very, very important keyword because it directly relates to what I'm selling. It's the first thing I'm offering, so I'm going to put it at a 10 out of 10, versus maybe "how to open a coffee shop," which looked at some content marketing that I might do in the future, but it's not a high priority right now for me from an importance standpoint.

Then all of these metrics, so it's like this metric combined with this metric combined with this metric should give me some form of potential. I want to come up with an algorithm. You could come up with your own, or you could use one of the tools. Tools like KeywordTool.io and Keyword Explorer have an algorithm that combines these types of scores to give you a consistent one for potential. The idea is I want high volume, low difficulty, strong click-through rate opportunity, and high importance. That should give me a good potential score. Then, hopefully what I can do is just sort by this potential metric, and now I get my prioritized list of keywords.

If you don't take this data-driven approach, you can wind up just in a world of hurt where you're targeting the wrong keywords and not being as intelligent as you could be. You can do this with something like AdWords and then an export to Excel or to Google Spreadsheets. You can do this with a tool like WordStream, who does a great job of it particularly for paid search, and you can leverage some of that for organic too. Like I said, KeywordTool.io. Obviously, Excel and Google Spreadsheets. Then Keyword Explorer does this right inside the tool as well.

Step four: Determine keyword targeting & new content creation needs & priorities.

Now what I want to do is I want to determine my keyword targeting and my new content creation needs and the priority of those processes.


So after I look at this, I might refactor a few things and say, "Wow, you know what. That is pretty strong. Even though I set it as a low importance, I'm kind of interested. I'm more interested in this 'how to open a coffee shop' than I was previously, based on the metrics that I saw there, the opportunity I think I've got."

So here's my prioritized list.


  • (A) I'm going to start by optimizing my homepage for "coffee beans online." I've decided that's the best keyword that I can possibly target on there. That's what I'm going after.

  • (B) I want to create a new coffee accessories page. Maybe I didn't have one before. I see that that's a high opportunity and high potential keyword. I want it. I need to create a new page. Now, I also need to go get inventory relationships established with all my accessory providers so that I can actually ship folks that stuff.

  • (C) I've decided that I really like that "how to open a coffee shop," and I want to create a guide. That's going to be one of my key content marketing pieces and, therefore, I'm going to go interview 10 successful coffee entrepreneurs, folks who've opened some successful shops. Then I'm going to assemble some content, build a survey, target 500 coffee shops in the U.S. - maybe that I already have relationships with or that I don't - so that I can get a survey of data back. I'll outreach to each one individually, or I'll have my SEO content person do that. Now, I'm going to create that guide based on the feedback that I get from there. Now, it's data-driven, and I have a bunch of people who are likely to help support it because they've contributed to it.

  • (D) Finally, I might say, "Hey, I really like that 'best independent coffee roaster.' That keyword looks real strong to me. I want to target that one too." That's also going to go into my content marketing efforts, so I'm going to establish some criteria for that one. I'm going to do some research, and I'm going to send out awards to the winners after we pick those through whatever process we decide.

This is a phenomenal way to go through keyword research and keyword targeting to get the content and the optimization priorities that you need for SEO. I think if you choose the right data and the right tools, you use multiple sources, you intelligently build the right kinds of lists, you use metrics to prioritize, are data-driven rather than just pure intuition, and you prioritize your work based on this, you can have phenomenal success.

All right, everyone, look forward to your feedback and comments. Certainly, if you haven't given Keyword Explorer a spin yet, I'd encourage you to do so. I think it's pretty cool, but obviously, there are lots of good competitors out there too, and you can check them out as well.

Hope to see you again next week for another edition of Whiteboard Friday. Take care.

Video transcription by Speechpad.com


Sign up for The Moz Top 10, a semimonthly mailer updating you on the top ten hottest pieces of SEO news, tips, and rad links uncovered by the Moz team. Think of it as your exclusive digest of stuff you don't have time to hunt down but want to read!

No comments:

Post a Comment