Prioritizing Keyword Research: Finding Needles in Haystacks

Today we’re going to take a deep dive into how to prioritize keywords.

There are no shortage of tools to generate keywords, and no shortage of metrics you can attach to them to evaluate their potential. But how do you turn your giant spreadsheet into something that enables you to take action?

How do you go from a sea of ideas from numerous tools to something you can turn into a roadmap and actionable to-do’s?

You need a process for prioritization.

Keyword Prioritization: Turning Spreadsheets into Action

keyword research process

Prioritization Criteria

There are a few ways to think about prioritizing and segmenting keywords.

They should all be based on your business and your goals, rather than some arbitrary score spit out of a tool.

Each search query has many unique qualities – some tools report dramatically different volumes, regional differences, and even something as sharp as 2-3x swings in conversion rate by simply going from a singular to a plural.

They’re different – and their audience intent is different – and they should be treated as such.

One way to think about keywords is to segment each of them across 3 different criteria:

  • Volume, or How Many People Search For It?
  • Buyer Intent, or How Close is That Searcher to a Commercial Action?
  • Competition, or How Many People Are Trying to Rank for It?

For your business, one of these might be more important than the others.

For example, if you run a celebrity gossip blog, you might think about volume a great deal, while buying intent might be unimportant. By contrast, if you run an ecommerce store, buyer intent might the most important criteria for you.

Every business is different and your mileage may vary.

So let’s take a deeper look into what these criteria might mean volume vs. intent vs. competitiveness.

Volume: Quantity has a Quality All Its Own

Volume is the clearest way to prioritize search terms – after all, who doesn’t want a lot of traffic?

The Google Keyword Planner provides a helpful volume estimator, which is also included in Term Explorer:keyword volumes in keyword planner The Term Explorer interface:

keyword volumes in term explorer

However, before you get excited, having found a way to swan dive into an ocean of cat video traffic, it’s important to understand a few big things about keyword research tools and volumes.

Volumes Are Often Seasonal

Most businesses have some sort of seasonality. In many categories the seasonality is obvious and straightforward – people often look for water slides in the summer, coats in the fall and winter, and flower delivery around Valentine’s day, and again at Mother’s Day.

But even B2B categories often have quarterly, yearly, and mid-month demand cycles.  And keyword research for blogs is a different beast all together.

It’s very possible that giant number of searches for a term are entirely seasonal, and you might find yourself waiting a great deal of time for those searches to come back.

You can use a tool like Google Trends to check out search volumes in regards to yearly demand and seasonal demand:

seasonality demonstrated

SEMRush also captures some of this information in their keyword report:

seasonality in SEMRush

Volumes are Often Local

Local differences in language and pronunciation filter through to search volumes, along with differences in demand. In some places, a truck is a lorry and a water fountain is a bubbler.

Understanding these local differences is key – you want to optimize for how your customers are searching – not how someone that’s out of your target market searches.

Or you might find yourself on the wrong side of a cultural misunderstanding.

crawdad geo search volume

crawfish geo search volume

It’s also worth noting that what is not a brand name in one place might very well be a brand name in another – and what Google sees as a local or transactional query in one area might not be one in another.

rodeo in austin - google thinks you want to go to a rodeo

rodeo in san francisco - people think you wanto know what a rodeo is

For example, Google serves dramatically different results in San Francisco, CA, and Austin, Texas, to searchers looking for “Rodeo.”

Make sure you understand the local behavior and implied intention in your SERP – if something is a local query in one geography and a general/event query in another, that should deeply affect your strategy.

Volume Can Change Over Time

At the risk of business cliche, follow Wayne Gretsky’s advice and skate to where the puck is going.

Search volume can change dramatically over time – not just seasonally, but simply disappearing:

declining interest in left shark

In addition to big, event-driven spikes, there are often larger, consumer behavior-driven search trends.

paleo v atkins - really the same thing

In this case, that blog about low carb diets you started might be better positioned for the future (and the New Year’s spike) by talking about paleo diet recipes instead of Atkins diet recipes.

Buying Intent

In many business models, traffic by itself isn’t that interesting.

Traffic that purchases things, subscribes to email lists, and represents the right sort of potential customers, by contrast, is exceptionally interesting.

To that end, you’ll want keywords that convert – not just drive traffic, and certainly not just the ones that come from of Google Keyword Planner.

Often the long tail phrases that don’t drive tons of volume still drive many interested new customers – especially on aggregate.

Buying intent is going to be different for every business – and ultimately, the relevance of a keyword to someone else’s transaction with your business is not that interesting to you.

You can monitor this using:

  • Existing Organic Search Landing Pages – using a tool like SEMRush, Authority Labs, or one of the enterprise-level SEO performance management platforms (Conductor, BrightEdge, seoClarity, etc.)
  • On-Site Search Data – If you have an internal search engine somewhere on your website, check out the conversion rates of different terms internally and get an idea of what converts and what doesn’t.
  • Paid Search Data – currently one of the more accurate sources of conversion data by query now that (not provided) rates approach 1, if you’re running extensive paid search campaigns, use that intelligence to understand how keyword searches turn into purchases.

You can also use external data – cost per click on paid search is often a good indicator in more robust verticals. As pricing is auction-based, for this method to work at least two of your competitors will need to know what they’re doing.

In some areas, I wouldn’t depend on this, while in other verticals, like travel, that tend to have multiple extremely savvy paid search advertisers, this can be pretty accurate.

Words and Buying Intent

One of the more interesting aspects of search marketing is often the direct relationship between the length of the query and level of buying intent (there are of course exceptions to this rule – “emergency locksmith” is a short query with lots of intent.)

This creates an interesting scenario – where there is often more revenue available optimizing for specific, low volume high intent searches, rather than broad keywords with many more searches per month.

Generally more specific modifiers signal more specific intent – for example, “Mens polo shirts” has nothing near the buying intent associated with “White mens lacoste shirt size 4 pique”.

You can also see this pattern with:

  • Plurals – ecommerce store owners often find the plural term possesses much more buying intent than the singular
  • Coupon or Discount Code terms (often these visitors are literally on a shopping cart checkout page)
  • Manufacturer product codes and SKUs
  • Product names with descriptive modifiers
  • Exceptionally specific searches with vertical modifiers (EG “email marketing software for florists”)

Competitiveness

In addition to volume and buyer intent, you have to look at competitiveness when you look at search keywords.

Will that juicy, high-volume, high-intent keyword take a year of dedicated work from a team? Or is it something you can get with your existing authority?

Often the best way to calculate these is through SERP analysis and looking at existing link metrics for rank potential – how many links would you need to rank for one of these terms?

How authoritative are the sites that rank for these queries today?

You can do this manually, using your favorite link data set, spreadsheet tool, and source of cheap labor, or you can let a tool like TermExplorer’s Keyword Analyzer handle the heavy lifting for you:

difficulty in term explorer

CPC can also be very helpful here – although it is not a strict measure of organic competitiveness, it can be a very helpful way of gauging search competition.

Prioritizing: What to Tackle First

Based on these three metrics – competitiveness, volume, and purchase intent, you can categorize and segment your search terms.

prioritizing keywords matrix

 

  • High Intent, High Volume, High Competition – the “Whales” –  These are the terms that have significant purchase intent and significant volume. Because of this, they have significant competition as well.

These will take a long time to rank for (and may be impacted by new Google ad units) – so while these should be a part of your plan, don’t exclusively hunt whales – after all, if your budget is cut or your client ends your engagement, it doesn’t matter how great your plan is.

Another point about these terms is that alternative search-driven marketing strategies than “build page and rank it in search engines” might be appropriate for these terms – can you buy a website that already ranks for these terms? Can you enter into an affiliate partnership with a website with this traffic? Can you do some digital PR and place an article that ranks for this sort of term on a high-authority publication? There are lots of ways to influence people searching for keywords and bring those searches back to your business.

  • Low-Medium Volume, High Intent, Medium – High Competition – the “Deer” These terms represent terms you can rank for with some work, and have some commercial value. These are good medium-term goals – and a great way to start building out a footprint you can continually expand.

Generally I try to think of keywords along these criteria:

  • What can we do in the near term that will help us show success and finance future investments in search visibility?
  • What’s the low-hanging fruit? What can we turn on in the next sprint and see an ROI from?
  • What can we do in the medium term – attacking more competitive terms – and build the machine that will enable us to pick terms, target them, and rank for them?
  • And what long term ‘whales’ can we take on eventually and aspire to get to?

Closing: Work Smarter, Not Harder

Effective prioritization is the key bridge between “a spreadsheet full of numbers and strings” and “an ever-increasing stream of traffic and revenue from organic search.”

By prioritizing your keywords through researching their volume, competitiveness, and purchase intent, you can find great opportunities and structure your work for maximum effectiveness.

6 Responses to “Prioritizing Keyword Research: Finding Needles in Haystacks

  • Thanks for the details. Very informative, and thanks for keeping it simple. I do believer the power and role of Keywords in Digital Marketing, and for sure, everyone should know how to play around the keywords around the objective of branding or keywords keeping in mind, intent, seasonality, competition, volume, local usage etc.,

  • Analytics provides tools that help you find the needle in a haystack more efficiently. Find Similar Documents and Keyword Expansion allow you to hone in on the unknown in your case documents without the need for additional review resources.

Trackbacks & Pings

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>