Keyword Research for Startups – or One of the Two Things Every Content Marketer Should Do

Keyword research. These two words sound about as sexy as an empty milk carton in a windy car park. But, unlike said piece of packaging, keyword research is insanely useful. It  helps to identify big content marketing wins without relying on trial and error only.

And the good news is that on early stage startup scale it relatively simple and quick to do.

Keyword research is based on 2nd grade mathematics. The more monthly searches a keyword has, the better. The less competing blog posts and pages are about the same topic the better. And finally, the more relevant that keyword is to the service or product you want to promote, the better.

High number of monthly searches + low competition + high relevancy = content you should be creating.

Let me give you a Pipedrive example here. Which one of the following 4 keywords would you start producing content for?

Keyword volume

“customer relationship management” sounds like a huge opportunity, and “sales pipeline report” kind of pointless, but it might be a mistake to start by creating content for the former and dismissing the latter.

Here are the same keywords with some additional information:

Keyword volume full example

You’ll notice that “sales pipeline report” and “sales quotes” appear to be less difficult/ competitive terms. But more importantly, when you factor in relevancy to what you offer, the picture changes completely.

While “customer relationship management” is a popular search term, people who type it into Google have very different motivations. It probably includes business students doing some homework, secretaries trying to decipher whether the email they just received should be deleted or forwarded, my mom trying to understand what does the company I work for really do, and many others. Very few of them are CEO’s thinking “I should probably get a new sales tool so let me google the dry conceptual term for the thing my team does every day”.

On the other hand, if you’re searching for “sales pipeline report” and I’m offering sales pipeline software with built-in reporting (see what I did there?) you’re kind of looking for me. I really want to reach you and the 89 other monthly searchers that are putting up your hand.

And if you put the three things together, it becomes evident that more often than not you shouldn’t probably focus on chasing high-volume generic keywords but the less popular and more relevant ones. Long tail keywords, as they’re called.

This is especially true if you’re an early stage startup with a relatively new and content-light site. It’s very unlikely you’ll start ranking for popular keywords in your niche before you run out of money. For a fresh startup 30 qualified visitors in 1-2 months is much better than 1000 visitors in a year.

Humble beginnings of Pipedrive content and SEO efforts

With Pipedrive we started with a blank sheet, zero search traffic and, characteristically for a startup, no money. Here’s what Hubspot’s now shelved tool called Website Grader (its current incarnation is Marketing Grader) showed me:

website grader

This is not a screen shot cut off at the wrong place. We really did rank for no keywords at all.

I drooled over keywords like “crm software” or “best sales tools” but a few evenings of reading the Moz blog and other resources quickly taught me that going after them wouldn’t be the best possible start. I decided to tackle the very long tail term “sales pipeline management software” first and some weeks later we were on the first page of Google. Even more importantly we had people coming to the site and signing up. I didn’t know I was doing keyword research at the time, but it worked and so we moved on to trickier keywords.

Here’s the Website Grader report a couple of months later:

website grader 2 months later

Today Pipedrive ranks for many more than the 5 keywords shown above. Our small and growing team has put a lot of effort into producing and promoting content, finding linking opportunities and improving our landing pages to get there. Effort that has been pleasantly productive thanks to the keyword research we have done on a regular basis.

Guess that was enough background. So how do you do keyword research?

Step 1. Compile a mega list of search terms that are even remotely relevant to what you do

The first step is to map all keywords people may be searching for when it comes to your domain expertise. In Pipedrive’s case this is:

  1. keywords that directly describe what we do (eg. “sales crm” or “sales pipeline tool”)
  2. keywords which kind of describe what we do or describe part of what we do (eg. “sales dashboard” or “sales tracking”)
  3. keywords that people use when they have problem related to the thing you do (eg. “how to manage sales” or “how to define sales stages”)
  4. things only vaguely related to what you do (eg. “best sales books”)

Possible sources for compiling your keyword mega list:

  • keywords that are driving traffic to you, free and paid that you can see in Google Analytics.
  • keywords that are driving traffic to your competitors. I use SEMRush but SimilarWeb can also give you that.
  • search queries Google suggest and that you can see when you type in a word and a letter … and refrain from pressing Enter. For example in Pipedrive’s case “crm h…” gives this list

google autosuggest“Crm (for) healthcare” looks like a solid long tail keyword. On the other hand, “crm hobbies” are two words I’ve never seen next to each other, so it’s either a great or a terrible idea to write content for it.

  • keyword tools like keywordtool.io and Google Keyword Planner
  • words that people search once they’ve hit your site
  • words that people use when they write to your support, comment on your blog, etc.
  • last but not least, brainstorming

Step 2. Estimate search volume

Step 3. Estimate keyword difficulty

I’ve lumped in the next two steps because you can get this data from the same source, at least partially. And this resource is Google’s Keyword Planner.

Now, because Google’s tool only gives you difficulty in Adwords context, I also get keyword difficulty from Moz and calculate a weighed difficulty based on these two sources.

Please note you can export a neat cvs file for all keywords from Keyword Planner in the end that lists search volume as well as keyword difficulty as Adwords team sees it.

Step 4. Assess relevancy, calculate Total Opportunity Score

The final step is the easiest and also the most important. Using any data you have about keyword history and and common sense (invaluable resource!) you add an index of relevancy. You’ll then get something I call Total Opportunity Score for each keyword you have thought of at that stage. 

Here’s a basic Google Spreadsheet template I’m using to do keyword research. Please note that  it can and should include more data fields such as current ranking in search results and the same for your most important competitor(s).

How to apply keyword research

It’s worth pointing out that completing this keyword research exercise doesn’t give you the authoritative answer on what content you should create, and in which order. If your content resources are limited, you’ll want to sense check and manually pick the keywords and topics you want to double down on. This way you can make sure you have a healthy mix of  relevant long tail keywords that can give you an almost guaranteed result within a reasonable time frame and higher volume generic terms that take more time to start ranking for, but can potentially be big home runs.

And finally, keyword research is not a “set and forget” type of things. Rankings and search trends change, new competitors and technologies arrive all the time. I don’t know what the consensus is on how often this should be done but I’d say at least once per year.

Happy keyword research. I’ll follow this up with a post about the other thing/process every content marketer should do.

PS. We’re growing our marketing team at Pipedrive (looking for Head of Content/SEO, Director of Performance Marketing, a copywriter, Designer / Creative Director, multimedia/video producer, a “sales whisperer” and a few others) so if you’d like to talk about marketing with me more often, get in touch.

Follow the discussion on Reddit

7 comments

  1. MRPEasy

    Andrus,
    again very very good step by step survival manual.
    Thank you.
    Could you please fix a dead link for Google Spreadsheet template…
    Thanks!

  2. BorisDus

    Hey, Andrus. Nice explanations of basics! It’s a good example how to explain what is SEO and its tactics to people. Thanks a lot. Btw, a widget for sharing to Twitter has too much of symbols for sharing. =)

  3. Karolina

    Hi Andrus,

    thanks for the tutorial. Right now i’m trying to do everything like you say. I have already selected a bunch of keywords and i have several questions:
    1. What to do, if keyword planner doesn’t provide any of data for particular keyword?
    2. The step 4 is mysterious. How should i count the opportunity score, total opportunity score and competitor traffic?

    Thanks a lot for your answers.

    • Andrus

      Hey Karolina, sorry for the late reply here..

      If keyword planned doesn’t show any data it probably has very low search volume in that language / search engine – but it still may be part of content marketing planning if the phrase is super relevant to what you do.

      And indeed, step 4 sounds a bit cryptic. Basically it’s adding a 1-100% indication of relevancy and calculating total opportunity score as Search volume X Weighed Difficulty X Relevancy. Competitor traffic is available from tools like SEMrush.

  4. Jack Duncan

    Andrus,
    Another great tool you can use to find good keywords is the Bing Ad Intelligence plugin for Excel. It consistently delivers great keywords that so many other tools don’t mention, and it has four or five types of keyword search features to find even more related terms.

Post a comment

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