Traditional keyword research has a tragedy of the commons problem. The more people that use a common keyword research sources, the less valuable those sources become.
When everyone is using Keyword Planner, Ahrefs, Semrush, Mangools KWfinder, and Google Autosuggest – the keywords from those sources will become either too difficult or will lose their intent. Big players, or Google itself, will show up and rank for those keywords.
As a small to mid-size publisher – you have to either become more creative, become more “long-tail,” or seek out more sources to find new keywords.
I’ve written before about “pre-qualifying content.” That process used different sources to confirm your own keyword ideas.
Finding new keywords from new sources is the reverse. It’s finding keywords that you have literally never heard of.
You would never know the root, modifier, or topic. It’s doesn’t show up in Google Suggest, and you’d never know how to search for it in traditional keyword tools.
Here’s where I’ve been looking lately.
Bing / DuckDuckGo Search
Bing and DuckDuckGo are usually seen as Google alternatives. But they have audiences in their own right.
And those audiences are different enough from Google that you can usually find new, interesting, and different keywords there that you won’t find on Google.
Head over to Bing / DuckDuckGo and try out their autosuggest with broad modifiers (how to, what is, etc).
Search within a broad topic that you are looking at and explore their related searches.
Lookalike Online Publishers
You might be familiar with the idea of a lookalike from Facebook Ads. They take a person that you want to target and try to find other people who share that person’s characteristics. It’s a powerful idea.
And it’s an idea that you can use to find keyword ideas.
Many SEOs know to look at the big industry publisher for keyword ideas, but never think to find “lookalikes” of their big industry publisher…in an adjacent industry.
And adjacent industries can be a goldmine for both keyword roots and modifiers.
For example, imagine that you own a small fishing website. You probably know the trick for mining Field & Stream for content ideas.
But what about looking at the Field & Stream equivalent in the backpacking, boating, or wildlife magazine segments?
If you had a Venn diagram, you’d see that they all have an audience overlap, even if they don’t have a keyword overlap. And that represents an opportunity.
Remember, the best keyword research understands the question behind the keyword. Explore those publishers to find content strategies and audience questions that you could use.
Look at their top-performing content and think about what you could swap in that would fit your website.
If Backpacker magazine did a profile on the top trails in the Southeast US, why wouldn’t a profile on the top rivers in the Southeast US work for your fishing website?
I would argue that it would. And best of all, any new keywords, headlines, or modifiers that you find function as pre-qualified content, since you know it has already worked for someone else.
I’ve written an entire post about using Wikipedia for SEO & Content Marketing. But Wikipedia is an especially good place to find keyword ideas because it is structured.
The process is to pick a category and systematically explore all the entries along with how they are related.
Read entries with an eye for keywords and concepts that need further exploration.
YouTube is a popular place for How To’s, tutorials, and visual content. You can (and should) use YouTube autosuggest to mine for keywords. They’ll have a different dataset from Google search.
But I recommend that you go a step further. Look at (or scrape!) top-ranking videos. Read their transcripts, tags, comments, and descriptions with an eye for new keywords and concepts.
If the video narrator uses a term that needs context or further explanation, note it. Look for how the narrator and video present information.
Even the best YouTube videos leave gaps that need to be filled. And those gaps usually produce Google Searches.
It’s a great way to find high-volume, long-tail keywords.
I wrote an extended guide to use YouTube for SEO & content here.
I would argue that Pinterest is one of the top search engines on the Internet, even though it is thought of as a social network.
Millions of people use it to start their search and bookmark their favorite answers.
While it does work better for some segments than others, I think it’s worth checking out for nearly every industry.
Like YouTube, you should look at both their autosuggest and their suggested topics. But go a step further and look at top-performing pins & boards.
Note any unique keyword angles, audience questions, etc. Usually, they have a completely different angle from Google that will provide a window into an audience’s true intent – which, again, will help you answer the “question behind the keyword.”
I wrote an entire guide to use Pinterest for SEO & Content.
Unlike Pinterest, Instagram is a pure-play social network. But it still throws off a lot of keyword data, especially for consumer industry segments.
Use the autosuggest on Instagram for modifiers, but also check out the Explore to find trending hashtags.
Take those hashtags and use them to understand trending topics, angles on existing keywords, and to find new, trending ideas.
Using Twitter search for keyword research is a bit like using Instagram, except that it’s heavily biased towards right now.
Twitter doesn’t have the same breadth that any other tool has. But Twitter’s speed, recency, and analytics can help you keep your content up to date and cutting edge.
I’ve written a full guide to use Quora for SEO & content. But the short version is that Quora has a bunch of experts answering specific questions with plain, human language.
You should mine both the topics and the keywords that contributors use in their answers.
As a bonus, if you sign up for Quora Ads, you can see the exact search interest that a question has.
You can use Reddit with a similar process as Quora. I’ve written a guide to use Reddit for SEO & Content here.
But the short version is to find one or several subreddits that your target audience is interested in. Sort by Top or Hot and start mining both titles and responses.
Everything is written in plain language, so you’ll find plenty of keywords that your audience uses, but that might not show up in a keyword research tool.
There are also plenty of automated Reddit research tools like Keyworddit.
Next to Wikipedia, Amazon probably has the largest repository of user-generated content on the Internet.
Their reviews are a goldmine for finding keywords that your audience uses while searching for products. In other words, they have the intent to purchase, which is critical for many publishers.
I’ve written a full guide to using Amazon for SEO & content here. I recommend starting with the Bestsellers in your category and then manually looking through the user reviews and their questions.
There are a few tools that can automate parts of this research, but it does not work as well as manually read reviews & questions.
Google Scholar Autosuggest
Google Scholar is one of Google’s most powerful, but least understood products. It searches the universe of scholarly journals, magazines, patents, and more.
In other words, it searches content that is actually rigorous and correct. It’s a huge opportunity to find academic jargon, theories, data, and more.
There is a universe of podcasts that cover every niche, market segment, and industry. But episodes are not indexed or analyzed in any systematic way.
Head over to any of the big podcast directories including Apple, Google, Spotify, and Stitcher. Look for podcasts in your industry. Look through top episodes, reviews, and descriptions. Listen to episodes that catch your eye.
Harvest & use any new keywords that you find.
Physical Books & Magazines
I cannot stress how underestimated physical books and magazines are for keyword research. They are structured, comprehensive, edited, fact-checked, and exist in every industry for every market segment.
The problem is that they are inaccessible for quick research. You have to, you know, read them? But that’s your competitive advantage over big publishers working at scale.
Buy some physical books and magazines specifically for keywords and topics that are not coming up in your traditional keyword research.
There are a couple of ways to speed this process up. One way is to use ebooks. Convert them to HTML or text. Search them quickly, or use an algorithm to parse them.
Another way is to use Google Books. I wrote a guide to use it for SEO & content research.
Google Surveys and Survey Monkey both make customer surveys much more accessible than ever in the past. They are still fairly costly but can be a good value when planning an expensive content or ad campaign.
One idea here is to ask open-ended questions and word association type questions to help trigger unique, qualitative keywords from real people.
Next Steps To Find New Keywords
There are a million variations of the cliche that “if you do what everyone else is doing, you’ll get what everyone else has.” Keyword research is no different. It takes work to find new keywords.
If all you do is go to Semrush or Ahrefs or KWFinder drop in a big keyword and sort by keyword difficulty…then you’ll never get ahead. Same with Google Autosuggest or Keyword Planner.
But searching out new sources to find keywords that your audience is using, but that you don’t see will help you get ahead.