English Español Deutsch Français Italiano Português (Brasil) Русский 中文 日本語

SEMrush Toolbox #11: Topic Research tool




Craig: Hi, and welcome to today's SEMrush webinar. We are back for another SEMrush Toolbox, and today we're going to be talking about the Topic Research tool.

Today, I'm going to be joined by A.J. Ghergich. Hopefully, I pronounced that properly A.J.?

A.J.: Nailed it. Nailed it.

Craig: Thanks for coming on today. Just before we start, what we'll do is just spend a couple of minutes just letting you introduce yourself to the audience, for anyone who doesn't know of you. Can you tell me A.J., just a little bit about what you do, and your background, and stuff like that?

A.J.: Yeah. Long time SEO, been in that game since the early 2000s. I have been a serial entrepreneur during that time, and recently for the last five years, I owned a boutique SEO and constant marketing agency. 

We grew from a 20 person team to 150 person team in the last few months. Right now, I am president of a digital intelligence practice group at BRADO, which is a brain growth and innovation firm focused primarily on health and life sciences, but we have a lot of consumer-based clients as well, like Verizon, Google, Salesforce, Anheuser-Busch, beer and dog food, always pushing that.

Craig: How was the jump, out of curiosity, from the small amount of staff right up to, was it 120 you said?

A.J.: The integration is tough. It takes a lot of work, it takes a lot of trust. I wouldn't have brought my team over if I didn't trust the group, that we bring it together. What we brought together was about four to five different practice groups, where each group was really good at one thing. 

My group is coming in to do SEO content marketing, that's all we did. It's not as much of a change. 

Craig: It seems a sensible way. When you said it initially, I thought you had maybe hired all these staff all by yourself, and stuff like that. It makes sense to grab people who are good at a specific area and pull them together.

SEMrush Topic Research Tool Overview

We'll start the webinar. I'm just going to start to show my screen, and I'm going to show you guys the tool initially. As normal, I'm going to show you the tool, where to find it, and some of the brief things that the tool does. Then we will have A.J. go into more detail in how he uses the tool on a day-to-day basis.

Today's choice is the Topic Research tool. When you log into SEMrush, you'll have your navigation down the left-hand side. If we go to under Gap Analysis, you'll see Topic Research. If you click on that, it will take you to the Topic Research tool.

The Topic Research tool, we all know content is the biggest thing and something that most websites will always struggle for. It always has been over the years something that, just to rack your brains; sometimes you forget things. The Topic Research tool basically pulls these ideas together for you.

The tool checks out your rivals’ content and gives you ideas on what to write. Now you can either type in a topic here; just search a topic. Or, you can put a domain name to find content on. As always with most of the SEMrush tools, you have to select the correct database. 

I'm going to be different from A.J. because he's based in the US, I'm based in the UK. We can basically go right down into the city that I'm in, which is Glasgow. You can go right specific into that. Make sure you get your settings correct and foremost.

Now the topic that I've done previously, and the only reason I've done it previously is the tool takes a little bit of time to work in some instances, is used cars. When you type in your topic, just type in used cars, and then click on the green button there, Get Content Ideas.

The next screen is basically going to land you on a nice overview page. The great part about this is going to give you the top ten headlines, based on your competitors, and it's also going to give you 10 interesting questions, and obviously, it's about used cars. There are loads of different questions there that are going to be relevant to used cars.

Over on the right-hand side, you can prioritize the topics by volume, difficulty, or topic efficiency. Volume for me is always going to be the one that I'm most interested in, purely because if there's volume there, there’s scope to get more traffic. 

You can see here, used cars NI, which stands for Northern Ireland, has 201000 searches. If we click on that, down below it's going to give us headlines, and some questions based around those topics. It'll tell you the difficulty score, so we can zoom in here, and you'll be able to see that the score for all these kind of searches is 72.01% out of 100, and topic efficiency is high. 

That is very relevant for someone who generically sells used cars in the UK, as long as you service Northern Ireland. The cards are going to show you a wide range of different options.

You also again the questions. You've got, what is? Why? Which? Should? How, and would? Which are all questions that people are typing into google, so these are good questions that you can blog about. 

You have related searches here as well, which you can click on, and it'll pull up the data for that. As you scroll through all the cards, it's going to give you different ideas. You click on any card that you want, and it will pull up all of the headlines and the questions.

Now on the right-hand side here, if I zoom in, you can add this to your favorites. If that's a keyword that you really want, and you can go through here and tick as many different ones as you want, and maybe some questions as well. That's cards explorer. 

Explorer is going to give you sub-topics, and it's going to give you a bunch of keywords, the content ideas. It's also going to give you an idea in terms of Facebook engagements, backlinks, and total shares that these topics have had as well.

Something there that you can grasp from that is, if something's shared a lot on Facebook then it's likely to be popular. Someone might have just done massive amounts of paid social on it, but I think engagement is something that you should think about while looking at this data.

Click on Overview, and it will give you an overview of everything: “Top 10 headlines by backlinks”, and again you can add some of this stuff to your favorites. 

Up at the top here, the big green button, you can export this stuff to a spreadsheet if you like. Now, what I'm going to do is click on this little tab here, which is Favorite Ideas. When you're doing your research, you may click on a wide range of different cards, and you may only want to pick one or two ideas from each of those cards. 

Having these things, marking them as favorites; it's going to have it all listed for you so that you don't have to go back through or write it down with a pen or whatever. It's quite good that you can do that.

There are other tabs here, so you can also use this tool to create an SEO template. You can also schedule a content task. You can also send this information to Trello, so if you wanted to have a processing place where someone has done the topic research and sent it on to Trello to get your content writer to do that content, then that's also a good option. 

The Topic Research tool is great because SEMrush's data is constantly improving. This is just laying it all out in a plate for you. You are always going to be stuck for ideas or looking for new ideas. Your mind may go blank, or you just want to get data that your competitors are possibly using, whether it's questions or headlines, or whatever it may be. That is essentially what this tool is doing for you. 

That is pretty much what the tool does. I'm sure A.J. will go into the tool and how he uses it, in much more detail. What I'm going to do is stop sharing my screen and let A.J. come in.

A.J.: I'm going to show you a framework that I conduct marketing from within, and how well this SEMrush tool fits within that framework.

Google as a Database of Human Intentions

If you remember only one thing that I say, remember this: Google was not a search engine. What Google actually is, is an artifact. I would posit that Google is not only an artifact, but it's also the most important artifact that humankind has ever created. The reason it's an artifact and the most important human artifact is because Google is actually a database of all human intentions.


Google has every single question that humanity has asked, on a global scale, for 20 years. They have every answer returned, for 20 years, on a global scale. Then they have every path taken, so every action we took. They know what you read, you didn't read, what you clicked on, what you clicked on two weeks after you read an article that sparked an interest. 

If you look at it again, every question, that's all of our thoughts, wants, and desires. Then they have all of our knowledge, every answer returned. I refer to Google as the database of human intentions. I'll show in a second how that subtle change can interact, or drive your marketing focus.


When you ask Google a question, Google uploads you to the database of human intentions. Small, simple transaction; you are telling Google your questions, which means you are telling them your thoughts, wants and desires, and every time you do that you are adding to the database of human intentions. You become more, and more a part of that ultimate database of human intentions.

That combined with this, makes Google a very interesting target for us in its marketplace. This concept is also key. Google is one of the last judgment-free zones in the world, and here I'll give you three quick examples of how that is accurate. 

Go right now, if you're still on Facebook, one of those people, post on Facebook who you're voting for in the next election. I dare you. It doesn't matter what country you're in, someone's running for something. Go ahead and post on there to your friends, who you're going to vote for in the next election. 

Guess what you're going to get from your friends: judgment. You're not going to do that. Walk up to a coworker ask them how much salary they make, ask how much they paid for their house. You're not going to do that because what will you get? Judgment.


Here's the thing, you don't have to do any of those things because you can just Google them. You will go to Google and tell Google all your political readings. You will bear your soul to Google in a way that you wouldn't do to your friends. 

All this is at your fingertips, and you don't need any judgment, Google's not going to judge you for anything you ask it. Think of anywhere else in your life where that is true. Ask your spouse some of these questions, you're going to get judgment. Google is an artifact, and it's the last judgment-free zone that many of us interact with on a daily basis, on a global scale.

Create Your Own Database of Customer Intentions

Once we realize that Google has a database of human intentions, we go about creating a database of customer intentions. A little DIY version of what Google has, by mining and using tools like SEMrush, to mine out those questions that people are typing into Google, to get to what intentions each of those customers have. A little slice of Google, just specifically for your customer base. There are two steps in this that I'm going to cover today.

The first is simple, we're going to mine search data. We're going to mine data from Google, Bing, Quora, we're going to use tools like SEMrush to get to those questions, just like we saw in the Topic Research tool. The data that we can gather is pretty staggering how quickly we can get it, and how insightful it is. 

Here is that Mind Map tool from the Topic Research section in SEMrush. I've put in Bariatric surgery, so you can see this is just a visualization. If you're an Excel person, that's great, just export it to excel and you'll be fine. If you're a visual learner, this is a good way to click around and explore. 

What I'm really interested in, in this tool, is the questions that people have around Bariatric surgery, which is weight loss. I want to mine out, and export all of these questions that are literally at my fingertips with SEMrush. We'll go through why we're doing that in just a second, but I also want to give you some additional ways to add to what SEMrush is giving you.

There is a chrome plugin, it's called Scrap Similar, there is a bit.ly link that I am showing on the screen, we can maybe pop it into the comments, that will take you right to it. It's just a free plugin. Install it, and you can scrape Google, and really actually any site. I scrape Quora a lot, but I'll show you how that works right now.


These are People Also Ask boxes, I know most of you are familiar with them, just ask a question on Google, and you'll see them. Some of these go 200 deep but expand them like 20 times, right-click, and then click Scrape Similar, as you can see me doing on the screen. 

There's literally an export button that sends that to Google Docs. You can hit export to Google Docs, or Excel, and you have this raw data, which is the easiest way I've found to grab PAA's with the questions. You combine that with what SEMrush has, and now you start to build up again a very good start to a database of your customer intentions, as you do this for each of the keywords that are looking to write for.

Mining Google for “People Also Ask” Suggestions

Also, you want to mine suggestions from Google. When I say that Google is a trusted space, a safe space, where it's judgment-free. Look at what happens when you type in, bariatric surgery into Google. That's weight loss surgery again, and these are things that people will say to Doctor Google, but they won't say to their primary care physician. 

Can I drink alcohol and get pregnant, chew gum, smoke weed and take ibuprofen? Hopefully not in that order, but it's pretty interesting what people will say to Doctor Google and not their physician. We could get all these suggestions, SEMrush has access to this as well, and you can mine them with the PAA's and combine that with the topic research questions, and even the Keyword Magic tool at SEMrush. 

You can build up a very powerful and efficient database of all the questions that your customers are asking about everything that you offer, and even your competitors. What is on your customer's mind, that is unspoken? I guarantee you they're speaking it to Google, and you could access that information with tools like SEMrush.

What we do is, we take these questions, and we use natural language processing to group them, and some visualization tools to create interactive graphics. You don't have to do this, I'm just showing you how you might take some of these topics and group them.


What I'm showing you here though is how to group things, and I'll show you a way that you can do that easily, but you want to group all of these intentions, all of these questions into subtopics so that you can start to make sense of them. Then you can start to think of them even like web architecture.

In search, we always talk about intent. What is the intent of that query? I think that is a phenomenal thing. If a lot of people stopped to understand the intent of each query they try to aim for, they could save themselves a lot of hassle, and that hassle being wasting their time on things that just don't mask the intent of the query. Intent is good, but I think focusing on outcomes is better. 

Once you understand the outcome your customer needs and desires to complete it, it completely changes how you're going to go about marketing to them.

For instance, let's say I have an outcome statement that I want to reduce anxiety. That's my outcome. "I want to reduce my anxiety. How will I go about that?" That can be different ways. I may go for a run, I may drink a glass of wine, I may take a pill. The outcome remains the same, people focus way too much on the product, on the horse getting you there faster, when the outcome was just to move there faster.

Keep that in the back of your head as we look through this, and the reason why those questions are amazing is, your customer is literally telling you the outcome they desire by the questions that they ask Google. They are signaling that intent, and they're signaling the outcome they need. If you address that better than your marketplace, you will win because it's not the product, it's the outcome.

Grouping Customer Intentions

I'll move onto step two, which is to take that database that you've created and to activate it into a customer-intent engine that powers your big campaign. I use natural language processing. We use Google Cloud to compartmentalize or deconstruct every single one of those questions. 

This is a question that I was looking at for Zirtek, the allergy medicine. We got all the questions people were asking about spring allergies, and then we used natural language processing to break those down. This costs tenths of a penny, and you don't have to be that tech-savvy, you can just go to Google cloud right now, and you can do a free demo of it.

I'm going to give you a second option. This is a tool called Online Ngram Analyser. If you look on the right-hand side in the yellow, there is another bit.ly link, that bit.ly link will take you right to this tool, it's just a read or write. 

In this case, we were looking at gut health, so that's a big topic right now that your gut biome is the gateway to your overall health. I used SEMrush, a combination of the Topic Research tool and Keyword Magic tool. We put in a few thousand questions on good and bad bacteria in the gut, and instantly you press the button, and this will give you the Ngrams, which think of this as the root phrase, or keyword inside of the question. 

This is a dead simple way to remember that wheel I showed you, where I had grouped everything of natural language processing. There are 35 questions on killing gut bacteria, good and bad. You could start to group those so that when you go to create content, or you go to create your website architecture, it can mirror the groups that your customer is telling you that they have questions about. 

Then address those intentions, and address those outcomes that they're actually looking for. That's a simple way of gathering that information, and grouping it without having to have a data scientist on hand.

Another way inside of the Topic Research tool is the explorer tab. I love this for one simple reason. If you look once you get all your questions and your URLs in there, look at the subtopics. It basically is doing a version of that for you. It is grouping them up by a root topic, or concept. You could export this right to excel.

Whether you created an excel, or you do some fancy NLP, doesn't matter, or you just use that Ngram analyzer and SEMrush combination. You should have a large collection of your customers’ questions, and they should be grouped by a theme or a subtopic. 

Base Your Website Architecture and Content on Answering Customer Questions

Once you have that you can compare the current state of your website and architecture to the database of intentions, customer intentions that you've created, which will lead to this. This is the website architecture your customers needs you to become.

You have your current state, you collected all of the questions, and you grouped them by subtopic. Then you say, "Here are our gaps, here's where we have a leak." We don't even answer any of these questions. "How should I lay out the architecture?" Good news, your customer already told you how to lay out the website architecture. 

"How should we write the piece of content that goes under each of those little content hubs?" Good news; we've already collected all the questions, those are your subheaders. Answer those questions, you'll be rewarded by Google. 

This is a simple framework of using these types of tools to power everything from SEO, from technical SEO to content marketing, even paid search.

Mine questions with SEMrush and Google, you can do that. You can either use the Google cloud to group them, you can spend a hell of a lot of time in Excel; if you're good with that to group them, or you can use Ngrams, or you could use the SEMrush explorer section to group them, but you want to group them up. 

Then you want to use that to drive your content creation, all the way from your architecture to the individual hard goals. Once you do that, you are going to harvest a ton of free traffic from Google.

When you build a database of customer intentions, it has all the questions that your customers need to know. Guess who has that database? You and Google. Google is only looking for one thing, for people to have a good user experience when they use Google. 

Where do you think they want to send them? To documents that closely mirror the questions and needs that they have. This will fill your pipeline with grateful customers who know that you get them. You understand the outcome that they're actually trying to achieve, which will lead to happier customer outcomes because you are attracting the right customer because you understood and addressed their needs.


As I said in the beginning, Google is not a search engine, it's an archive, but it's also a doorway. Google is a doorway into every single one of your customers’ unfiltered thoughts, needs, wants and desires. All you have to do is step inside. 

Craig: Connie is saying that they're considering using the topic tool to replace BuzzSumo, would you recommend that move?

A.J.: Look, the political way for me to answer this is, if budgetary concerns are an issue, and many times they are, then consolidation is your answer. If you can get a similar result from one toolset then do it. If you have the budget... I have SEMrush and I have BuzzSumo, so I'm not going to say that I don't use them all. 

If you have the budget, I think any SEO will tell you there are certain times, like for instance, sometimes I'll use screaming frog, sometimes I'll use Sitebulb, and sometimes I'll put it into an audit in SEMrush. I have access to all three, but it's a budgetary question. Yes, you can get very similar results with SEMrush, like BuzzSumo, and I think that you will get by pretty well.

How to Map Your Website

Craig: The next question we've got is..."How do you quickly map the current state of your website, is it simply just pulling your current sitemap?"

A.J.: Yeah, exactly. Taking the sitemap along with all copy, so, however, you want to do that. You can use SEMrush for that, you can use Screaming Frog, whatever you're going to use. Just creating that, and pulling it into a database. Then looking at it, look at your architecture.

Take your architecture and delete it from your mind. Then get every question your customer asked, do the Ngrams, and just look at those little root phrases like lose weight, side effects, blah, blah. You can start to see how the hub architecture of your website is just written by your customer. 

They're telling you the exact hub architecture they want, and then as you drill down, if you go into one of those side effects, well now you know here's the 50 questions they have. Well is that going to be three pieces of content, or one long form because a lot of the questions are duplicated? Your customer can lead you to the architecture that they need, and it'll mirror what Google knows that people want because it's completely customer driven.

Craig: Next question we have is from Jules Web, "Does what you're explaining translate to creating a blog post for each subtopic, and a parent post that outlines, and links to all the subtopic blog posts?

A.J.: That's a great way of doing it. Yes. This could be done in blog architecture, it could be doing in landing page architecture, like a hub. Yeah, that's a great way of doing it, and yes you don't want to just make things in silos, you want to interconnect them. 

Once you know that, "Hey, I have the topic of weight loss," and then I have my subgroups for if you're doing surgery. Well, there's going to be very clear subgroups. Once you see all the questions you know the interconnectivity. Once you know someone's question you can pretty much predict with high accuracy the next five things they want to know. Well, that's how you're going to connect that into a hub.

Finding Customer Questions for Esoteric or Vague Topics

Craig: Jock's asking, "What if the topic is very vague, and there are no unique crisp questions that exist on the web? Where would you go from there?"

A.J.: Well I would challenge that there are no questions. I've never met a niche that didn't have questions, I just made one of these databases for a company that their only product is bone cement. A thing that I didn't know existed, but that's when you have, say a surgery to repair a bone or something, that's like that cement that they put into your joint. That's all they do, bone cement.

We were able to find a lot of data, around 200 questions that surgeons had around bone cement, and the applications of it. What causes infections, what the proper use is, and certain surgical procedures. They were there, we just had to dig. 

If you want to post in the comments maybe what subject matter you're looking for, that could be good. Also, if you're having difficulty go over to the keyword magic tool in SEMrush, I think is amazing at questions. Put in the magic keyword tool, put your topic in there and I'll be shocked if you don't find a lot of questions and you can add multiples, and multiples are just created and exported to excel. It's a great tool for questions as well.

Craig: Excellent. Time for one last question... are there any case studies that you published that outline your framework, and how they have worked for clients?

A.J.: Well, that is going to be coming soon. As I mentioned earlier, we are a combination of about five different practice groups, so we've just formed over the last year our new agency, our new website is coming out in a few weeks. Very soon we will have case studies. 

If you follow me on Twitter @SEO, I will be posting a lot of those. I'm going to be posting a lot more on that framework specifically, and outcome-driven marketing, focusing in on the outcomes that your customers are trying to achieve and working backward from there.

Craig: Excellent. Excellent. That is us sadly out of time, but for anyone who wants to ask A.J. any questions, his twitter handler is @SEO, believe it or not. It's outrageous that he has that, pretty jealous of that. Yeah, you can get A.J. there, and we will have his slides under the video later on. I'm sure you can shoot him up on Twitter if you've got anything to ask but thank you A.J.

A.J.: Thank you.

Craig: Has been a great presentation. A lot of people are all saying they got a lot out of it. I think the feedback already is that you gave a lot of good knowledge there, and a lot of good tips. Thank you for that. Thank you to everyone who came along to watch. We will see you at the next SEMrush Toolbox.

All levels
Discover SEMrush