How to Research For A Blog Post That Ranks [& Gets Cited by AI]

In the guide, I will show you how to research your blog posts.

I will walk you through the research process I have used to write 600+ blog posts and share practical tips for finding high-quality sources. 

So you can create content that is genuinely helpful for your readers and outperforms competitors in search engine results. 

Let’s dig in!

Here’s a video where I explain the whole process:

Here’s how I gather and organize information when I’m writing a blog content outline.

When a brief lands in my inbox, and it’s a topic I’m not familiar with, I plug the primary keyword into Google.

I upload the 5-10 top SERP results to Google NotebookLM and generate the Audio Overview.

That’s how I get a general understanding of the topic and what the competitors write about. 

If I’m familiar with the topic? 

I might skip the Audio Overview, but it’s still useful to have all the competing content in NotebookLM. So you can pull information from it without switching between tabs and looking for it manually.

Search intent is the reason why the reader runs the search on the keyword. And the best way to find it is by looking at the competing articles: their format, content, and visual presentation. 

The rationale? If the articles rank, they must give readers what they need.

For example, if we search for “blog posts”, 3 out of the top 5 results are how-to guides. They teach the reader how to write high-quality blog posts.

That’s the format I would try to emulate if I wanted to rank for the keyword.

Can you get search intent from a keyword research tool?

Sure, I’ve just searched for “blog post” in Ahrefs, and it returns that the keyword has informational intent. 

But how useful is this insight?

Not very. It tells you nothing about the articles that rank for the keyword. And most blogs target either informational or informational/commercial keywords.

Here, I dig deeper into competitors’ content to understand what they do well and what they don’t.

To start with, I plug all top articles into ChatGPT and ask it to:

Here’s a screenshot listing some of the unique questions answered by articles about e-commerce returns.

Reddit is a treasure trove of target audience insights. That’s where readers go when SERPs fail them. 

I use Gigabrain to discover questions Redditors ask about the topic and problems they want to solve. Then feed the results into ChatGPT and ask it to cross-reference the results with the competitor analysis results (from the previous step). 

Here’s what the results looked like for the article about e-commerce returns.

That’s 8 questions that competitors fail to answer, most of them about practical ways to minimize the impact of customer returns on the business.

I tackled them in the main article body and answered a couple more in the FAQ section. 

After competitor and Reddit research, I send out my requests for SME contributions by:

I’ve discovered that asking for SME help at this stage hits a sweet spot.

Because I already know if the article will benefit from SME insights (not all articles do), I know what questions to ask them, and it gives them enough time to contribute.

Next, I move on to the most powerful resources: the client’s internal assets: their case studies, templates, frameworks, and internal data. 

These are priceless because competitors can’t access or easily use them without referencing the source. This alone may be enough to create unique content that radiates authority in every direction.

⚡ A pro tool hack: Create a separate NotebookLM file with all your clients’ case studies and internal reports so you can easily access the data.

Google is still the most reliable way to find primary sources, like research reports, case studies, and original data from reputable sources. 

Here are the most useful search operators:

Here’s a confession: I rarely use these operators these days. Google’s Advanced Search panel is way more intuitive — and foolproof.

And again, anytime I find a useful source, I upload it to the NotebookLM project file. So I can extract relevant data points when I start writing.

I never trust AI to find data, especially if I’m after up-to-date stats. But Deep Research is excellent for expanding the research scope beyond the first page in Google. Something that would take hours if not days otherwise. 

I use three chatbots with the Deep Research functionality —  Perplexity, Gemini, and ChatGPT — and I run the search on all three of them at once.

Then export the results and use ChatGPT to synthesize them. 

Not every blog post needs academic references. But for some topics, like healthcare, education, HR, or management, peer-reviewed sources can add depth — and authority.

Google Scholar is fine for finding relevant academic articles. But it’s a bit basic.

These days, however, I lean into AI-powered tools — Semantic Scholar and Consensus – that not only find the sources for me but also summarize them and synthesize the results. And assess their quality!

Bloggers often overlook books in their research.

And I get it: Finding relevant information in a book is like finding a needle in a haystack, and they’re often out-of-date the moment they hit the bookshop shelves. 

Still, I always run a quick Google Books search to see if I can find any evergreen insights or quotes. 

I’m also in the process of converting my Kindle books into PDFs and uploading them to NotebookLM, so I can search them more efficiently.

If books give you timeless insights, newsletters give you what’s happening right now. And they often include exclusive insights, case studies, and data that sometimes never appear in blogs.

I subscribe to dozens of newsletters across the niches I write about regularly, like SaaS, SEO, product management, marketing, and e-commerce. 

The challenge? Staying on top of all of them. 

I’ll be honest with you: I don’t have time to read all of them.

So, I use Zapier to automatically pull the newsletter content into a Notion database, where I can search it all in place whenever I’m researching a piece of content. 

Taking the software I write about for a spin is the most reliable way to discover details that competitors miss. Because — let’s face it — most writers rarely use the tools they review in their blog posts (even if they say they did to tick Google boxes). 

And readers can see it immediately.

So, at the very least, I always sign up for the free trial.

That’s often enough to assess the onboarding process, test key features, and spot usability quirks.

For more in-depth tests, I ask my client to share their premium subscription with me — or purchase one if they haven’t got it.

And I:

When researching BOFU content, I dig into customer reviews on platforms like G2, Capterra, or Clutch. 

I use them to spot recurring pain points, find product pros and cons, and capture customer vocabulary. 

Thanks to these, I can write about the products objectively and authoritatively, even if I can’t test them myself. In a language that the customers understand. 

Podcasts and YouTube videos provide unique insights, fresh perspectives, and expert soundbites.  

The reason? They’re more difficult to process than written text, so many writers don’t bother. 

They used to, anyway.

These days, audio recordings are as easily accessible as blogs, newsletters, or books. 

Here’s how I go about podcast research:

With YouTube, it’s even easier because there’s no need to download the files; just copy/paste the video URL, and NotebookLM takes it from there.

⚡ A pro tool hack: Whenever I need a full transcript of the episode or video, SembleyAI is my go-to transcription tool. 

LinkedIn and X feeds are often better sources of current expert insights than contributions on PR platforms.

Here’s an example of a LinkedIn post I used in an article about 360 feedback, as it offered unique insights that no other article in SERPs covered. 

That’s why I always scan the two platforms for relevant posts that I could use in my content.

And save them religiously in my swipe files in case I can’t find them in the future (LinkedIn’s search engine isn’t the most reliable or user-friendly).  

That’s it, folks, here are the tricks of the trade that I use to write well-researched content for my clients. 

If you’re a writer, I hope this helps you when researching your blog content.

If you’re an editor or content manager and would like to work with a freelance writer who goes above and beyond to create unique content, get in touch!

get in touch

You still here? Jolly good. Here are a few FAQs about research I often see.

I need 3-8 hours to thoroughly research a blog post. This depends on how familiar I am with the topic and what resources I already have in my Resource Vault.

Articles that require SME insights are more time-consuming as the contributions take longer to source. 

I check the publisher, the author, and the publication date. Reports from established organizations, like Gartner, McKinsey, or brands like Ahrefs, Semrush, SparkToro, or Surfer, carry more weight than anonymous blogs.

I also look for stats no older than 2–3 years unless I want to illustrate historical trends. And if I can’t trace a claim back to the original report, I won’t use it.

When researching my articles, I use:

I stop once I can explain the topic clearly without leaning on competitor articles.

If I have credible sources to back up my claims and know how my angle will differ, I move to the outline.

Research shouldn’t be endless — it should give you enough confidence to start shaping the piece.

To avoid overwhelm, I centralize everything in NotebookLM so I can easily ask for specific information or data points. 

I also have a Resource Vault in Notion where I keep track of all the sources I’ve used in previous articles in case I need a specific data point again. However, I update it less and less now as it’s easier to organize and recall data and facts in NotebookLM.

Good research always takes time, but you can speed it up and make it more in-depth with AI-powered search tools like GigaBrain or Consensus. 

With AI chatbots, like ChatGPT or Gemini, it’s also easier to synthesize information and extract relevant insights. Which means you can work with more sources and gather more data.

Finally, NoteboookLM is perfect for organizing your research, so you can easily access the info you need — whenever you need it — without going through the sources manually again and again. 

When I work for a B2B company, the content ideas come from keyword research and product or service offerings.

For example, when I worked for Honestly, the employee survey platform, we targeted keywords related to employee surveys and employee engagement to showcase the tool’s features and present it as a solution to customer pain points.

Topicality is another factor. A lot of the content I write covers trending topics.

For example, lots of my recent articles about SEO and link-building focus on increasing visibility in AI search results.

When it comes to my blog content, it’s less about keyword research at the moment and more about sharing helpful advice and hacks I learned as a writer and editor.

This article was originally posted on tatarek.co.uk on 9 September 2025.