How I write research-based posts

Post image

I've been writing research-backed blog posts for a few years now, and I've been lucky enough to have great success with some of my work. It's been featured on Lifehacker, Fast Company, Inc, Business Insider, Huffington Post, and I also helped to increase the Buffer blog's traffic by over 50% in my first month there, and doubled it in my second month, mainly as a result of some well-received research-backed blog posts.

As Groove founder Alex Turnbull said in a post for the Buffer blog, research makes your content more credible. And while stories can help us remember things, using statistical evidence has been shown to be more persuasive than simply telling a story.

If this is all new to you, or you just want to get a look inside my process, this guide will break down the steps of writing a research-based blog post from start to finish.

Since there's a lot to get through, here's a list of sections to help you jump around quickly:

Choosing a topic

I tend to have a list of topics laying around most of the time. I've written about how I get ideas before, and I always make a note of ideas I have that could turn into content topics later on. My current ideas list looks like this:

topic ideas

What I've found, though, is that when I first think of a topic, that's usually the time when I'm most excited about exploring it. I keep my list of topic ideas for the times when nothing comes to mind immediately.

Sometimes I use these lists just to stir up new ideas. You can see that lots of the cards in that screenshot are grey instead of white—those are the ones I haven't touched for a long time. Whenever I come across some research that relates to one of these I'll add notes to it. Sometimes an idea will sit there for months before I come across some research that suggests a different angle to me, and that's what will finally get me to write that piece of content.

When I'm having trouble choosing a topic I ask myself these questions:

  • Who is the audience?
  • What are their interests?
  • What do they need help with?
  • What am I curious about at the moment?
  • What have I come across lately that sparked my interest?

Research doesn't just help me create content. It's also a process for coming up with new ideas. Looking through my RSS feeds or a tool like Prismatic, I can see what other people are writing about a topic at the moment. If I look at Reddit or Quora, I can see what questions people are asking, or what they need help with. In the gaps between these two things is where successful content ideas come from. Adding in my current interests and what I've been reading about lately gives me a way to find a unique approach for each piece of content.

The magic happens when you put all of this together:

What your audience needs help with + Your unique angle - Content published by others


Researching is often the most time-consuming part of creating content. From finding what you need to pulling it all together into a coherent narrative, there's a lot of work that goes into each blog post.


If you don't have a process set up for collecting research as you read, now's the time to get started. There are two main parts to this process: saving bookmarks and taking notes.

I try to save and tag anything I read online that might be useful at some later point. I currently have 6,600 bookmarks. I keep a folder of bookmarks called Research, that contains any studies or articles I might want to reference in my content—this folder has 374 bookmarks so far. This makes them easier to find than throwing all my bookmarks into one place.

A lot of the studies and articles I save come from Google Alerts. If you haven't set up Google Alerts before, you should try one right now for your own name.

Google Alerts

Just type in your search term, how often you want to get results via email, and what to include in your search.

Google Alerts set up

Google helpfully shows you a preview of the results you'll get to help you fine-tune your alert settings before you commit. Once you're signed up you'll get new results delivered to your inbox.

Google Alert preview

I receive all my Google Alerts emails once a week. I work through them in a block, opening interesting links in my browser, and then as I read through each article I bookmark any that might be useful later. Because you can set up Google Alerts once and forget them, this research material comes to me every week and keeps my research folder topped up without me spending time searching out new studies or articles in my areas of interest.

Google Alerts emails

The rest of my research bookmarks come from articles I happen across on Twitter, in my RSS feeds, and interesting finds from Reddit, MetaFilter, or Medium. These don't have to be links to studies or academic papers; sometimes they're blog posts about someone's personal experience, or graphs of data that I think might be useful. I err on the side of saving too many bookmarks so I always have lots of reference material at hand.


I use tags to organise the bookmarks within my research folder, so I can filter by "sleep" or "learning", for instance, depending on what I'm writing about.

Bookmarks with tags

I also have a bookmarks folder for everything I've written. This comes in handy when I write about related topics or different angles on the same topic, because I often go back to the same studies I've referenced in the past as a helpful starting point.

bookmarks of my writing

The other part of collecting research as you read is taking notes. Note taking is how I save and remember quotes from books I read, interesting stories I hear on podcasts, or random thoughts I have when I'm exercising. Putting these down on paper further solidifies them in my mind, so I can better connect them to my research or future notes.

Studies have shown note taking is most useful (at least in terms of learning) when it's combined with reviewing your notes. So don't fill up a notebook and then let it wither on a shelf. You'll need to take it down from time and time to keep those notes fresh in your mind.

You don't have to use a notebook, though. Author Ryan Holiday uses an index card system to organise his notes. Each card has a single note on it, and is filed in a card box that's organised by category. This makes it really easy for Ryan to find all his notes related to a topic when he wants to write about it.

Ryan's cards

If you're a fan of notebook tools like Evernote, you can use tags in a similar way. This might even be more useful, since you can cross-reference your notes between difference categories, whereas a physical index card can only live in one category at a time. Having said that, science shows handwriting your notes will make them sink in a lot better than typing.


When I've found all the research material I'd already collected that relates to my topic, I read through it all. This helps me pare back to what's most useful and get a grasp on the direction my research might take. When I've decided what to work with from my existing materials, I start searching.

I usually start with Google to get an idea of the keywords that will help me find what I'm looking for, and to find some articles with good introductions to the topic. Sometimes I start my research with a keyword that isn't used in studies so I need to figure out the right words to search for before I can even start collating research.

If I don't have any saved in my bookmarks, other articles about the same topic can be useful to help me get my head around the basics before I start digging into more specific research. I start with a normal Google search so I can turn up these "introductory" articles that I find useful.

Once I've read any articles I found and I have a good idea of the keywords to use, I move on to Google Scholar and Statista to find studies and papers on my topic. I quite often find links to studies in other articles on the same topic too, which can be more efficient that trawling through pages of Google Scholar results.

I try to limit myself to about five or six open tabs at a time. I open many more and skim them, but close the ones that aren't useful. When I've got about five that seem useful, I move on to taking notes. A really useful tool for this process is Evernote's Clearly browser extension. Clearly cleans up articles much like Evernote or Readability, to let you read them without ads or a distracting sidebar. It also gives you a highlighter tool so you can mark sections of the article you want to come back to.

Evernote clearly

If you have an Evernote account you can save these articles with your highlights. I find it useful to just leave the tabs open in my browser so I can switch back and forth between them, using my highlights to remember which parts are most useful for what I'm writing.

Many times I'll come back to the research stage later, but I've found the quicker I can get some words on the page, the easier it is to find some momentum with the post.

It's way too easy to get bogged down in research. The more you read, the more you realise how much you still don't know, and before you know it you've wasted a whole day searching and reading and you've got nothing to show for it.

Don't be afraid to start writing before you've got the whole picture in your mind. As soon as you have some interesting points in mind or an idea of what your outline could look like, get those words on the page. The more you write, as counterintuitive as it might be, the better your research efforts will be, because they'll be informed by the progress of your content. Without writing anything to steer your reading, you're leaving yourself open to the blackhole of the internet and you'll find it hard to come back.

xkcd comic

Evaluating your research

One of the most common questions I get about researching content is this one:

How do you know what's high-quality information and what's not?

It's tough, for sure. The more research you do, the better idea you'll get of which sources you can rely on and which ones are sketchy. There's definitely a good helping of gut feeling at work when evaluating research.

However, there are some things you can rely on as general rules to help you out. These are the main ones I stick to.

1. Always go to the source

Citations comic

A trap I fell into when I first started putting lots of research into my writing (and one I have to remind myself not to do, still) is quoting a quote. Although I like to start with other articles and blog posts about the topic I'm researching, this is only to get my head around the topic in a general sense.

Quoting a study or a fact without following through on another author's research to verify it is lazy research. Whole books are written on the shoulders of shaky or misinterpreted evidence (and sold millions of times over) so if an author is new to you, don't trust their interpretation of the facts—go directly to the source. Only then can you (and your readers) be satisfied that you've represented the research as accurately as you can.

Sometimes the source is no longer available, or it's behind a paywall, or you can't understand it as easily as another author seems to have. This is where point #2 comes in.

2. Don't misrepresent information

I'm not a scientist. I don't really know how to read studies and understand all the mathematics that determines how statistically relevant information is. This is something I want to improve on, but for now I do my best to check for these things before citing a study or academic paper:

If I don't understand it, I don't use it.

Sometimes academic writing goes over my head and I don't fully understand the results. If I'm not confident that it confirms a claim I'm making, I look for more studies.

If the population is too small, I take the results with a grain of salt.

The population is the group of people in a study (or animals, or whatever group of things is being studied). If I'm trying to prove a point and I find a study of, say, 10 people that backs me up, I'd be remiss to pass that off to my readers as proof of my claim. In that case I would either not use the study at all, or I'd mention it but make clear to my readers that the study was small, and therefore we can't draw from it any accurate conclusions about the average person.

Here's an example where I've done this: in a piece titled 8 Surprising Ways Music Affects and Benefits Our Brains, I included this section:

It’s not just kids that can benefit from musical training or exposure. Stroke patients in one small study showed improved visual attention while listening to classical music.

The study also tried white noise and silence to compare the results, and found that, like the driving study mentioned earlier, silence resulted in the worst scores. Because this study was so small, the conclusions need to be explored further for validation, but I find it really interesting how music and noise can affect our other senses and abilities — in this case, vision.

If the study focuses on a particular group, I don't apply it to the average person.

Sometimes you have to look carefully to find this out, but studies often focus on particular groups—students, people from a certain geographical location, or people of a particular age group. This is often out of necessity, because these people are easier for the researchers to get access to (in particular, university students are used in lots of studies).

I might come across a study on exercise, for instance, that shows short interval training is as effective for fitness as long, steady exercise. But if that study was made up only of, say, 20-25 year-old men, it's not accurate to say that finding applies to everyone. Even if the group is large, if they're of a particular demographic it's important to either note that when you cite the study, or find one that applies to more people.

I made this mistake in a Buffer blog post about the science of happiness, where I wrote "happiness is maximized at 13.9°C". In fact, the study I used to back up that claim used data from Japan, which means the findings aren't necessarily relevant to people living elsewhere.

Sometimes you have to make mistakes to realise how you can do better.

3. If you can't support your point, cut it

I actually went through this exact situation when writing this post. In the editing section below, I suggest you read your post out loud. When I was drafting that section I went on to say you should record yourself reading the post and listen to it, since listening and speaking use different parts of the brain so you'll get a new perspective on your work.

This is the kind of claim you can't just go making without evidence. But I wasn't entirely sure I was right. If I knew I had some research filed away to back this point up I would have kept writing and come back to it during the editing stage. But I wasn't sure about this one, so I went searching.

I typed a bunch of words into Google before I found anything useful. Just to give you a realstic idea, these are some of the terms I tried:

  • take in information talking listening
  • learn differently talking listening
  • how are talking and listening different
  • talking and listening different language skill
  • talking and listening different language skill brain
  • talking and listening different brain

Useless Google results

That's six Google searches before I actually found something useful. Which either shows how much effort it takes to find useful research, or how bad a Googler I am.

Useful Google results

With this sixth search term, I actually found some research that totally refuted my claim. It turns out there's a lot of overlap in the brain areas used for speaking and listening.

When that happens, I just cut the point I'm trying to make. I don't argue with the research.

4. Learn to be sceptical

Above all, be sceptical of anything you read that isn't backed up by reputable sources. It will stand you in good stead to be sceptical rather than accepting everything you come across if you want to deliver well-researched, respectable content.

Research-based writing is hard work. It takes a lot of time and effort, and it can be incredibly tempting to cut corners (trust me, I feel that too). But you must remember to give your readers the respect they deserve.

Help your readers by doing the work for them to back up your claims. The more they can trust you, the more confident they'll feel about accepting what you write and putting it into practice.

Structuring your post

So how do you take all of that juicy information and pull it into a complete blog post? First, you start with an outline.

In fact, I start all of my posts with outlines. It helps me see the big picture and make sure I'm telling a coherent story. Our founder, John, does the same thing.

Your outline structure will depend on the piece you're writing, who will read it, and where it's going to be published. But as a very general rule, here's what I usually start with:

  • Title
  • intro
  • Subheading 1
  • first main section
  • Subheading 2
  • second main section
  • Subheading 3
  • third main section
  • conclusion


This is a very broad starting point, and I often change the format depending on the content. For instance, in some pieces I'll end each main section with a useful takeaway for the reader, much like Alex often does on the Groove blog. For other pieces, I'll add a fourth section that focuses on actionable points from the post just before the conclusion.

John's most common outline is different again:

  1. Introduction of the thing
  2. Value of the thing
  3. The THING
  4. Case study of the thing
  5. Conclusion

One thing I never go without is a title. I absolutely must have a working title. It doesn't have to be good, but it has to give me an idea of the point of this post. The title is like my north star—it helps me focus and stops me from going too far off-topic.


Depending on the topic, there are two ways I approach this stage. One is to go directly to a blank Markdown document and start writing from memory. If I've got a good handle on the topic after all my research, this is a good way to get words on the page and overcome the blank page syndrome.

It also helps me to synthesise the information I've taken in and make sure I understand it clearly by testing my knowledge from memory. After the first draft is done, I go back to the research stage and fill out what I've put in my draft (and correct anything I remembered wrong).

For other topics, particularly for posts which will contain lots of quotes, I copy-and-paste a lot of the research material directly into my Markdown document. I put the URL for each quote or chunk of text right under it so I know who to credit for it, and I write my draft at the top of the document, pulling in the source material as I go. If you're smarter than me, you might just use two separate documents side by side—one with your research and one to write your post in.

Once I have something on the page to work with, I like to add some more structure to the post. This helps me get my purpose clear in my head and gives the post a better flow from the beginning. It can be a lot harder to restructure a post once it's filled out and I've written each paragraph carefully to flow into the next one so I like moving things around at this earlier stage.

I'll do a lot of cut-and-pasting at this stage to rearrange points I've made or adjust the outline entirely so it flows better. If I'm working on a long, sprawling piece, or if I'm unsure of the structure I have, I'll work through the bits and pieces I've got from top to bottom and summarize the sections on paper. Then I can look at the paper and play with the order of the sections based on a short summary of what each one covers. It's a good way to "zoom out" from the piece as a whole. (Again, you could do this in another document. I don't know why I've never thought of these things before.)

For this post I wanted some advice on the structure from Ghost founder, John, so I created a summary snippet in Slack, which looked like this:

Slack summary snippet

Once the structure is in better shape, the draft needs filling out. This is the process of taking points I've barely touched on, bullet points, or notes to myself and expanding them into readable sentences. I like to work from top to bottom, getting a rough idea of the way the post flows as I work on it. Any facts that need checking or links I want to add usually get skipped at this stage so I can focus purely on getting words on the page. The primary goal is to make sure I say everything I need to. Clarity and polishing come later.

The introduction can be a tricky beast. Sometimes I get so stuck with the introduction that I skip it entirely and come back to write it later when I'm editing the piece.

Lately I've found that writing anything, as bad as it might be, can be the perfect approach for the introduction, because it gets the words flowing and helps me build up some momentum before going into the first main part of the post. I often cut my entire introduction and rewrite it during editing, because I let myself write such rubbish the first time around.

Just to make this super clear, because it's one of the most useful tips I have to share: I will literally write an intro during the draft stage that says something like, "I have no idea how to start this post. I want to talk about this topic but where to begin?" This kind of nonsense can be surprisingly helpful in lubricating the gears of my mind and helping me to find a way into the topic.

If you have as much trouble with introductions as I do, think of your draft intro as a run-up to your first "real" section of writing. Use it to get yourself into the flow of putting words on the page, and to find your voice. The less you force it to be "right" the easier it will come, I've found.


The editing stage is for major fixes. Any quotes I need to add attribution for, any links I need to fill in, or names I need to double-check get fixed now. I read through the post from top to bottom, making edits as I go.

During this stage I also reword anything that sounds clunky when I read it out, and cut out any words I don't need. The word "that" is a common offender. Take a look at the next thing that you write and you'll probably find that you use the word that more often than you need to.

I tend to read my work out loud even as I'm writing it, just as a habit, but if you don't, take the editing stage as an opportunity to do this. Words never sound the same out loud as they do flat on a page, so this trick helps me pick up clunky sentences I would have missed otherwise.

The other main part of the editing stage is to swap out or add sources to make sure my points are all well backed-up, and to add extra research time for any parts of the post that need a bit more "meat".

Final touches

I refer to this final stage as "polishing". The drafting and editing stages are where you massage your work into the form you want. It's a bit like molding clay: you slowly, gently push and prod at the work until it's in the basic shape you wanted. Then comes the polishing.

Polishing involves another full read through from the top. As silly as it might sound, I do look for rhythm in the way my words read. I only make minor changes during polishing; I change a word here or there, add some punctuation I missed, add italics or bold for emphasis.

As I polish a piece, I'm looking for the points that jump out, the points a reader will take away with them, any parts of the post that are ambiguous and need more clarity added. And of course, the way it reads.

Before publishing a post or sending it off to an editor, I ask myself these questions:

  • Am I proud of this work?
  • Would I put it in my portfolio?
  • Is there value here for the reader?

The last question is the most important. I want readers to finish every piece I've written and feel like they've either learned something or that I've piqued their curiosity and they now want to explore the topic further. Either of those is a win in my books.

I'm only ever writing for the reader.

write for the reader

Remember to get words on the page as soon as you can, and be prepared to spend a lot of time massaging your work into the best it can be.

Research-based content takes effort, but the more you do it the more naturally it will come.

Image credits: Wikipedia comic via xkcd. Citations comic via xkcd.