This is the third piece in our 3-part series on how to conduct research like a pro.
In the first part of this guide, we discussed the value of good research.
In the second part, we went over ways to use widely available internet resources to research and establish expertise and authority.
Now, it's time to talk about social media.
Social media has fresh, relevant, and even disruptive content.
It’s a channel that should be built into your research process.
But it’s also easy to get drowned in the opinion-onslaught. People can post whatever they want without checking facts or remaining unbiased.
This is a tricky dichotomy.
So how do you use social media for research without getting distracted or collecting biased information?
I’ve put together this guide with the exact strategies I use.
And if you’re researching across multiple mediums, here’s a quick TL;DR on other channels.
When to use each channel:
Why use social media channels?
Social media is current, democratic, and easily accessible—making it an essential research resource.
With the rise of social media influencers there is more content on it today than ever before. People are posting opinions, ideas, and recommendations every day. And while not all of it is credible (more on this later), it plays a role in building and driving conversations.
So here are my processes for researching via social media platforms:
Twitter is popular for short-form, raw, and quippy content.
It has been my primary research resource for many-a content pieces — especially those revolving around trends, current events, and fresh takes.
Here’s how I research via Twitter:
Topic: ChatGPT’s role in Content Marketing
Step 1: Type your keyword into the search box and filter by “Top” or “Latest”, whichever version best suits your needs.
Step 2: Use the advanced search feature: If you click on the three dots next to the search box, you can access Twitter’s Advanced Search feature.
You can use this to search for exact keywords and phrases.
For example, I was curious about if ChatGPT would replace writers so I looked up the exact word “replace”.
You can also filter by accounts, engagement, and dates.
Here’s an interesting result from my search. This person wrote a satirical post explaining exactly how he replaced his entire marketing team with ChatGPT. And while he was joking in the Twitter thread, his hook attracted a lot of engagement.
Step 3: Explore all aspects of the argument by going deeper into the replies and quote tweets.
For example, here is a helpful quote tweet. Because of this, I have a direction for further research, a story to build a metaphor from, and parallels to draw.
Step 4: Look up industry experts and see if they have an opinion on the topic.
For example, Rob Lenon is a popular Twitter influencer who tweets about audience building and writing. So I looked him up to see if he had posted about ChatGPT and sure enough:
Now I can further dig into the prompts he shares and write about their use cases for content marketing.
For example, here Matt Mic uses ‘99%’ but he doesn’t actually mean 99% people. He means ‘most people.’
The process of researching on LinkedIn is very similar to that of researching on Twitter.
You can use LinkedIn’s filters to refine your research as narrowly as possible.
Here I’ve filtered my results by “Post” and sorted them by “top match”. Had I wanted to be even more specific I’d have specified the author industry and date.
Let’s look at this post by Kevin Indig:
Kevin Indig is a thought leader in the content world and two of his top predictions are about ChatGPT and the use of AI tools for content.
Here’s how I can use this research:
- I can quote him directly on ChatGPT
- I can dig deeper into Google’s ChatGPT competitor and why Google is interested in building it.
- I can interview Kevin for my own ChatGPT blog.
But here’s the thing about LinkedIn: there’s almost as much value in the LinkedIn comments as the post itself.
For example, this comment by Philipp Gotza highlights an interesting quote, adds a personal opinion and an example. This parallel between TikTok and ChatGPT is apt, detailed, culturally relevant and can make for a relatable comparison section in our blog.
Method 2: Another way I use LinkedIn for research is by looking for events around my topic.
Here are some I dug up for ChatGPT:
Method 3: You can join LinkedIn groups for hyper-relevant information. You have access to a community of professionals with the same interests as yours. You can use it to identify and connect with subject matter experts for interviews.
Method 4: You can use the platform's "Pulse" feature. Pulse lets you publish long-form content and reach a large audience. By following the Pulse posts of industry experts and thought leaders, you can get leading ideas and analysis of trends in one place.
Method 5: Use the platform’s ‘Learning’ feature. (You can find this in the drop down menu under “Work”)
Just type in your topic in the search bar and explore the results you can use.
(Note: I couldn’t find exact results for ‘ChatGPT’, so I searched for the next closest term “AI in Content”)
You can filter these by:
- Language
- Type
- Time to complete
- Level
- Continuing education units (Available in “all filters”)
- Subjects and topics (Available in “all filters”)
- Hands on practice (Available in “all filters”)
- Software (Available in “all filters”)
A lot of courses that LinkedIn provides are ‘ongoing’. This means that fresh information is continually added to the course.
Pro tip: Chapter quizzes often highlight common gaps in people’s understanding of a topic. In my experience, it has been an invaluable resource for creating a relevant but unique angle.
Another thing I always do when evaluating a course is look at this brilliant section:
It gives me reviews from learners, information about the instructor and evaluates if they’ve conducted more courses.
Here I can see that Martin Waxman has created a lot of courses around digital and content marketing. Not just that, he has been actively updating courses. Each of these courses also has a high number of learners. All of this information works to cement the instructor as an authority.
Another advantage of this section is that I can see if this course is part of a bigger series.
Here I can see this course is a part of the series “Courses watched by recently promoted business professionals.” This information is invaluable if I’m writing for a similar audience.
You can look up your topic in the Reddit search box.
The filter options you’ll have are:
- Posts
- Comments
- Communities
- People
Posts
Explore posts which request unique perspectives.
Doing this sets the stage for interesting, critically thought out inputs from readers.
For example, this post asks “I plan to use ChatGPT as my AI cofounder to create a business from scratch. Code, marketing content, everything. How far do you think this can go?”
Here’s an interesting response:
Comments
Here are two (actually useful) ways to use Reddit comments for research:
- Analyze the sentiment and engagement of comments on Reddit posts related to "ChatGPT" or "language models" to understand how the public perceives the technology and its potential uses in the field of content marketing.
- Identify and reach out to users who have shared their personal experience or case studies of using ChatGPT in their content marketing strategy, as a way to gather first-hand information and insights.
For example, this opinionated take by Sam Harris offers a critically thought out explanation of what ChatGPT can and cannot do for content marketing.
In writing this piece, I would use the idea behind Sam’s take “ChatGPT is already being used to write content you’re reading, and you probably don’t even know it.”
Not only is this statement accurate, it’s also a strong opinion that would make for an excellent introduction and hook in an article.
(Note: I’m not recommending that you plagiarize someone’s work. I’m recommending you use unconventional ideas for inspiration and weave them into your article to create a piece that is culturally relevant.)
Communities
Reddit communities are as raw as it gets.
I’ve used them very often to look up public opinion, new applications, and helpful use cases.
The best use case for Reddit communities for research is to look up upvoted topics and actively brainstorm how to use creative ideas.
For example, I came across this creative use case of ChatGPT.
After a lot of brainstorming, I realized I could structure this set of commands to generate a Twitter thread. I just needed to define what a Twitter thread was and feed my LinkedIn post, and ChatGPT would accurately repurpose my post.
As a result, content marketing teams would need to invest less in content repurposing (since it would become practically free with ChatGPT.)
This idea goes straight into my article.
People
Use this to connect with people talking about ChatGPT and keep a track of what the experts say.
If you do this long enough you can find trends emerging.
Other ways to use Reddit for research (with examples)
- Utilize Reddit's "Ask Me Anything" (AMA) feature to conduct a virtual interview with a ChatGPT expert or professional in the field of content marketing. This can provide an interactive way to gather information and insights for your research.
- Use Reddit's "Challenge" feature to create a competition related to ChatGPT for content marketing, to get creative and innovative ideas for your research topic.
- Utilize Reddit's "Live" feature to host a live Q&A session with experts or users experienced with ChatGPT for content marketing. In addition to providing an interactive way to gather information, it also provides an opportunity for others to ask questions and gain insight.
Instagram is a great place for research if you’re writing for a B2C or D2C audience. Here’s a quick breakdown on why:
- The visual richness of Instagram makes it scan through relevant content in an engaging format (you don’t have to read too much)
- With “Instagram Influencer” now being a recognized profession, you can access niche content created with one person’s (or influencer’s) point of view.
- Instagram hashtags can be used to discover and research specific topics, industries, and communities relevant to a B2C audience.
- Instagram’s ‘explore’ section will repeatedly surface relevant content for research.
- The "Stories" feature on Instagram can provide valuable insight into your audience's daily lives, habits, and interests, guiding your tone and voice.
Here’s how I carry out research on Instagram:
Step 1: Look up “ChatGPT” in the search box.
You can filter Instagram results by:
- Top
- Accounts
- Audio
- Tags
- Places
For trending topics like “ChatGPT”, I normally go with the top results because they are relevant and frequently updated.
Here’s an example:
As you can see, there are carousels, reels, and memes galore.
Step 2: Limit your browsing time and shortlist 6-8 resources you’re going to explore for research.
You can go down a lot of rabbit holes with Instagram, so it’s important to restrict your browsing time. I like to shortlist a few resources and analyze how they contribute to my topic.
For example:
I’d definitely use this meme as an introduction in my article “ChatGPT for Content Marketing”:
It clearly communicates how marketers feel about the questions about ChatGPT replacing their jobs.
Inspired by this meme, my introduction would look something like this:
“ChatGPT has been the talk of the marketing world since the very night it launched. People are tense about it replacing content marketers. But is there any truth to those rumors? Let’s find out.”
Step 3: Dig deeper into profiles of content creators who’ve built engaged audiences.
And by engaged I don’t mean high follower counts but an audience that comments and keeps communication going.
For me this was Alexandra Fasulo (aka the Freelance Fairy) for the ChatGPT piece.
Disclaimer: I do not agree with Alexandra’s business ethics at all.
But she has built an engaged audience and people ask her relevant questions—which further my research.
For example, this reel on ChatGPT use cases.
Now I personally disagree with most of these use cases. Selling AI-written content to clients is unethical. But let’s look at the comments section:
Willowann makes a relevant point about Google punishing AI-written content.
Now as someone writing about the future of content marketing with ChatGPT, this is highly relevant.
I can do a deeper dive into whether Google plans to punish AI-written content and write a relevant disclaimer for my audience.
These are the top ways I use Instagram for research but there are multiple others.
I stick to these because I find them reliable and unbiased research methods.
The most important thing while researching on social media is to maintain your focus and check your research for biases.
Customer Reviews
Customer reviews are easily accessible, reflect on the real-time experience, and render plentiful research material. ChatGPT, being an AI tool, has reviews on ProductHunt, SourceForge, and personal blogs as well. Here’s a breakdown of why they work:
- They’re recent. Reviews are concurrent, and you can find new opinions on a tool to form your impression and write about it.
- They’re helpful in decision-making for your readers. Whenever they want to buy something, friends and family give recommendations. Your researched content can be a piece of friendly advice for them, but more in-depth.
- They help filter through the fluff and reflect prevalent issues in the tools, and you can use them to warn the readers.
Step 1: Search “ChatGPT” customer reviews on Google console
You’ll get a bunch of review tools and personal blogs about ChatGPT and it’s experience. For customer reviews, I scroll to the 3rd or 4th page, because new reviews can also be at the bottom of the SERPS.
Google pulls up the top-ranking content (value-based content), and also helps me understand the sub-topics that interest the people from the “People Also Ask” section.
Let’s see how the first-ranked review by Marc Abraham adds to my research:
Marc Abraham is the head of Product Management at Intercom and the author of "My Product Management Toolkit" and "Managing Product = Managing Tension." He's done this extensive review of ChatGPT with his prompts on medium.
Marc's review is helpful in my research because he not only delves deeper into the advantages of Ai written content but also discusses the topic it refuses to create content on.
His review helps me add a "Limitations Section" for my post. He specifically discusses the refusal of the tool to comment on religious or current affairs topics. I can use this tip to create related prompts, which will further save time on prompt research and topic restrictions.
Step 2: Customer Review Websites
Review websites like G2, and Capterra, are also the top-ranking results for my "Customer Review" search. Here, I can contrast and compare the impact of ChatGPT for people with varying use cases. These reviews are composite but digestible and help me find unique points to raise and test in the content.
Here’s a review I liked:
To build a cohesive view of the tool, I can combine pros and cons from multiple reviews, find similarities, and single out common hurdles and benefits users derive from the tool.
ChatGPT is still in its Beta phase, so I didn’t find 100s of reviews on these websites, but once the paid version goes live, I’ll be able to understand its user experience from multiple TAs like businesses, writers, product managers, etc.
Step 3: Compile different reviews to form advice
Customer reviews are best for investigative/advice pieces. It helps form the conclusion if the product is valid or not by reading multiple reviews.
A quick glance at these two reviews indicates that ChatGPT saves time, is ad-free, free-to-use (for now), and has a team that’s active on websites.
An active team indicates they’re eager to make improvements and launch the new features soon to transition into a paid tool. This helps me advise my audience to practice and test the tool’s free version, so they don’t hit roadblocks with the paid one.
How can I use customer reviews?
These customer reviews reflect on the lack of transparency and practice in using the tool. So, I’m going to create examples from the tool to create different prompts and reference them from the reviews that say, “It helps save time,” and compare if it’s the truth.
Next, the reviews say it doesn’t have access to google. So, I’m going to create a prompt that requests for a 2023 trend and check if the dialogue result lacks the answer to it, and how it compensates for this drawback.
Here is the result:
As evident, the model’s training data cuts off in 2021. But it can provide me with generic statistics to compensate for its lack of knowledge. The customer review was helpful as it mentioned the problem in a singular sentence which further led me to inspect and discover an eccentric trait of the tool.
Another way this research contributes to my content is to help me create a list of common questions to answer in the post.
In the case of tools that have reviews a plenty online, I use them to create subheads and also find a converging point for customers to highlight the most relevant things for customers to notice.
There are other ways to collect customer reviews, like surveys, DMs, and interviews, but this is the best for time-bound tasks.
The last word on research
Now that we are at the end of our three-part guide to research and I’ve walked you through every single tactic I use, here are my five unconventional tips:
1. You don’t need to research in one go: you might have come across a writing ‘best practice’ to research>outline>write>edit, but you really don’t need to adhere to it.
I don’t.
I go back to research multiple times while writing and sometimes even while editing—to cross-check facts or find more relevant data.
2. Ask other writers: sometimes, you can spend ages looking for a data point or idea that a niche writer might be very familiar with. Get in the habit of reaching out and asking (and also giving back.)
3. Always save your links (even if they’re the first SEO result): I’ve lost a lot of key ideas by not noting down links when I came across them. I spent hours searching for them.
4. If you are a niche writer, bookmark a few favorite resources (website, podcasts, newsletters) that you always find yourself going back to. It’ll save you a lot of time in the long run. You won’t have to go down research rabbit holes every time you’re assigned a new piece.
5. Refer to the guide I’ve listed at the top of this page so you can streamline your research sources before you begin the actual work.
And finally, have fun with research. Allow yourself to learn new things and be drawn toward what interests you.