There is no shortage of AI marketing tips and tricks to weed through on LinkedIn. Finding useful ones you can apply right now is a little harder.
That’s why we were so blown away by the responses when Amanda Natividad, VP of marketing at SparkToro, asked her LinkedIn followers for their AI marketing use cases.
The roughly 100 replies are packed with creative prompts and ideas from marketers who use them daily.
We wanted to see how they worked IRL, so we plucked out a handful of those ideas and tried them for ourselves. There are new ways to reposition products, instantly reframe content, mine Reddit for consumer insights at scale, and a lot more.
We couldn’t test all the brilliant ideas, so we also listed several options we’ll try in the near future. Have a look through these tips and see if there are a few you haven’t tried yet.
Contents
5 uses for AI in marketing we tested
The goal is to make this list as practical and usable as possible. So, we picked AI use cases with the biggest payoff for the shortest learning curve.
We also used a variety of AI tools to show off their capabilities. And we stuck to the free versions so anyone can replicate these experiments without a paid subscription.
1. Learn how your products are actually positioned
We all know what we want people to think of our products and services. But sometimes, it’s hard to know how they’re actually perceived in the wild. Steve Lamar, a founder of several tech businesses, shared a clever way to learn how his SaaS marketing calendar was positioned.
Here’s how he did it. Steve asked his AI tool, “What are the best marketing calendars.” When his product didn’t make the cut, he asked why. The AI offered a few cons that helped him reposition his product. Very clever.
Ready for more? Get the Emergency Guide to AI in Marketing.
The test
I asked ChatGPT to provide a list of the best scheduling apps available. Here’s what it gave me:
I then Googled scheduling apps to find one not on the ChatGPT list. I picked Rippling because they had a Google Ad targeting the key phrase “best meeting scheduling apps.” That brand actively tries to position its products as a solution for that key phrase.
Following Steve’s lead, I asked ChatGPT why it didn’t include Rippling in the original list. It offered some interesting feedback.
I don’t know how Rippling wants to position its products. But if the goal is to be a universal scheduling app for any business, this feedback would be really helpful.
ChatGPT also gave links to a review of the platform and a Reddit discussion about it. The brand could use them to update its website, guide influencer marketing initiatives, and jump into Reddit discussions to change how people think about its product.
How to use this tip
This was such a useful and straightforward idea. You could use it for just about any product (e.g., I also tried “best mountain bikes” and “best social media scheduling software”).
There are plenty of ways to refine this tactic for different audience segments. For example, you could ask, “What are the best mountain bikes for kids?” or “What’s the best budget-friendly social media software.” If your product doesn’t show, use AI to figure out why.
Take it a step further and ask AI how you can reposition your product for a particular market segment. Rippling could follow up and ask how to position its app for a broader audience or specifically for small businesses.
2. Mine subreddits for customer intel
Reddit has become an increasingly influential marketing channel. It’s a great place to connect with dedicated groups of like-minded people. Plus, Google places Reddit answers prominently on its SERPs. We spent a few months learning how Reddit marketing works and wrote about it.
Rebekah May, global content lead and SEO manager at Haypp Group, said she uses Reddit for a different purpose. She has her AI tool mine relevant subreddits to help her build customer personas.
Specifically, Rebekah asks her AI tool to review a subreddit and “find most commonly asked questions, terminology most often used, and help me draw conclusions about challenges and solutions.”
The test
This was one of the most interesting tests I ran. It surfaces so much useful feedback that I had to keep myself from continuing down too many Reddit rabbit holes.
I chose Semrush as a test subject because I knew it was a popular topic on Reddit. I used Gemini for this test since it can access links (some tools, like the free version of Claude, can’t).
To start, I gave Gemini the link to a Reddit conversation titled “Semrush – is it essential or are there better alternatives?”
Gemini spit out a brief summary of the thread. Then I asked it a series of questions, like:
- What are the most common adjectives used to describe Semrush?
- What are the most common negatives mentioned about Semrush?
- What challenges do people say they have that Semrush solves?
Every question netted a lot of helpful information. Here’s what Gemini listed as common adjectives:
And here’s part of its answer to the “common challenges” question:
Rebekah said she used this tactic to help build customer personas. I could also see a lot of value in this for positioning statements and even guiding future ad campaigns.
How to use this tip
If people are talking about your product or industry on Reddit, you basically have a free and unfiltered focus group from which to learn. You can use Gemini (or any other AI tool that can access links) to mine those conversations for insights in many ways from dozens of subreddits in minutes.
Besides the questions I posed in the test, try questions like:
- Why do people say they bought [competitor’s] product?
- What are the most common alternatives to my product that people mention?
- Are there any specific product features mentioned most?
- What criteria do people use to decide if my product is worth the cost?
- Are there any use cases where my product is mentioned as particularly useful or unuseful?
Of course, you can also just ask Gemini to make a list of questions you should ask (that’s what I did).
Free resource! How to Create a Value Proposition From the Ground Up.
3. Quickly find contextual stats and data
If you’ve ever searched for relevant data to include in your work, you know what a pain it is to find it on Google. Inevitably, you’ll land on the same out-of-date statistic every other marketer used. Then, you’ll spend 30 minutes chasing down the source, only to end up in a never-ending spiral of posts that link to each other.
Sound familiar? Several commenters said they broke the citation circle by using AI as a research assistant.
The test
There weren’t many specifics in the thread explaining how people use AI for research. This is a use case I’m familiar with, though.
Most of the time, I need to see if there’s data to back up a point I want to make. AI’s ability to understand context comes in handy here. I also want the data to be recent.
With that in mind, I fed this prompt to Perplexity (currently my favorite free tool for research):
“Provide five statistics from 2023 or 2024 that support this statement: ‘Consumers prefer to buy from businesses that prioritize sustainability.’”
Here’s Perplexity’s first response:
A good start, but it’s pulling from statistic roundup posts. I want first-hand citations. I asked it to recreate the list “using only original sources.”
I checked each citation and, yep, they’re all original studies. No second-hand stats this time.
I gave Google’s Gemini a crack at it with the revised prompt asking for original sources.
This information is great, but it didn’t include links to each source. And when I asked it to provide the links, it couldn’t do it.
How to use this tip
This one’s easy to replicate for most research purposes. There are a couple of important caveats to consider.
First, not all AI tools have access to up-to-date web results. Double-check where your gen AI of choice gets its info.
Second, verify the source of any stat you get. For all its progress, generative AI is still prone to make stuff up.
Look at the response Katie Parrot, a freelance writer and marketer, got from Claude when she questioned its data sources:
4. Write offbeat analogies
Analogies help translate complex ideas. They make concepts more relevant and digestible for readers. But they can be hard to come up with, especially when you’re under deadline pressure.
That’s why I love the idea Hardik Lashkari, a content writer in the finance industry, gave in his reply. He uses AI to help come up with “weird” analogies that grab attention.
The test
I started with the basics and asked Claude to “Write three brief analogies that will help people understand what generative AI is.”
They could use some tightening up, for sure. Not a bad start, though.
But we want something weird and unexpected. The kind of analogy that knocks readers out of skimming mode and into understanding mode.
To get there, I asked Claude to “start over and write three weird, unexpected analogies for generative AI.” It came up with a blob of pizza-eating pizza dough, a book troll, and a kaleidoscopic chameleon.
Fun? Yes! Useful? Maybe. At least as a creativity springboard.
How to use this tip
The analogies we got from Claude needed some editing, but it understood the concept and delivered decent options to get us started.
Since the goal is to make a concept more understandable and relevant, you might ask your AI tool to position your analogies for a certain audience. For example, here’s how Claude would explain generative AI to a sports-minded reader:
The beauty of gen AI is that it remembers your previous requests so you can iterate on them until you’re happy. Try refining your analogies by audience segment, tone (funny vs. serious), or industry.
5. Reframe content for formulas
Copywriting formulas are really useful for creating influential copy and content. AIDA (attention, interest, desire, action), for example, helps writers frame headlines, sales copy, and blog posts to walk readers through their decision-making process.
Roy Povarchik, organic media lead at Riverside.fm, uses AI tools to reformat copy around copywriting frameworks like AIDA. He said it helps him “get out of his head and not be fixated.”
The test
I grabbed several random pieces of copy and fed them to Claude to test out Roy’s use case. First, I tried a blurb from Buffer’s sales page. Here’s the original:
“Grow your audience on social and beyond
Buffer helps you build an audience organically. We’re a values-driven company that provides affordable, intuitive marketing tools for ambitious people and teams.”
I prompted Claude to “rewrite this text, framing it with the AIDA (attention, interest, desire, action) formula.”
The output was a bit jargony. But it’s a great way to see new options, especially when your brain is stuck on one version.
Next up, I thought I’d see what Claude could do with a press release for a product launch. I found a recent one from HubSpot.
Here’s part of Claude’s rewrite:
I wouldn’t publish it as is, but I really like how it highlights pain points in the “Interest” section and immediately soothed them in the “Desire” copy.
I also really liked how it separated and labeled each portion of the framework so it’s easy to review.
How to use this tip
You can plug just about any copy into a gen AI chatbot and ask it for a rework. I tried some headlines and a few social media posts, and Claude had no problem with them.
I wouldn’t expect ready-to-publish content, though. There was a lot of marketing-y phrasing I’d remove. It did do a good job of combining challenges with solutions, though. I’ll definitely use this as a co-writing sparring partner to see if it can suggest new ways to organize my content.
5 more AI uses for marketing we didn’t test
There are several more great ideas embedded in the replies to Amanda’s post. Many required a lot more time or a paid tool I didn’t have access to.
All that said, these are some of the best ideas from the comments I plan to try soon.
6. Create blog posts from voice notes
Lade Falobi, head of marketing at SocialKit, said she records “voice notes with my rambling thoughts and then asks AI to turn them into posts that actually sound like me.”
I can’t tell you how many times I’ve had the perfect thought, only to lose it when my fingers hit the keyboard. Or came up with the solution for a sticky problem while walking the dog and struggled to remember it later. Ladi’s tip solves that.
It also sounds like a great way to keep your personal tone in what you write.
7. Surface gaps in under-performing posts
Ellen Kaross, global marketing communications manager at Certn, said she finds informational gaps to boost underperforming articles using AI.
For example, she’ll prompt, “If you were a _________ (ICP), what outstanding questions would you have about X topic after reading this blog post?” Then, she adds a Q&A to fill in the gaps AI identifies.
Updating existing content will be a critical trend for content marketers this year. You could take it a step further and ask AI to optimize an old blog post for answer engines like Perplexity.
8. Break creativity ruts by asking for an opposite strategy
Rolf Mistelbacher, CEO and founder of Fresh van Root, “juices creativity” by “asking AI to do the exact opposite of what I expect it to do.”
He gives an example prompt: “Create the complete opposite car ad of what I would expect from a car ad for Jaguar.”
A common thread throughout these tips is to use AI to break out of creativity ruts. Rolf’s idea does exactly that.
Here’s another twist. Ask AI to write the opposite of a competitor’s website copy. Then, use it as a springboard to create a standout sales sheet or website that differentiates you from them.
9. Simplify complex concepts
Jesse Friedman, a fractional product, brand, and content marketer, runs copy through AI to make sure he’s made complicated topics approachable. “I mostly use it to verify if I’ve summarized complex concepts correctly,” he wrote. “We all know of the risks of hallucination when you ask open-ended questions, but I find it’s very good at critiquing and suggesting improvements.”
One of the most important jobs of a content marketer is to translate information provided by an expert to an audience of non-experts. I like having an AI double-check to find better ways of doing that.
10. Pull content ideas from meeting recordings
Jason Vu Nguyen, co-founder of Bloomstory, set up automation to extract content ideas from video calls. In his words, “I’ve set an automation workflow to record video meetings, have the transcript analyzed by ChatGPT/Claude to pull out any potential content ideas and put it in a well-structured document and uploaded to a Google drive (I can also send this to a slack channel) for me or my team to review and pick the ideas that resonate most with us.”
He added that they then write the posts internally and use Claude to verify tone and punctuation.
This would be an absolute game-changer for almost every marketer. Imagine setting up a regular cadence of calls with customers, the sales team, leadership, or SMEs. Instead of furiously taking notes, you can just talk free-form and, when you’re done, have a list of ready-made topic ideas waiting for you.
Play is the key to learning AI for marketing
Play has always been an integral part of learning. That stuck out to me as I worked through the ideas from Amanda’s post.
The results I shared here are only a portion of the things I tried. In every case, I plugged different prompts into multiple tools to see what would happen. I’ve done it many times since and continue to learn new ways of integrating AI into my daily work.
The point is, don’t restrict yourself to achieving one specific outcome. Set aside time to chase random theories and head down rabbit holes. Make note of what works, and iterate on what doesn’t. And if you’re feeling generous, share it with the rest of us so we can learn, too.