ChatGPT is truly nothing short of amazing. The program generates code, solves crosswords, and spits out alternate endings to The Sopranos. Tell it what to do, and you’ll get nearly exactly what you want. Naturally, this is giving entire industries pause, including content marketing.
But AI in content marketing has been around for quite some time — so what is it about Open AI’s ChatGPT that’s causing such a stir? As it turns out, ChatGPT quite literally takes AI to a new level.
A brief overview of AI in content marketing
The use of artificial intelligence in content marketing is nothing new. Marketers have used these smart technologies for years to assist with rote yet essential tasks. With once-lengthy and complex tasks taking minutes instead of hours or days, teams are free to apply their brainpower to creative thinking and problem-solving.
Take the Associated Press (AP), for example. The news organization put AI to use in 2014, producing business news stories regarding corporate earnings.
Importantly, the AP’s AI wasn’t out to win Pulitzers. Its stories on business news earnings were quite formulaic, swapping out names and figures with each new report. The changes from story to story weren’t drastic, with few details edited in each article. This sets AP’s AI use apart from a publication like tech news website CNET, whose AI-written articles left a lot to be desired – more on that later.
With such small tasks in the rearview mirror, reporters could focus on breaking news and in-depth investigations. The payoff was clear: The use of AI saved valuable time and, therefore, money.
This application isn’t just limited to journalism. AI tools have a place throughout the marketing industry, assisting folks with a seemingly endless list of crucial tasks. Just a handful of examples include:
- Data analysis
- Pitch generation
- Content topic generation
- Content outline generation
- Consumer insight analysis
- Predictive analysis
Why is ChatGPT different?
What moves ChatGPT out of the sidelines and into the limelight is its ability to understand a question’s intent and respond in kind. You don’t need to simplify the commands or put your query into a certain format. ChatGPT can understand what you’re asking and reply with remarkable, human-like clarity. We’ve come a long way from asking Jeeves.
ChatGPT is built on what’s called a large language model, or LLM for short. An LLM learns by processing more and more data, building upon what it knew before. The more data, the more complex and comprehensive the responses.
Early LLMs predicted the next word in a sentence. ChatGPT — and the LLM it’s built upon, GPT 3 — goes further because it was trained to respond like a human would, not just to finish a sentence. That was achieved using what’s called Reinforcement Learning with Human Feedback. This method pulls from tons of internet content, from blog entries to social media posts, to help the AI expand its capabilities.
4 ways content writing agencies can leverage ChatGPT
AI isn’t the future — it’s here. Content writers need to learn how to leverage it to keep up with the times and stay relevant. Start with these four ways to leverage AI in content marketing, with a practical example for each.
1. Help generate ideas
Feeling stuck for inspiration? You can turn to ChatGPT.
In this blog, we’ll use the example of topic ideas about content marketing. ChatGPT was prompted to identify three long-tail keywords on the topic. The ideas that came back — content marketing strategy for small businesses, B2B content marketing best practices, and measuring the ROI of content marketing — are all viable topics that can be of value to a content marketing agency’s audience. With these ideas and a general direction in mind, writers can conduct independent research to determine the keyword’s potential, find other related keywords, and further refine the topic in a way that works best for their audience.
2. Speed up drafting
Good writing takes time, but not everyone has that luxury. ChatGPT can produce snippets that can be worked into larger drafts and rewritten to fit your needs. Impressively, ChatGPT content is fully original and passes plagiarism checks.
In this example, GPT was prompted to share two ways to measure the ROI of content marketing. The result was:
“One way to measure the ROI of content marketing is by tracking the number of leads generated from the content. This can be done by including unique tracking links in content, which lead to landing pages where users can provide their contact information. By measuring the number of leads generated from the content and comparing it to the cost of creating and promoting the content, you can calculate the ROI.
Another way to measure the ROI of content marketing is by analyzing the increased website traffic and engagement. This can be done by tracking metrics such as page views, bounce rate, and time on site. By comparing the metrics before and after the content marketing campaign, you can see if there has been an increase in website engagement and calculate the ROI based on the cost of the campaign.”
3. Help write outlines
Outlines, a key part of the SEO content writing process, ensure that the content is written in ways that give it the biggest chance to compete. Outlining, though, is a painstaking process — it involves in-depth competitor scoping, topic research, and balancing company offerings and values, all without veering away from the framework that works best for SEO. With ChatGPT, you can generate an initial draft of the outline that can inform your final version.
When prompted to write an outline for a blog about ways to measure the ROI of content marketing, ChatGPT produced:
Definition of content marketing
Explanation of why measuring the ROI of content marketing is important
Setting goals and objectives
Identifying specific business objectives for your content marketing
Setting measurable goals that align with those objectives
Overview of the key metrics to track for content marketing ROI
Discussion of tools and technologies that can be used to track metrics
Analyzing the data
How to use data to understand the effectiveness of your content marketing
Techniques for identifying areas of improvement and opportunities for optimization
Measuring the impact on sales
Techniques for measuring the impact of content marketing on sales and revenue
Examples of how to use data to demonstrate the ROI of content marketing to stakeholders
Recap of the importance of measuring the ROI of content marketing
Call to action for readers to start tracking their own content marketing metrics
4. Help with repetitive work
Content marketing involves lots of copy besides blogs, like email copy, tweets, and meta descriptions, to name a few examples. These short descriptions may seem simple to write, but there’s an art to succinctly summing up 1,000 or more words in just a few characters. ChatGPT can help.
Turning to the content marketing ROI example once again, ChatGPT was prompted to produce social media captions to promote a blog on the topic. (The hashtags came unprompted, although we highly recommend you review hashtags manually before sharing them.)
“Want to know how to measure the ROI of your content marketing efforts? Check out our latest blog post for tips and strategies on setting goals, tracking metrics, and analyzing data to optimize your content marketing strategy! #contentmarketing #ROI #digitalmarketing
“Want to see the return on your content marketing investment? Check out our latest blog post on measuring #contentmarketing #ROI and see how to track metrics and analyze data to improve your strategy! #digitalmarketing”
“Measuring the ROI of your content marketing efforts is crucial for understanding the impact of your strategy and identifying areas for improvement. In our latest blog post, we share tips on setting goals, tracking metrics, and analyzing data to optimize your content marketing strategy. Read now and take your content marketing to the next level. #contentmarketing #ROI #digitalmarketing”
Don’t like the options? Just press “regenerate response,” and you’ll get a fresh set of original ideas. Here’s an example of what a regenerated response could look like.
“Want to know how to measure the ROI of your content marketing efforts? Check out our latest blog post for tips and tricks on setting goals, tracking metrics, analyzing data, and measuring the impact on sales. #contentmarketing #ROI”
“Maximize the return on your #contentmarketing investment with our latest blog post on measuring #ROI. Learn about setting goals, tracking metrics, and analyzing data. #digitalmarketing #marketingstrategy”
“Are you struggling to measure the ROI of your content marketing efforts? In our latest blog post, we share practical tips and techniques for setting goals, tracking metrics, analyzing data, and measuring the impact on sales. Whether you’re a content marketer, a business owner, or a marketer, this post is a must-read! #contentmarketing #ROI #digitalmarketing #marketingstrategy”
6 reasons why ChatGPT still needs human help
For all its incredible applications, ChatGPT can’t walk and talk like Rosie from The Jetsons. The human touch is necessary for successfully implementing this tool’s suggestions. Here’s why.
1. Sure, it sounds human — but not human enough
The writing quality produced by ChatGPT is, well, pretty basic. There’s a certain rhythm to content produced en masse that experienced writers (like our staff writers) know how to spot and course-correct. To the untrained eye, there may not be a difference at all. But readers can recognize a rote, droll tone, and it can turn them off. Taking ChatGPT’s initial suggestions can certainly kickstart the writing journey, but experienced writers need to take that content the rest of the way.
2. There are greater implications for brand voice
Brand voice is crucial for distinguishing your company from competitors and positioning your company as authentic and trustworthy. ChatGPT’s content reads quite neutrally — and that’s not always a good thing. To those of us invested in brand voice and creating distinctive copy, the content generated by ChatGPT can detract from the brand differentiation that marketers work tirelessly to develop and maintain across all marketing channels. And if you want to stand out, sounding the same as everyone else can be quite a detriment.
3. “Smart” doesn’t mean it’s error-proof
It recently came out that tech and electronics publication CNET quietly used AI to author articles on its site. Aside from the fact that CNET did not disclose the articles were AI-generated, it didn’t take long to discover that their AI-written articles were riddled with errors.
Remember how ChatGPT learns what it knows: It’s fed massive amounts of data from the internet. Any content marketer can tell you how much junk clogs the pipes. Those errors, miscalculations, and false perceptions have apparently made their way into ChatGPT. Whether a tiny error in word choice or an article that’s totally off-base, those inaccuracies can damage your business and brand. Experienced fact-checkers are still a must to verify the information making its way to publication.
4. It won’t always know the latest information
The ChatGPT disclaimer states that “[ChatGPT] has limited knowledge of world and events after 2021.” It may only leave out a single calendar year, but think about how much evolved in 2022. This can have significant consequences for content quality and accuracy.
This is particularly true for rapidly-evolving industries like cannabis or psychedelics, and for industries like tech, where significant innovation made in a year’s time can entirely change the premise of an article. If not caught by an experienced content writer or subject matter expert in cannabis, tech, or any other of these aforementioned industries, you run the risk of publishing outdated content that makes your brand look outdated too.
5. We’re still learning how it can affect search engine results
There’s no shortage of reports coming from SEOs who generated articles using ChatGPT and watched the content rank at the top of search engine results pages (SERPs) in days. With these observations, though, comes murkiness around the details that shape the why and how of this success story.
We don’t always know which niche the article was published for, the keyword they selected (or that ChatGPT suggested), how competitive you need to be to rank for that keyword, if the articles are meeting searcher intent, and the like. We also don’t know how long that content will stay there. It’s certainly an exciting development that holds promise, but it’s not a home run yet.
6. It cannot replace meaningful and purposeful content
Finally, and most importantly, ChatGPT is not yet capable of understanding your intent. Why does your company need to write an article on content marketing ROI? Should your company write this article? What spin does it need to successfully work for your company, content writing strategy, and overall marketing strategy?
Only the brainpower of your marketers and advisers can take ChatGPT’s content from idea to reality. Even when the suggestions are technically sound and factually correct, people who know you and understand your business need to make the final decision about its practicality and usability.
Where does this newest advancement of AI in content marketing leave us?
There’s a lot of power in ChatGPT. The tool offers boundless opportunities to save time, analyze data, get inspiration, and generate information that content writers can leverage to go far. ChatGPT frees up time for professionals so they can achieve greater things in content writing, content marketing, and marketing and advertising as a whole.
What is concerning, though, are SEOs and marketing professionals calling for the end of the human touch. As we’ve demonstrated here, ChatGPT has limits that only experienced content writers and trusted marketers can work around.
There are also tremendous ethical implications that this article hasn’t touched. The racial bias baked into these systems is a serious problem that needs to be addressed. There are opportunists who see AI as a way to displace workers. The list goes on. On top of all the practical reasons why ChatGPT can’t replace quality content writers just yet, serious conversations need to be had — and advanced — as adoption becomes more widespread.
But we’re only at the beginning of this exciting new chapter for AI in content marketing. And as more people experiment with its possibilities — and indeed, push its limits — we can continue to reach for the stars.