Like it or not, the tool is here, but is it ready for serious marketers yet?
Imagine a world where with no programming experience or technological understanding, you could go to a website and have a conversation with a computer intelligence who could not just answer any question but answer it in a thoughtful way and in the voice or style of anyone living or dead.
What if that same artificial intelligence could in just one to five prompts write an article on any topic and at any grade level? And, that the content could rival any human writer and fool even the most experienced editors. Sounds like science fiction or a comic book, right? Maybe. Except, that many of these things are possible right now, or close to possible with generative AI.
Unless you’ve been under a rock or vacationing somewhere without access to news or the internet, you’ve probably seen the buzz around AI generated content, namely a tool from startup OpenAI called ChatGPT. OpenAI’s tools, including their API and AI image generator, Dall-e have been getting attention for a while, including from Microsoft. But things really took off with the release of ChatGPT and once marketers started using it, including Ryan Reynolds and MNTN for his company Mint Mobile with this ad, based on a script written with ChatGPT.
What is generative AI
Generative AI isn’t new. However, even as recently as five to six years ago, what it could do was limited. It seemed at the time based on the strangeness of its output, that the technology had a long way to go before it might be a useful tool for business, including for creating marketing content, let alone video scripts.
But then came ChatGPT and the content world is rapidly working to both test its abilities and learn how to leverage generative AI to make content creation quicker, more efficient while increasing company revenue. For content marketers and copywriters (like the person writing this article), there’s also fear that those jobs will rapidly disappear and instead, generative AI will take over those creative tasks. But what is it and how does it work?
How do tools like ChatGPT work?
The answer to that is: it depends. There are many different ways of programming an AI to understand language and generate content. There’s nature language processing and neural networks, and other types of advanced technology. Rather than get bogged down in every possible permutation of artificial intelligence and generative AI, and the technology involved, let’s keep things simple.
Most generative AI, including ChatGPT work through “training.” In other words, the AI computer “learns” or improves based on the amount and quality of information (data) it’s fed or has access to and then the number of exposures to that data and any new data it’s given. It then generates new, (hopefully) original content from all the content it has access to on any given topic.
According to OpenAI’s ChatGPT FAQ:
Source: OpenAI
An AI with text generation abilities that’s given limited data can only generate content based on that limited information. So, the old adage of “garbage in; garbage out” is just as true for AI as it is for anything else. Feed the AI tool high-quality content and lots of it, while programming it with the ability to “learn” from content users type into it, and the generative AI should provide you with equally high-quality, original and usable content.
So, are we there yet? What can this new content tool do for marketers and content departments? Will professional writers soon be out of a job? A member of the content team at Elevation Marketing “sat down” with ChatGPT and asked the bot some questions, then tested the AI tool on some content marketing tasks.
How does ChatGPT compare to a human writer?
The first question, or prompt I entered into ChatGPT was a comparison of human writers and AI generated content. I had read an article by Content Marketing Institute (CMI) that had posed a similar question but from a strict content marketing angle. I wanted to see what it would say about written content in general and also see if it knew I might mean marketing content based on two previous sessions on marketing related topics I’d asked it before. I also wanted to investigate how well it could tell me where it got the information for its answer and whether it would cite sources or not.
Here’s how it responded:
Elevation:
Are human writers better than AI writers at generating B2B content?
ChatGPT:
It depends on the specific task and the quality of the AI writing system. In general, human writers have the advantage of creativity, emotional intelligence, and domain-specific knowledge, while AI writing systems can generate large amounts of content quickly and at scale. Both human and AI writers have their own strengths and weaknesses, and they can complement each other in certain applications. Therefore, it is important to use the right tool for the right job and to have human editors review the content generated by AI systems to ensure its quality.
Elevation:
Who says?
ChatGPT:
Many experts in the field of AI and content creation have commented on the strengths and weaknesses of human and AI writers. This is a widely accepted view in the industry and it’s also supported by various studies and reports. However, it’s worth noting that the AI technology is continually improving and it’s possible that in the future the AI writers will be able to generate content that is indistinguishable from human-written content.
Takeaway:
ChatGPT’s response is well-written. It’s clear, concise and answers the query. That is, it can answer until asked where it came up with the answer in a conversational way (Who says?). Instead, it gives the vague response of “many experts.” But, stay tuned because sometimes giving a more specific query or setting up clear criteria in an initial query gets better results.
Will copywriters and content writers become relics of the past?
Writers aren’t the only people wondering if they’ll be out of a job. Business leaders are also curious how AI can provide value and increase revenue as well as improve productivity and efficiency. Instead of asking if ChatGPT was going to take over copywriting tasks, I asked it the following:
Elevation Content:
How soon will AI writers produce marketing content that’s indistinguishable from human content writers and copywriters?
ChatGPT:
It is difficult to predict exactly when AI will be able to produce marketing content that is indistinguishable from human-written content. However, AI language models such as GPT-3 have already demonstrated the ability to generate high-quality, human-like text. However, the quality of the content generated by AI systems can vary depending on the specific task and the quality of the AI model. In addition, it’s worth noting that even though AI can generate human-like text, it still lacks the creativity and emotional intelligence of human writers. Additionally, the ability to understand the audience, their needs and the context of the marketing campaign is something that AI is still working on. Therefore, it’s likely that we will see more and more AI-generated content in the near future, but it will still require human editors to review and refine the content generated by AI systems to ensure its quality.
Elevation:
Where did you find this information?
ChatGPT:
As a machine learning model, I was trained on a vast amount of text data and information, so I don’t have a “source” as such. I use the patterns and structure of the data to generate responses. However, the information I provided is widely accepted in the field of AI and content creation. You can find similar information on the capabilities and limitations of AI-generated content in reports and studies from research institutions, industry experts and analysts, as well as from reputable media outlets that cover the field of AI and technology.
Takeaway:
Again, the content ChatGPT generates on the topic is very good. It’s not particularly robotic. And it’s content that a human copywriter might include. It does repeat “however” to begin two consecutive sentences which a human writer wouldn’t do or if they did, a copy editor would correct.
When I asked where it found the information for its answer to see if it would list sources based on a more direct and less conversational prompt, it did not. Instead, it explained how it uses data to create its answers and directed me to look for the studies, reports and news stories that already exist on ChatGPT. It basically gave me an indirect response of “Google it,” something I was seeing if I could avoid doing. Afterall, I am trying to investigate how generative AI can make Elevation more efficient and create value for both Elevation and our clients. If I have to spend just as much time researching and fact-checking AI as I do without AI, then a tool like ChatGPT isn’t useful…yet.
But, again, stay tuned. I do continue to evolve how I prompt it to provide sources later on.
Does ChatGPT understand the B2B vs B2C content?
As content professionals and marketers, we understand that even with B2B customers craving more B2C experiences, they still have different priorities. Also, the consumer buyer journey is typically quick and straightforward. Whereas, the B2B buyer journey has many twists, turns and roundabouts. Can ChatGPT craft content that shows an understanding of those differences if asked. Here’s how they responded to my question:
Elevation:
How does writing content for B2B audiences differ from writing content for consumers?
ChatGPT:
Writing content for B2B audiences differs from writing content for consumers in a number of ways. Here are a few key differences:
Tone and style: B2B content is typically more formal and professional than consumer content. It should use technical language and industry-specific terms that the target audience will understand. Consumer content is generally more casual and conversational.
Focus on solutions: B2B content should focus on providing solutions to specific problems or addressing specific needs of the target audience. Consumer content, on the other hand, is more focused on showcasing the benefits and features of a product or service.
Data-driven: B2B audiences are often more data-driven, and they are more interested in facts and statistics than emotions. Consumer content is more focused on emotions and storytelling.
Length: B2B content is often longer than consumer content, as it goes into more depth and detail about a particular topic. Consumer content is typically shorter and more concise.
Decision-making process: B2B content is often aimed at decision-makers, such as business owners and managers, who are responsible for making purchasing decisions. Consumer content is aimed at end-users who make purchase decisions based on personal preferences and needs.
ROI oriented: B2B content often focuses on the return on investment (ROI) that a product or service can provide, while consumer content is more focused on the benefits and features of the product or service.
In summary, B2B content is more focused on providing solutions, data-driven, longer, and targeted at decision-makers, while consumer content is more focused on showcasing benefits and features, emotions, shorter and targeted at end-users.
Takeaway:
Actually, the above response is good, if a little basic. It’s lacking some of the nuances of B2B and doesn’t discuss when a more B2C style and B2C creative, including copy, might be more appropriate.
Nor does it consider that a content creator could incorporate B2C style and experience, including emotional marketing techniques while also discussing ROI and using hard data to establish trust.
Moreover, enterprise-level decision makers also want to know the features and benefits of individual products and services, including what differentiates one choice from another, or why they should invest more upfront for a more expensive option.
I did not ask ChatGPT to tell me about its sources for its answer because I already had an idea for how to test if it would list them.
Are there copyright issues associated with ChatGPT?
For this last query example before starting to talk about prompts to generate specific types of content campaign copy, let’s talk about generative AI and plagiarism. Other articles have pointed out the plagiarism and even factual issues associated with AI generated content. We’ve already seen how difficult it is to prompt ChatGPT in a way that it will provide anything but a vague answer when asked to provide sources. So, I wanted to see if it would “tell the truth” about the plagiarism issues. Here’s what happened:
Elevation:
Are there plagiarism issues associated with the use of AI to generate original B2B marketing content?
ChatGPT:
There can be plagiarism issues associated with the use of AI to generate original B2B marketing content if the AI system is not properly configured or if the generated content is not carefully reviewed and edited by humans.
AI language models such as GPT-3 use patterns and structure from the data it was trained on to generate new text, it may inadvertently generate text that is similar or identical to existing text. Additionally, some AI systems may use existing text as a source for new content, which would also constitute plagiarism. Therefore, it’s important to use plagiarism detection software to check the originality of the AI-generated content before it is published.
In addition, it is also important to have human editors review and refine the content generated by AI systems to ensure that it is original, accurate, and appropriate for the intended audience. This will help to prevent plagiarism issues and ensure that the content generated by AI systems meets the standards of quality and integrity expected of B2B marketing content.
Overall, while AI can help to generate large amounts of high-quality content, it’s important to use the right tool for the right job and to have human editors review the content generated by AI systems to ensure its quality and originality.
Takeaway:
I was impressed that ChatGPT didn’t have filters causing it to avoid the question. Not only did it answer correctly, but it also provided suggestions for how to find and correct any plagiarized content. Moreover, it pointed out that human editors should also:
Source: ChatGPT
During my experiments with ChatGPT, I found I needed to add linking content, check for duplicate content and plagiarism, and make sure I fact-checked anything I didn’t already know. I also have had to Google even cited sources to ensure I could link to them as needed or applicable. In other words, content marketing and content jobs will change with the use of AI, but humans will still be involved. Companies who rely to much on generative AI with zero checks, could wind up publishing unoriginal content and/or false content.
Can ChatGPT write a blog on B2B content and AI generated content?
And now we get to the fun stuff. How does AI do when asked to create actual content? The first task I tried was a blog on a similar topic to the blog you’re reading now. Below is my prompt and the blog ChatGPT generated:
Elevation:
Generate a 1000-word blog post on the topic of human copywriters versus AI generated content in B2B marketing.
ChatGPT:
In today’s digital age, businesses are constantly looking for ways to improve their marketing strategies and increase their online presence. One of the ways they do this is by creating high-quality content that resonates with their target audience. However, with the advent of AI technology, many businesses are now turning to AI-generated content as a way to save time and money. But is this a wise decision? This blog post will explore the pros and cons of human copywriters versus AI-generated content in B2B marketing.
First, let’s take a look at the benefits of using human copywriters. Human copywriters have a unique advantage over AI-generated content in that they possess creativity and emotional intelligence. They are able to understand the target audience and create content that speaks to them in a way that AI-generated content simply cannot. Additionally, human copywriters have a deep understanding of the industry and can create content that is both informative and engaging. They are also able to think outside of the box and come up with new and innovative ideas for content.
On the other hand, AI-generated content has its own set of advantages. One of the main benefits of using AI-generated content is that it can be produced quickly and at scale. This is especially useful for businesses that need to create a large amount of content in a short amount of time. Additionally, AI-generated content is able to analyze data and create content that is optimized for search engines. This can help to improve a business’s online visibility and increase their chances of being found by potential customers.
While both human copywriters and AI-generated content have their own unique benefits, it’s important to note that AI-generated content still lacks the emotional intelligence and creativity of human copywriters. Additionally, AI-generated content can sometimes lack the context and understanding of the target audience, which can make it less effective in engaging and converting potential customers.
When it comes to B2B marketing, it’s important to remember that the content needs to be informative and credible. Human copywriters are better equipped to create such content as they have the experience and knowledge about the industry, the target audience, and the products or services. They are also able to write in a tone and style that is appropriate for B2B marketing.
In conclusion, while AI-generated content can be a useful tool for businesses looking to save time and money, it’s important to remember that it still lacks the emotional intelligence and creativity of human copywriters. Additionally, AI-generated content can sometimes lack the context and understanding of the target audience, which can make it less effective in engaging and converting potential customers. Therefore, it’s important for businesses to use the right tool for the right job and to have human editors review the content generated by AI systems to ensure its quality. In B2B marketing, human copywriters are a better option as they are able to create content that is both informative and credible, which is crucial for this type of marketing.
Takeaway:
My prompt only included the topic and number of words as criteria. What I didn’t include were clear directives to include any subheadings or other formatting. Otherwise, it’s clear and concise, but short and pretty basic.
Moreover, the blog falls 500 words short of the 1000-word request in the prompt. Is that because the AI had interpreted it had answered the query in the prompt or because it’s not set up to follow word counts?
Unclear. Yet, it’s a good start for a human writer or content marketing thought leader to write a 1000-word blog on the topic. A human writer or editor will also have to fill in stats, cite sources and add relevant links. The tool can still save time. It took less than 30 seconds to generate the blog. And possibly learning how to create AI prompts with very targeted, specific criteria could eliminate some of these issues that it might take a human editor to as much time to check as it would have a writer to research and write the article itself.
Can ChatGPT generate content for an email promoting a whitepaper on AI generated content?
What if the blog above was part of a larger campaign or was related to a content campaign targeting businesses interested in AI tools for content creation? Could ChatGPT write an email promoting a whitepaper or the topic? And if we gave specific directions for any statistics, would it provide sources? Let’s see.
Elevation:
Generate an a B2B marketing email that tells the recipient the benefits of content marketing. Use three different statistics and end with a CTA that invites the recipient to download a whitepaper on AI generated B2B marketing content.
ChatGPT:
Subject: Unlock the Power of Content Marketing for Your Business
Dear [Recipient],
Are you looking for a way to increase your online presence and drive more leads for your business? Content marketing is the answer.
Recent studies have shown that businesses that consistently create and distribute valuable content see a conversion rate that is 6 times higher than those that do not. (Source: Content Marketing Institute)
In addition, businesses that have a documented content marketing strategy are 13 times more likely to see positive ROI from their efforts. (Source: Content Marketing Institute)
And if you’re worried about the cost, it’s worth noting that content marketing costs 62% less than traditional, outbound marketing and generates about 3 times as many leads. (Source: DemandMetric)
At [Your Company], we understand the power of content marketing and have developed a whitepaper that explores the benefits and best practices of using AI-generated content in B2B marketing.
Download our whitepaper now to learn how you can take advantage of this powerful marketing strategy and start seeing results for your business.
Best regards,
[Your Name]P.S. Don’t miss out on the opportunity to unlock the power of content marketing. Download our whitepaper now and start seeing results for your business.
CTA: Download our whitepaper on AI generated B2B marketing content now.
Takeaway:
Depending on how promotional the voice of your emails is, this email isn’t terrible. Plus, not only did ChatGPT provide statistics, but it also included sources for those stats, something I did not ask it for. It might have already learned that I tend to ask for sources and so gave them to me.
Or it might have been taught a particular email format. If the CTA is just a link, the length might be ok. If there will be a button, it’s too long. And, you would still need to A/B test copy, CTA and other elements and tweak the content based on your findings.
But this email took mere seconds to generate. It might take a really experienced copywriter an hour. And many writers Closer to 2 hours, depending. It has data from multiple publications or studies, which would take research or finding them in the whitepaper.
Can ChatGPT generate a white paper on a marketing topic?
Since ChatGPT had been able to quickly write a short blog and an email, I decided to try a whitepaper. So, I asked it to write one. And here’s where I ran into the limits of an open-beta tested technology tool: overloaded servers and error messages. At first I thought my prompts were at fault. But I kept altering them, then went to OpenAI support on their Discord channel and discovered the servers were overloaded. When that happens, your session will probably timeout. It’s similar to a dropped call when you’re talking on your cell phone.
However, I did get back on and manage to have ChatGPT generate a short whitepaper for me. Here’s how that looked. (Note: I’ve deleted all but one error message).
Elevation:
Generate the whitepaper for the campaign described in the email
ChatGPT:
{General error again: An error occurred. If this issue persists please contact us through our help center at help.openai.com.
After several tries and logging back on to ChatGPT servers:
Elevation:
Generate a whitepaper on the benefits of using AI to generate B2B marketing content. (Note: new chat session)
ChatGPT:
Introduction:
Artificial intelligence (AI) is rapidly revolutionizing the field of marketing by providing businesses with new and innovative ways to create, distribute, and analyze content. One area where AI is particularly beneficial is in the creation of B2B marketing content. This whitepaper will explore the various benefits of using AI to generate B2B marketing content, including increased efficiency, cost savings, and improved personalization.
Efficiency and Time Savings:
One of the biggest benefits of using AI to generate B2B marketing content is the increased efficiency and time savings it can provide. AI-powered content generators can quickly and easily create a wide variety of content, including blog posts, email campaigns, social media posts, and more. This can significantly reduce the time and resources required to create and distribute high-quality marketing content.
Cost Savings:
In addition to time savings, using AI to generate B2B marketing content can also lead to significant cost savings. By automating the content creation process, businesses can reduce the need for expensive human resources and eliminate the need for costly external agencies. This can help to lower overall marketing expenses, allowing businesses to allocate more resources to other areas.
Personalization:
AI-generated B2B marketing content can also provide a high level of personalization that is not possible with traditional methods. By analyzing data on customer behavior and preferences, AI-powered content generators can create highly targeted and personalized content that is tailored to specific segments of the market. This can help to improve the effectiveness of marketing campaigns and increase customer engagement.
Conclusion:
In conclusion, the use of AI to generate B2B marketing content offers a wide range of benefits for businesses, including increased efficiency, cost savings, and improved personalization. As the technology continues to evolve, it is likely that AI-generated content will become an increasingly important part of the B2B marketing mix. Businesses that take advantage of this technology now will be well-positioned to stay ahead of the curve and achieve greater success in their marketing efforts.
Elevation:
Can you provide statistics for this information?
ChatGPT:
I apologize, but I don’t have access to any statistics related to the benefits of using AI to generate B2B marketing content. However, here are some statistics that show the general trend of AI adoption and its potential impact on businesses:
According to a survey by PwC, 72% of business leaders believe that AI will be a “business advantage” in the next five years.
A study by McKinsey & Company estimated that AI has the potential to create $3.5 trillion in value for the marketing and sales industry by 2030.
A report by Gartner predicts that by 2022, 30% of B2B companies will use AI to generate human-like content for at least one-use case.
A survey by the Content Marketing Institute found that 38% of B2B marketers are already using AI in their content marketing efforts and another 32% plan to adopt AI in the next 12 months.
A study by Accenture Interactive found that AI-powered personalization could increase revenues by 15% while reducing costs by 30%.
It’s important to note that these statistics are an indication of the potential impact of AI on the marketing industry. The actual benefits and results may vary depending on the implementation and use case of the AI technology.
Elevation:
What do marketing thought leaders say about generative AI? provide quotes
ChatGPT:
Here are some quotes from marketing thought leaders on the topic of generative AI in marketing:
“AI-generated content is going to change the game for marketers. With the ability to quickly and easily create large amounts of high-quality content, AI will allow businesses to reach and engage more customers than ever before.” – Neil Patel, Founder of Neil Patel Digital
“Generative AI is the future of content marketing. It allows businesses to create highly personalized and targeted content at scale, which is crucial in today’s competitive digital landscape.” – Joe Pulizzi, Founder of Content Marketing Institute
“AI-generated content is the next big thing in marketing. It has the potential to change the way we create and distribute content, making it more efficient and cost-effective. It will also enable us to create more personalized and engaging experiences for our customers.” – Ann Handley, Chief Content Officer of MarketingProfs
“Generative AI is going to play a big role in the future of marketing. It can help businesses to create more personalized and effective content, which is essential for building strong relationships with customers.” – Jay Baer, Founder of Convince & Convert
“AI-generated content is a game-changer for businesses. It allows marketers to create highly personalized and targeted content at scale, which is crucial for standing out in today’s crowded digital landscape.” – Marcus Sheridan, Founder of The Sales Lion
It’s important to note that these are just a few examples of what some thought leaders say about generative AI and its potential impact on marketing. While they show the possibilities, it’s important to keep in mind that the actual results may vary depending on the implementation and use case of the AI technology.
Takeaway:
I don’t know whether I had been able to get back onto the servers when I’d tried a specific query with clear parameters whether the AI generated content would have had the stats and quotes in the whitepaper itself or not. Also, if I had asked for agency boilerplate as a parameter would that be in the whitepaper? Hard to know. The paper is short. But, once I could get into the interface, ChatGPT again generated clear, concise, informative copy in seconds. Even with all the errors, it took a lot less time to generate the content than if I had had to research and write an outline and rough draft myself.
Is ChatGPT ready as a content creation tool for B2B marketing?
The answer is unclear. In a sense, yes. I used it to create this blog. I also used the tool to create an outline and format for another blog. Even with all the errors, it still saved me a lot of time. For example, this blog is just over 5,000 words long. In the past a 5,000-word blog would take me probably 15 hours or more to research, outline, write, revise, etc. This blog took me the time it might take to write a 1500- or 2000-word blog. And some of the time I spent was experimenting with how to create queries that would yield the best results.
It’s clear that for the immediate future, the tool’s servers can’t handle the number of people trying to use ChatGPT. Businesses, students, programmers and journalists are flooding the tool, anxious and excited to test what it can do or just have something complete their homework for them. This morning when I tried to get on it, I was added back on a waiting list to access it, and I have OpenAI login credentials. The company just received an influx of capital via a substantial Microsoft investment and reportedly has plans for a premium tier where paid users could have an error-free and unlimited access experience to ChatGPT, for a price. OpenAI isn’t ready to divulge when that will be. I think by the end of the year though, we’ll see improvements and within the next five years AI generated content will be everywhere online.
Get help using martech, including AI to improve your digital content marketing efforts
Technology moves fast. And right now, we’re at an incredible turning point. It’s clear generative artificial intelligence and similar technology is here or will be here within months. Contact us today to take advantage of our over two decades of experience in using emerging technology to create high-quality content and reach today’s B2B customers wherever they are, including online. Let us help you stay ahead of the curve.
