Introduction to AI
In recent years, buzz about artificial intelligence has been nearly impossible to miss in news articles and social media threads as discussions about its powerful applications ballooned. In the context of marketing, there have been headlines about how AI can be used to make decisions about product pricing, identify customer trends, perform detailed data analyses, generate photos, draft written content, and more. The subset of machine learning related to creating content—including images, text, audio—is called generative AI, and this article will be focused on the applications of generative AI in marketing.
In a recent report, Botco.ai asserts that “Generative AI (GenAI) is fundamentally changing what it means to be a marketer.” The report details the swift adoption of GenAI in marketing, detailing how—in a survey of 1,000 marketing professionals—approximately two-thirds utilize GenAI to initiate projects (e.g. brainstorming, initial drafts, outlines) and nearly one-half use AI to create final products. The trend was especially prevalent for B2B companies, which had a 78% GenAI adoption rate compared to 65% for B2C companies.
My background
As a marketing professional specializing in social media and content development, I was eager to experiment with how I could incorporate artificial intelligence into my workflow to not only improve my efficiency, but also the quality and effectiveness of my work. My core responsibilities include researching and writing copy for blog articles, creating social media graphics and posts, and developing content for web pages and other marketing assets. These are tasks that GenAI can undoubtedly perform, perhaps even faster than me. However, being human allows me to add emotional components, edit copy that sounds awkward, and take projects in more of an innovative direction (rather than one based on previous data and learned patterns). So, the question becomes, how can I—and other marketers like me—create the best outputs in the shortest amount of time by combining experience and creativity with AI capabilities?
Thus far, I’ve been in the group of B2B marketers who use AI in early project phases to spark ideas or act as a thought partner. Primarily, I use Chat Generative Pre-Trained Transformer (ChatGPT), a GenAI tool developed by Open AI that answers questions and prompts with text responses. In this blog article, I’ll describe three ways to use ChatGPT for content development (research, short-form posts, and data analysis) and offer insight into optimizing prompts and detecting patterns. I will also cover potential limitations with using AI.
Research
One application of GenAI for content development is to help bring yourself up-to-speed on industry topics. For example, as a marketer in the technology sector, you may be challenged with writing detailed descriptions about complex subjects and products that you have not previously used. To appeal to your target audience, such as IT decisionmakers, you need to be informed on the subject and ensure that your writing is specific and accurate. By asking ChatGPT for a summary on the subject, you can quickly get a sense of what the product or service is as well as what problems it solves. You can also ask follow-up questions about areas that you would like more clarification on. This is more efficient than sifting through articles from a search engine, as you may find repetitive information and need to pick out the information that addresses your specific question.
After you do a ChatGPT search, however, you may still want to read the top results from a search engine. This confirms the accuracy of the information, provides you with alternative viewpoints, and includes the most up-to-date information. While powerful, ChatGPT has its limitations. When it comes to research, one of the greatest criticisms of the tool is that its responses may be biased due to the data sets it was trained on. One Scribbr article says, “If biased inputs are what determines the pool of knowledge the chatbot refers to, chances are that biased outputs will result, particularly in regards to how it responds to certain topics or the language it uses.” Moreover, until recently, ChatGPT was only trained with data up to September 2021, so its outputs did not reflect most recent news updates and discoveries. All in all, because ChatGPT can provide an abundance of customized information in seconds, it is helpful for initial research, but you may also want to do additional internet searches to confirm your learnings.
Short-form posts
A second way to use GenAI for content development is as a starting place for short-form social media posts. Creating content for social media allows you to exercise creativity and experiment with trends, making it a fun task. However, it can also be time consuming—especially when you experience the “blocks” in your creative process that all marketers inevitably face. Perhaps you are struggling to craft a snappy title for your campaign, or maybe you’re stuck on where to start writing. For a cadence of posting once, or even multiple times a week, the number of hours you spend on creating content for social media quickly adds up.
This is where artificial intelligence tools such as ChatGPT come in. They can inspire different angles for you to adopt when building out campaigns, minimizing the amount of time you spend brainstorming. For example, you can ask ChatGPT to create potential posts based on content from your organization’s website and leverage the emojis, hashtags, and themes that are generated. You can also try including specific topics or target audiences in your prompt. Alternatively, when planning thought leadership content, GenAI tools can help you determine if articles from other reputable sources are suitable for your brand to repost. Simply copy and paste the article into your prompt and add the command for your chosen tool (e.g. ChatGPT) to summarize the information given. A quick scan of the summary should allow you to determine if the article is relevant.
When using ChatGPT to create content for social media, it is helpful to provide context and be specific in your prompt. You might clarify the platform you are posting to, the goal of your post, and other pertinent details. For example, you might type:
You are a social media expert. Please write five potential posts for LinkedIn introducing *example topic.* Please also tie it back to my company’s *example service* and target *insert audience.* Here is more information about my organization: *insert copy*
From there, you can work on improvements such as shortening the copy or adding a catchier headline. You can also prompt the tool to provide an entirely new set of posts. One thing to be aware of is that this prompt (and similar variations of it) may yield posts with the same format. Usually, there is a summative headline with emojis surrounding it and a call to action as the last line. However, contrast in the structure and format of your posts helps to create a more dynamic, engaging content calendar, so you may want to invest time in making edits and infusing creativity. The call to action can also seem forced or cheesy, so you can use your own judgement to edit or remove it. For example, when resharing certain thought leadership articles, it may be inauthentic, or even inappropriate, to have an aggressive call to action. Additionally, to avoid creating the same post as someone else who used a similar prompt, you can add details to and customize generated posts. Even when you spend time on these changes, including ChatGPT in your workflow should still speed up your social media content creation responsibilities.
Data analysis
A third way to leverage AI for content creation is for analyzing campaign performance, which helps inform future marketing and content strategies. While directly applying formulas and generating an Excel file is outside the scope of ChatGPT’s ability, the tool can propose metrics for you to further investigate to determine the effectiveness of a campaign. For example, you might type:
You are an expert social media marketing manager trying to evaluate the growth in followers on your *specific platform* page since you started posting once a week. Which metrics do you use to show the change in growth each month?
You can customize the prompt to be “the best reports and statistics” to pull from a particular platform (e.g. LinkedIn or Google Analytics). After you build the report, you can utilize ChatGPT to help identify mistakes in your formulas. Type the formula and describe the mistake in the prompt, and ChatGPT can provide you with a list of potential errors. You can add additional details such as specific rows where the error occurs or your hypotheses of factors impacting the results. With the campaign performance reports, you can evaluate whether your current strategy is working and make decisions about future strategies.
TSRM Group
Since 2015, TSRM Group has led with the idea that there is a better way to deliver sales and marketing services to the channel. Take the best things about a traditional marketing firm—the ability to respond quickly with the right team to meet client needs—combined with highly trained, process-oriented sales professionals, and a razor-sharp focus on IT.
Deeply expert in both IT and marketing, our focus is on crafting messages that both educate and move people, and then delivering them efficiently. Our clients are from a wide swath of the North American tech sector—IT solution providers, manufacturers, and distributors. We are committed to providing each client with subject matter expertise and battle-tested tools and processes that enable them to build a robust and thriving demand generation and customer acquisition engine.
0 Comments