The possible applications of generative AI in marketing seem endless. Getting started feels daunting for many firms. Here are 7 ways to use Chat GPT in professional services marketing right now.
Since launching in November 2022, Chat GPT has taken the world by storm. Some experts warn of AI as an existential threat to society. Pundits predict massive job destruction almost daily. Others see it as the next wave of frothy productivity-enhancing technology – juicing start-up valuations and catalyzing a new bull market.
While any of these statements could be true, they’re not very useful right now. Yes, generative AI will change the labor market as all technologies do. Yes, it will unlock organizational productivity. But, beyond just streamlining time-tracking, what should you be doing with generative AI right now in your professional services firm?
More specifically, what could you be doing with Chat GPT in professional services marketing to make your efforts more effective and efficient? We’ve been experimenting with this, and other AI-related technologies, for the last 6 months. These are the 7 ways we’ve found the technology most useful:
- Writing More Elegant Research Questions
- Identifying Different Angles on a Topic
- Testing a Topical Outline
- Performing Mindless Tasks
- Shortening Content
- Writing Metadata for SEO
- Writing Promotional Content
Let’s look at each use case a little more closely.
#1 – Writing More Elegant Research Questions
Thought leadership is the engine of most professional services marketing programs. And research is the fuel for that engine. One of the most difficult tasks of research is getting the wording “just right” on a research question. Generally, firms know conceptually what they want to ask. But writing a prompt in plain English that virtually anyone can understand can be tricky.
Chat GPT is a useful tool to simplify and reword research questions. Take this common, slightly awkward question on revenue growth used by one of our research partners – “What do you estimate as the percentage of change in your firm’s revenue so far this year compared to the same time period a year ago?”
A simple request to make the question shorter and more elegant, returned this: “How much has your firm’s revenue changed this year compared to the same period last year, expressed as a percentage?”
Note how the edited version is a little easier to understand, but an even shorter version would leave the survey respondent lost:
#2 – Identifying Different Angles on a Topic
Positioning your firm as a thought leader on a pressing client issue requires building a “body of work” on that topic. Your job is to fall in love with your clients’ problems and make it your permanent quest to seek better ways to solve them.
Yet, at times, marketers feel like they hit a topical wall. “We’ve been writing about supply chain disruption for 3 years straight. How many more ways can talk about ‘supply chain reconciliation?’” Sound familiar?
Chat GPT is a useful tool to explore topics from different angles. On this topic, Chat GPT returned a handful of excellent related topics … streamlining supply chains, optimizing inventory management, mitigating supply chain risks, and achieving seamless operations … within seconds.
Note how some of these options aren’t very good (unifying supply chain operations?), and need to be ignored:
#3 – Testing a Topical Outline
In this regard, Chat GPT is the greatest gift ever to thought leadership marketers.
At Rattleback, we’ve found the best way to develop a compelling argument in your thought leadership is to start with a structured outline. An effective outline describes a client’s business challenge, the conventional wisdom on how to solve it, the short comings of that approach, and your new, better solution.
As a large language model, Chat GPT is essentially consuming all the collective information on which it is trained. This includes pretty much everything on the Internet up through 2021 and every new query made daily. When you ask Chat GPT to outline a topic, it’s essentially taking all that collective information and converging on what it sees as the conventional solution.
A simple request to outline a topic is a loose summary of conventional wisdom. A comparison of that outline with one you’ve developed in close partnership with a subject matter expert (but not shared with Chat GPT) will help you see the novelty (or lack thereof) of your firm’s proposed solution.
#4 – Performing Mindless Tasks
Just yesterday I used Chat GPT to alphabetize a list of 25 software companies I had assembled for use in a client’s research study. A task that might’ve taken a couple of minutes took seconds. Pile all your mindless tasks together and you’ll free up days or weeks of time by the end of the year.
#5 – Shortening Content
“I would’ve written something shorter if I had more time.”
You need the intro on that website service page to be no more than 10 words, but you have 25? You need that headline to be no more than 25 characters, but you have 100? We’ve found Chat GPT to be incredibly useful in generating quick, light edits on sections of content. Best of all you can ask it to quickly generate a few options so you can make the final call on what to keep and what to cut.
#6 – Writing Metadata for SEO
Increasingly, we’re found Chat GPT to be useful in SEO. Specifically, it’s good at writing the metadata for an article based on the article text and our recommended target keyword. If you give it clear and specific prompts it can write your full meta description, suggest a search-friendly URL, and draft a page title. You can even give it character limitations for each asset.
Keep in mind it doesn’t have access to any real-time search data. So, it has no way of knowing how competitive or difficult any search phrase is or if the meta data it’s providing is meaningfully tied to your business objectives.
#7 – Writing Promotional Content
Clients lean on us to help them develop a compelling point-of-view on how to solve a pressing client challenge. Often, this work is packaged as an article, research report, or simple blog post. The value in the relationship is helping them sharpen their thinking, develop their argument, and make it as coherent and compelling as possible.
That said, once it’s time to hit publish, a whole slew of activity kicks into motion. Social posts, email marketing activities, PowerPoint slides, and other supporting resources need to hit the market quickly. We’ve found that Chat GPT can be quite useful to assist in drafting much of this copy.
From generating options for email subject lines to writing Google ad headlines and body copy, Chat GPT is adept at developing first drafts for very tactical content when given a human-written foundational piece of content to draw from. Of course, all the content it generates needs edited by a human. But, frequently, it’s a great starting point.
Just for a little fun, I had Chat GPT write a limerick about this article — check it out…it’s not too bad:
In marketing, a tool they found,
Chat GPT, so smart and renowned,
With its help, they explore,
New angles galore,
Professional services’ insights, unbound!
3 ways definitely not to use it
There are surely hundreds more ways to use tools like Chat GPT in professional service marketing to make your firm’s marketing approach more effective or efficient.
That said, before suggesting first steps, it’s important to point out a few ways we would NOT suggest using Chat GPT or any other tools like it. Do not use it to:
- Perform Any Meaningful Secondary Research – The system simply throws off too much false information. We’ve tested it to perform secondary research and it frequently makes up people, sources, or both. Worse yet, it names and quotes sources so quickly and easily when asked that it evokes a false sense of confidence in its responses.
- Develop an Original Outline or Content Draft – While many consumer marketers are using Chat GPT, and tools like it, to develop initial outlines or copy drafts we simply don’t see how it could ever be up to the task in this regard. As mentioned previously, the very nature of the tool is such that it’s designed to converge on conventional wisdom. This is the last thing you’re looking to do in developing thought leadership.
- Make Sense of Primary Research – Just because you’re having what feels like a real conversation doesn’t mean that the system really understands what it’s reading or saying. As a large language model, it’s essentially very good at understanding how words and phrases most logically and frequently fit together. It’s modeling language based on things written by real humans. As such, it’s not really “thinking” about what it’s saying … at least, not yet.
Now is a Good Time to Test-and-Learn
So, what should you do right now? Take a test-and-learn approach. Develop your “prompt-writing” skills. Try some of the tasks we’ve suggested above. See what works for your marketing team and what doesn’t. Need a little help? Give us a call.