When ChatGPT roared into the market in 2022, the communications world was abuzz about what it meant. It was the biggest change since typewriters moved to word processors 35 years ago—but whether it would help or hurt the PR industry was an open question.
It was also a question I was happy to let others figure out. Working with an AI client meant our firm had to adjust the media strategy, but outside of that, I mostly sat back and watched.
Fast forward a few years, and the lay of the land is starting to become more clear. Firms are using AI to drive more value for content via LLMs, some are creating entire operational systems based upon AI, while others are using it to accelerate their speed of performance.
Our firm is using AI for all three areas. But the story to focus on is how we used this not-so-new-anymore technology to solve a scaling problem that many PR firms face: how to serve a sizable client without over-investing in people whose jobs are reliant on the whims of a single client.
Preparing for Growth Opportunities
Some public relations firms have a single anchor client on which the company may live or die. We’re past that stage, but definitely have some clients whose revenue is larger than others. Earlier this year, one of those clients hinted at a big opportunity. They were pushing for some new business that would take us from “busy” to “very busy” with them. (Their words, not mine.)
Here’s the challenge: it’s great from a revenue and profit standpoint, but not so great at keeping hard-pressed teams from burning out without new capacity. Plus, making new full time hires on the strength of one client is a risky business practice. Having one-time contractors service a single client is both expensive and creates logistical headaches because such people are usually not smoothly incorporated into company processes and culture.
That’s where AI is a readymade solution for scaling. For the price of a ChatGPT subscription, the firm expanded capacity immediately in a way that could also easily cut back without impacting the team.
We asked Samantha Riel, CEO of Scale Without Chaos, to build customized writing assistants and training processes for the entire team. She didn’t just show us which tools to use—she showed us how to use them for our specific workflows, writing styles and client needs. By the time she was done, the team had an entire content creation and editing process that is being implemented across the entire company.
Executing When Growth Arrives
And then the real test arrived. Right after the training ended, a text arrived from that big client: They needed us to find five authors, write five op-eds, and place those op-eds across multiple verticals in days.
They also offered a substantial bonus for each completed and landed piece. We couldn’t say “no.”
Without AI, any team would have been up until midnight throwing together content and letting other projects die on the vine. Internal priorities would be thrown aside, and the short-term revenue boost would result in a burned-out writing and editing squad. Plus, spouses would have something to say about missing dinner, and kids might forget what you look like.
Instead, the team worked normal hours (for the most part), landed dynamic authors, and secured placements at desired outlets in three verticals. And the client happily paid a significant bonus.
Machine Success Built on Human Intelligence
One of the big problems with AI is that many people think they can just tell the machine to do the work. That results in…content that looks like it was written by AI.
But just as QuickBooks is useless without a solid understanding of business finances, AI systems are useless without having smart people put smart things into the system.
And being smart about AI means taking the time to invest in it the right way. For our team, that meant hiring an outside expert who can create customized, repeatable processes. Other teams may be able to build the processes internally.
Here’s what our ChatGPT-optimized op-ed creation process looks like now:
-
Get briefing information from the client (humans talking to humans).
-
Interview the author to ensure voice and narrative compatibility with the content and placement goals (more humans).
-
Put the relevant background information into ChatGPT (human skill).
-
Provide context (human skill).
-
Provide concrete, direct and specific feedback to several iterations of the draft.
-
Run the final AI draft through our extensive structure and copy edits, plus two rounds of fact-checking. Four sets of (human) eyes will see the piece before it goes to the author and client.
And perhaps the most encouraging part of the entire story? Despite the fears of many public relations professionals, nobody—the client, outlet editors or authors—asked if we used AI. People want great copy, and they don’t care how they get it.
Dustin Siggins is Founder at Proven Media Solutions.