“You have a copywriter on staff?” – A friend working at a large TV station asked, with much surprise in her voice.

She continued: “Can’t you have AI do that now?”

While I forced a smile and did my best to answer calmly, I was hiding a deep sense of frustration. I see copywriting as some of the highest value-add activity in all direct response marketing, whether we’re carefully crafting the unique voice of a brand we partner with, or writing a sales letter that makes or breaks the performance of a sales funnel.

It’s the words on the page that get someone to take action, and yet much of the marketing profession is racing to see who can offload responsibility for choosing said words to Claude the fastest.

Now, I’m no anti-AI crusader who refuses to accept change isn’t quietly knocking on but rather smashing through the door of every marketing agency regardless of their niche. Neither do I believe copywriting is some sacred ritual reserved only for the initiated few who spend their weekends dissecting print sales letters from the 60s (which is actually a lot more fun than it sounds…).

What I am saying is there is a balance to be struck between human character, and the speed and scale of AI. In this sense, I believe the losers of this transition will be those on either end of the spectrum.

The winners will be the agencies and copywriters that keep creating original human works but use AI to get different versions of these creations seen on more placements than a human ever could.

To draw a line, I differentiate between two types of copywriting: copywriting for impact, and copywriting at scale.

Some pieces of sales copy are super-high impact. Think of a video sales letter telling your brand’s origin story or a product launch email where you’re seeking to connect with the people on your list on a personal level. These are ‘make it or break it’ touchpoints for your brand, where human connection is paramount.

On the other hand, we have copy at scale, where the volume of output at speed is more important than curating the one best thing every time. Think of coming up with 30 different angles for testing new static ads to see which one works, or creating email automations that deliver your message at every imaginable customer touchpoint.

As a general rule:

Copy written for impact stays visible to many people for a long time and has an easily measurable impact on your funnel’s performance, while copy written at scale is the opposite.

Here are some common examples:

When we start working with a new partner at Polarision, I like to create the “human foundation” first. Most of our projects involve writing a long-form sales letter or video script, and this is where we like to keep things old-school.

As we write this original piece of copy, we often end up ‘accidentally’ discovering many of the rules that will guide our brand voice and get clear on what we want to see or not see when working on this project.

In this phase of the project, we use AI in the following ways:

Once we have our sales letter and the brand-specific copywriting rules we discovered while creating it, we can add more AI to our workflow. The difference from before? We can now give it examples of what good copy looks like and communicate our preferences clearly, which is the first step to making AI actually take on the voice of your brand.

So no, we’re not about to fire our copywriter. Having a highly skilled human on the team leveraging AI to 5X their output is the winning play.

In the next blog post: I’ll talk about how we set up our Claude projects for each new partner, and what we put into an effective brand voice file.