This is peak nerd. It’s the nerdiest thing I’ve ever wished for, except maybe the full volume set of In Search of Lost Time.
But consider this an open letter to AI scientists out there wondering, “Hmm, to what other useful purposes could we apply natural language processing?”
You may not know this, but there’s a niche specialization called conversion copywriting. It’s marketing, but the word side.
And as conversion copywriters, we’re caught somewhere between craftsmanship and science.
It’s part artisanal, part data analysis, anxious about both.
Someday we’ll look back and laugh at how bespoke it all was.
Maybe not when AI is writing the control-shattering copy, but when we have a more precise science for zeroing in on what will really convert. (Aka, prompt readers of said marketing to take the next most valuable action.)
Not to say that we’re currently guessing.
We have methods. Ways of crafting copy based on actual conversion theories.
But while the numbers are wrangled into Google Analytics, automated… the qualitative data that is the lifeblood of conversion copy?
Those conversion insights come from customer conversations
As copywriters, we collect and analyze voice-of-customer (VoC) through a process so manual, we should post it to Pinterest.
- We troll forums, groups, social posts
- We scan for comments across blog posts
- We pour through chat transcripts, customer emails and survey responses
- Sometimes we even interview customers and then there are the transcripts…
All in the name of finding out what matters to the people who buy.
We do this even though it’s crazy…
Even though it takes too long…
Because until there’s a better way to extract from the collective psyche of our audience the insights that matter, all we have is voice-of-customer.
Only, here’s the problem. Customer attitudes and behaviours aren’t static, but most research is point-in-time.
Once we’ve painstakingly, line by line read the VoC from every possible source…
Once we’ve themed it and found the, “Oh, so that’s why they buy” insights…
Once we’ve crafted a sales argument around that research…
What happens next?
That intense, systematic listening to the customer tends to screech to a halt.
Until the next round of copywriting.
See, most companies aren’t systematically listening to voice-of-customer
I get it. For clients, just managing chat and email and social and customer support and surveys are enough.
Let alone going one step further and analyzing VoC to find the red flags and ah-ha moments that are clear signs the copy needs to shift.
Sure, they could put us on retainer to do this manually.
But what if we could pull all of that qual into one place, one tool? Automatically.
What if that tool was built on AI smart enough to analyze the themes and drop audience insights into buckets?
What if you could open that tool and at a glance see what’s going on in the minds of your customers?
With a tool like that, we’d have what we need to keep optimizing copy for conversions. We’d know with fair certainty:
- How the audience sees the problem
- Their barriers to taking action
- Other solutions they’ve tried
- Outcomes they most want
- Triggers that persuade them to take action
How valuable would that be?
If the prospect isn’t lighting you up already, consider this:
Too many companies are sidelined by shifts in the market they didn’t see coming.
By competitors stealing their loyal customers.
By new expectations and alternative solutions.
To be nimble and responsive – not just in copywriting but in innovation and customer experience – we need to always be listening.
AI-powered VoC analysis tools already exist for CX…
There are enterprise AI NLP tools springing up in all directions, telling marketing-relevant stories from Big Data.
Tools like InMoment promise to ‘transform customer data into the insights that create real business impact.’
I’ve spoken with reps from Chattermill and Thematic, both designed for enterprise customer experience programs but essentially exactly what we need. There’s still a human element involved in training the AI on a case-by-case basis to work within specific contexts, which is why these tools are currently only viable for companies willing to spend 4-figures a month.
We need something lightweight, scaled for small businesses and agencies.
And we need the AI to actually understand language better than a three-year-old. (No small task, right?)
I suspect we’re on the brink of getting our hot, little, conversion-obsessive hands on just such a tool. And when it’s available, I want to be among the first to have this on offer for clients.
Is this just a sweet, delusional dream or are you out there already developing this exact thing?