
Cadence OTC’s SVP, GM,
writes “AI comes in with enormous raw potential and zero context. It doesn't know your brand, your customers, or your financial targets. It's a gap in the onboarding.”
Let's be
honest about what many "AI strategies" actually look like: a review team, a few pilots, and a P&L that looks exactly the same as it did six months ago.
I've been there. For
a stretch, my team and I used AI like a more sophisticated Google. Summarize research? Great. Organize a spreadsheet? Sure. But we weren't moving the needle on anything that mattered.
The shift happened when we stopped asking "What can AI simplify?" and started asking "What would I hire someone to do if budget were
no object?" That became the role we onboarded our AI into. Turns out, that one reframe changes everything.
advertisement
advertisement
The Hires That Actually Moved Our
P&L
Hire #1 — The Customer Rep (and Surprise Insights Associate)
Most businesses deploy AI in support to deflect tickets.
Fewer emails, faster resolution, done. Respectable, but we found something more interesting hiding underneath.
Once we onboarded our AI on our brand voice and product
knowledge base, we didn't just get fewer support tickets. We got an insights machine. Here's what nobody tells you: customers share way more with AI than they do with humans. There's a perceived
anonymity that loosens people up. They'll tell Maya, our AI, exactly what's frustrating them in ways they'd never say to a human rep. Maya now takes those unfiltered frustrations and routes them
directly into content strategy and product updates. She's become our most honest focus group, one that runs 24/7 and never needs a recruiting budget.
Where we're headed:
accurate healthcare information is our priority. For example, roughly 25% of consumers don't know Morning After Pill, our emergency contraceptive is legal and available without a prescription. We're
changing that by building a fully automated loop where Maya surfaces content gaps and triggers blog posts with zero manual intervention. No coding, no searching. A system that listens, acts, and
updates.
Hire #2 — The Operations Assistant
As our business grew, so did the complexity. Smaller customers needed simple
invoicing. Larger ones had full EDI platforms. We needed to meet both without doubling headcount.
So we put AI on the genuinely unsexy stuff: organizing financial
documents, automating orders and receivables, streamlining warehouse tasks like label printing. None of it shows up in a TED talk. All of it shows up in operating expenses.
But the real win wasn't avoiding a hire. It was the quality of hire we eventually made. When we did bring someone on, we could hire for judgment instead of capacity, someone
experienced enough to look at our workflows and find the next thing to automate, not someone spending their days processing paperwork. That's a fundamentally different role
and AI made it possible.
The Bottom Line
EBITDA doesn't grow in the spotlight. It grows in the shadows, in the workflows nobody wants to own, in
the administrative drag that quietly bleeds margin. That's where AI earns its keep. In the boring middle of your business.
So if your AI investment isn't showing up in
your numbers, one question worth sitting with: would you expect a human hire to succeed if you gave them as little onboarding as you gave this LLM?
Like any new hire, AI
comes in with enormous raw potential and zero context. It doesn't know your brand, your customers, or your financial targets. That's not a flaw in the technology, it's a gap in the onboarding. And
it's on you to close it.
Three AI shifts that actually accelerate results:
Write a job description. Clear role, clear expectations.
Build a knowledge base. Give it the context it needs to be useful.
Create a feedback loop. Define what "good" looks like, and reinforce it.
The models respond.
Stop searching for the right prompt. Start building the right onboarding. Your P&L will feel the difference.
If you’re interested in submitting
content for future editions, please reach out to our Managing Editor, Barbie Romero at Barbie@MediaPost.com.