How AI is reshaping sales enablement

Let me start with a confession. Early in my career, I spent what felt like weeks building a sales enablement deck. Charts, competitive matrices, color-coded slides. I was genuinely proud of it. Then, about a month after I'd uploaded it to our enablement software, a sales leader asked me to build the exact thing I'd already made. He had no idea it existed.

That moment stuck with me. All that work, sitting there collecting digital dust.

And I think a lot of product marketers have been there. You build something you're convinced is going to change how the sales team operates. You know the product inside and out. You know the customer. But knowing the product isn't the same as knowing how it gets sold. There's a lot of nuance in that gap, and I spent a good chunk of my early career not fully appreciating it.

So that's really what I want to talk about here. How AI is changing the way we think about sales enablement, where it genuinely helps, where it falls short, and the part that no model can replicate.

Sales enablement is evolving beyond asset creation

For a long time, sales enablement meant building things. Battle cards. Pitch decks. One-pagers. Sales would ask, PMM would build, and the cycle would repeat. That was the job.

I think that's shifting. The role is moving toward something more strategic, more focused on driving insight and shaping how a sales team actually goes to market. It's less about orchestrating the perfect deck and more about orchestrating the perfect strategy.

That means understanding what the team truly needs, which is often different from what they ask for. When a rep says "I need a battle card," what they might actually need is help explaining one specific feature that keeps tripping up prospects. The battle card is the vehicle. The real need is the clarity.

And getting to that real need? That requires getting closer to the sales cycle than most product marketers naturally do.

Where AI genuinely helps

Here's where I'll give AI full credit: it's made certain parts of this job dramatically faster and more insightful.

Analyzing sales calls at scale

We use Gong internally, and I'd imagine a lot of you are using something similar to capture call transcripts. The honest truth is that no human can read every transcript. There just aren't enough hours. AI changes that.

What I've started doing is feeding those transcripts into Claude and querying them for patterns. What objections keep coming up? Where does the conversation stall? But here's the thing I'd encourage you to think about too: prompt for what's landing, not just what's falling short. It's easy to default to looking for problems.

But if something in your pitch is resonating with prospects, you want to know that so you can amplify it, not just fix what's broken.

Critiquing your existing assets

This one I've found genuinely useful. I'll pull in a batch of transcripts, upload the asset I want to evaluate, and then ask Claude to take on the persona of the customer we're trying to reach. From that perspective, I ask: how well does this asset answer the questions that actually come up in these conversations?

The gaps it surfaces are often things I wouldn't have caught on my own. Questions I'd never thought to address. Assumptions baked into the copy that don't hold up once you look at real customer language.

Querying customer feedback

We use a tool called Interpret internally, which I connect to Claude to query our customer feedback database. It's a way of finding where customers are hitting friction and then thinking about how we can explain things more clearly. You can do similar things by running AI across competitor landing pages, existing assets, or any corpus of customer-facing content you have.

NotebookLM as a shared knowledge base

This one I recommend a lot. NotebookLM lets you upload a set of sources, and when you query it, the responses are grounded only in what you've uploaded. No hallucinations from the broader internet. I built a notebook with competitive insights, messaging docs, and product information, and shared it with the sales team.

The idea is that when a rep has a question and I'm unavailable (I'm on the West Coast; a lot of our team is East Coast or Europe-based), they don't have to wait for me. They can query the notebook and get an answer that's rooted in the actual information we've developed. It's not a replacement for the relationship. It's a way of extending your reach when you're not in the room.

The shared context file: keeping AI outputs on-brand

Here's something I learned the hard way. When everyone has access to AI tools, they use them including sales reps. I had a couple of reps who built their own battle cards using AI, and I genuinely appreciated the initiative. But when I read what they'd produced, the messaging was off. The positioning didn't match what we'd carefully developed. The strategy wasn't there.