Most qualification calls fail in the first five minutes, and not because the rep asked the wrong questions. They fail because the rep doesn’t know, in concrete terms, what the product needs from a buyer before it can do anything useful.
I sell an AI gateway. It sits between a company’s software and the large language model providers it calls, handling routing, logging, spend controls, and failover. That sounds simple until you look at what it requires: the buyer has to be sending API traffic to model providers from their own systems. If they aren’t making API calls, there’s nothing for the gateway to sit in front of.
Recently I had a call with a prospect who looked good on paper. Senior title, a company size that fit, and an inbound form fill. Within the first few minutes I learned two things. They paid for a consumer AI chat subscription, the kind a person buys to draft emails and summarise documents. They had no API access and no engineering team calling models in production. They had also already bought a competitor’s product in the same category.
Either fact alone should have ended the call. Together, they made it clear this wasn’t a fit. Instead of stopping, I spent the next twenty minutes on a monologue: how the gateway works, why cost visibility matters, a customer story. The prospect was polite, I was thorough, and neither of us got anything out of it.
The lesson isn’t that the prospect was a bad person or that the lead was wasted. I had the information I needed by minute two and didn’t act on it, because I hadn’t written down the requirements clearly enough to check against.
Now, before any call on a product, I write down three to five hard requirements in plain language. For this one, they are:
- They run workloads that call model APIs directly.
- Someone owns those API keys and can see spend by team or application.
- They have a budget line for AI infrastructure, not just seats for individual users.
- They aren’t locked into a competitor until at least their next renewal.
Then I ask about these early, as questions rather than statements. “Walk me through how your team calls the models today.” “Who holds the API keys?” “What are you using now for routing and logging?” The answers come back quickly. If someone describes a chat subscription and a browser tab, that’s the signal. I say so plainly: “From what you’re describing, this probably isn’t the right fit yet. Here’s what would need to be true for it to make sense. If that’s not on your roadmap, I’d rather you didn’t spend more time on this.” Most buyers respect that more than a forty-minute pitch.
Disqualifying early helps in two ways beyond saving your own time. It spares the buyer hours in a sales process that won’t end well for them. And it keeps your pipeline honest. Every non-fit that lingers as an open opportunity inflates your forecast and crowds out the calls that matter.
The same thinking applies to lists. I recently reviewed a campaign list and found most of the rows were hospitals and universities, with job titles that had no authority over software spend. Nobody had defined what a real buyer for this product looked like, so nobody filtered for one. Bad lists and bad calls share a root cause: the team doesn’t know the product well enough to define fit.
So the practical step is simple. Before your next discovery call, write down what the product requires in one paragraph. If you can’t do that without checking the website, you aren’t ready to qualify anyone yet.