Charcoal marble with a vein of gold

How to tell a real B2B buyer from a time-waster: the five rules we use.

Score every inquiry against a few written rules before a person reads it. Send the ones that pass to sales, park the maybes, and drop the rest.

The short answer

Check five things before anyone spends time on an inquiry: that the company exists, that it is the kind of buyer you sell to, that the order is worth your time, that the contact details work, and what the message itself says about intent. Write the rules down, score every inquiry the same way, and route it on the score.

Why this matters more than more leads

Most businesses that ask for more leads already get inquiries. The problem is what arrives. One inbox receives wholesale requests, students, job seekers, vendors selling their own services, and scams, and a founder or a salesperson reads all of it by hand.

For one of our B2B clients, the screen now runs on every inquiry before sales sees it, and most of what arrives never reaches a salesperson's inbox. The work was never getting more inquiries. It was finding the few worth a call, quickly.

The five rules

  • Does the company exist? Look up the business. A real buyer has a website, a registered name, or at least a trace in search results. A domain registered last week is a warning sign. This single check removes most scams.
  • Is it the kind of buyer you sell to? Write down who you serve: retailers, distributors, restaurants, contractors. An inquiry from outside that list is not a bad person, only a bad fit. Say so politely and move on.
  • Is the order worth it? Set a minimum. Below it, the sale costs more to serve than it earns. Above it, the inquiry earns a fast reply.
  • Do the contact details work? A phone number that does not connect, or a free email address on a corporate order, lowers the score. A duplicate inquiry from the same sender is merged, not counted twice.
  • What does the message say? This is the only step where AI helps. A language model is good at reading free text and judging urgency and intent. It is poor at arithmetic and at consistency. So the model gives a judgement, and plain code does the scoring and the routing.

Three outcomes, never one inbox

  • Pass. Straight to a sales rep, while it is still warm.
  • Nurture. Real interest, wrong timing or size. Keep it warm.
  • Discard. Logged, so you can check the rules later, but nobody spends time on it.

This is the system behind our Qualification & Routing service, and it runs on the same layer as Business Automation.

What changed when the rules were written down

The pattern we see is consistent: once the rules are written down and search is pointed at buyer intent rather than cheap clicks, total inquiries usually fall and the share of real buyers rises sharply. Fewer inquiries, far more of them worth answering, often on the same budget.

How to start this week

  • Write your five rules on one page. If you cannot write a rule down, it is a feeling, not a rule.
  • Score the last 30 inquiries by hand. Count how many pass.
  • Decide where a pass goes. Name who answers it within a day.
  • Only then automate it. Automating a rule you have not tested only makes the mistake faster.

Common questions

Does AI decide who is a good lead?

No. The rules are yours and you can read them. AI only reads the free text of an inquiry for urgency and context. It never sets or overrides a score.

What if a good buyer gets filtered out?

Every discard is logged with the rule that caught it. Review the log each month and adjust the rule that was too strict.

How many rules do we need?

Start with five. More rules early on usually means more exceptions, not better leads.

Tell us what's happening in your business. We'll tell you honestly whether this is a fit, and what it would take.

A free call, no charge and no pitch deck.
If it's not a fit, we say so on the call.

Book a free call