Lead Qualification

Every day, a company collects contact data from people who are interested in its offering. In sales, such prospects are called “leads”: someone has filled out a form, downloaded a brochure, or handed over their business card at a trade fair. Most of these contacts never buy anything. Lead qualification is the process by which a company determines which contacts are worth calling. What gets checked, for example, is whether the person actually fits the target group, whether they have the money to buy, and whether they are authorized to decide themselves. In the past, employees did this over the phone; today, software handles a large part of the pre-sorting.

Why sales time is the scarcest resource

A conversation with a prospect costs working time, and working time in sales is expensive. If a salesperson calls ninety contacts and only two of them fit, most of the day is wasted. This is exactly why pre-sorting is so economically important. It determines how much revenue a team of the same size can generate.

For investors and analysts, this is a concrete metric. Companies report how many qualified contacts actually end up buying. If this rate rises, the cost per acquired customer falls. For software companies with subscription models, this figure is considered one of the most important indicators of healthy growth.

But there is also a downside. Those who sort too strictly discard contacts that would have bought later on. Those who sort too loosely flood the team with hopeless cases. Setting this threshold is a genuine management decision, not merely a technical question.

Points, criteria, and learning models

The classic method works with a point system. Each characteristic earns plus or minus points: a matching industry plus ten, a private email address minus five, CEO as a position plus twenty. Once a certain point total is reached, the contact is considered qualified and passed on to sales. These rules are set by humans, usually based on experience.

Modern systems instead let a program learn the rules from data. The model is shown thousands of past contacts along with the outcome: bought or not bought. The model searches for patterns in this data and estimates a purchase probability for each new contact. Such predictions often uncover correlations that no one would have previously formulated as a rule, such as which pages someone visited before filling out the form.

In recent years, language models have been added to the mix — that is, programs that can read and write text. They analyze notes from phone calls, email replies, and company descriptions. Some providers even let their systems send initial follow-up emails to clarify budget and responsibility. A common misconception is that such a system knows the truth. It only calculates based on past experience and regularly gets unusual cases wrong.

From cookie banners to Salesforce figures

People encounter this process as affected individuals without even noticing. Anyone who downloads a whitepaper on a company website is classified and ends up in a ranking. Filling out a contact form at a car dealership or an insurance company also triggers this evaluation. How quickly a call follows afterward often depends directly on the calculated score.

In business news, it is mainly vendor names that come up in this context. Salesforce, HubSpot, and Microsoft sell customer management systems in which this evaluation is built in. When such corporations announce new AI features, it very often concerns the automatic qualification of contacts. The market is therefore an important example of where AI is already making money today.

It is important to distinguish this from related terms. Lead generation means acquiring new contacts in the first place, for example through advertising. Qualification comes afterward and filters this pool. And because personal data is analyzed in the process, data protection plays a major role — in Europe, regulated by the General Data Protection Regulation.