Qualified leads vs contacts: why most companies' CRMs are full of noise

A CRM is a neutral tool. It records what is delivered to it. If marketing delivers a thousand contacts a month, the CRM has a thousand more records. If half of those contacts don't have a company name, another half don't reply, and a subset are competitors making a survey, the CRM still shows a thousand records. No tool has its own criteria for distinguishing real leads from noise.

This point seems obvious, but it has consequences that are rarely realised. The first is that the indicator “number of leads in the CRM” is almost always a vanity metric, not a business metric. Measuring base growth without measuring base composition leads to misleading conclusions about the commercial health of the operation.

The second is that sales people, faced with a full CRM but no prioritisation structure, tend to develop their own method of sorting. Each salesperson creates their own informal rules for deciding which contacts to work with, which nobody documents, which leave with them when they change companies, and which are not replicated with newcomers. The organisation loses accumulated commercial capacity because the criteria have never been shared.

The third is the cost in productivity. Misaligned teams can lose up to 60% of leads and cost B2B companies 10% or more of annual revenue. Aligned teams, on the other hand, achieve 24% faster revenue growth, 36% higher customer retention, and are 67% more effective at closing deals. The figures may seem exaggerated until you count up the time each salesperson spends screening contacts that will never convert, add up all the weeks of the year, multiplied by all the people on the team.

CRM, which should be a tool for commercial organisation, becomes a noise file when prior qualification fails.

The reasons are threefold and mutually reinforcing.

The first is metrics. When marketing is measured by “leads generated” without the required qualification, the quickest way to beat the target is to reduce barriers to entry. Short form, generic question, attractive but unfiltering bait. Higher volume of leads, easier to reach the number. Nobody is acting in bad faith, they're just optimising for what was asked for. The problem isn't behavioural, it's structural: the wrong metrics produce the wrong behaviour with predictable consistency.

The second is definition. “Qualified lead” rarely has a written definition in most companies. When it does exist, it's vague enough to mean nothing operationally. “Company interested in our service” is a phrase that describes anyone who clicks on the form. Without concrete and verifiable criteria, the qualification is made to the sensitivity of each person, at each moment, without consistency.

The third is process. Even when there is a clear definition, the moment at which qualification takes place is often poorly positioned in the funnel. Qualification by sales, already in a meeting, is too late - the cost of the meeting has already been paid. Qualification by marketing without direct contact with the lead is blind - it's based on assumptions about who filled in the form. Effective qualification has to happen between the two moments, with an explicit process that marketing and sales recognise as shared.

The most common objection to more demanding qualification is that it drives away legitimate leads. The implicit logic is: if we ask for more information on the form, fewer people will fill it in, and we'll lose contacts that might be good. This logic is flawed and worth explaining.

The form doesn't decide who is a qualified lead. The form decides who has the patience to fill in a long form. When the barrier is excessive (eight mandatory fields to download a brochure), it does indeed alienate contacts who prefer quicker solutions. But when the barrier is calibrated (three to five fields that any real buyer would answer without hesitation), it only drives away those who were never going to buy.

The key question isn't “how many fields should I ask for? It's ”what information distinguishes a real buyer from a curious one, and how do I ask for that information in a natural way“. Name, company, and size of operation rarely turn real buyers away. Forms that ask about available budget, decision stage, or expected timeframe are more likely to turn them away, but they filter better. The decision about which questions to include should be made based on the cost of a disqualified lead reaching the sales assistant, not on the feeling that ”it's a shame to lose contacts“.

A productive way to think of the form is as a first conversation. What three or four questions would a sales assistant ask in the first two minutes of a call to see if it's worth continuing? If these questions can be asked in the form, the first call starts at a much more advanced point, and the sales assistant knows from the outset which contacts are worth pursuing.

Most marketing automation platforms offer lead scoring functions: assigning points to each lead according to behaviour and characteristics. Managers set up rules, the system assigns numbers, and the sales team receives leads with an associated score.

In most cases, the lead scoring that is set up is cosmetic. It awards points for opening an email (one point), visiting a website (two points), downloading content (five points), filling in a form (ten points). The numbers are consistent but have no commercial significance. A lead with a score of 30 is not necessarily better than a lead with a score of 15. A high score may indicate genuine interest or it may indicate that the lead has been receiving automated communications for a long time.

Operational lead scoring is different in two respects. Firstly, it separates the fit score from the intent score. Fit measures whether the lead corresponds to the ideal customer profile: company size, sector, position, geography. Intent measures whether the lead is commercially active: recent searches, pages visited with high commercial value (pricing page, contact page), frequency of interaction in recent weeks. A lead with high suitability and low intent is nurturing. A lead with high suitability and high intent is ready for commercialisation. A lead with low suitability, regardless of intent, does not move forward.

Secondly, operational lead scoring is validated against real commercial results, not against theory. If leads with a score of 50 convert at 25% and leads with a score of 80 convert at 22%, the scoring system is wrong, no matter how logical it may seem. Calibration requires monthly monitoring of leads that have progressed and leads that have not, and adjusting the weights when the data justifies it. Behaviour-based lead scoring models, when well calibrated, can raise the conversion rate up to 40%, but this result depends on a closed loop between real conversion data and adjusting the rules.