Lead scoring applies a model to information about a prospect in order to support prioritisation. The inputs may concern suitability, observed activity or other relevant signals.
The score is an aid to judgement. It does not automatically establish a person's budget, intention or likelihood of becoming a customer.
Fit versus engagement
Fit concerns how well the prospect matches the business's intended customer. Engagement concerns recorded interaction.
Keep those concepts distinguishable. Someone can read many articles while remaining outside the service's scope, and a suitable buyer may make a direct enquiry with little recorded browsing.
How should criteria be chosen?
Use evidence from the sales process and the requirements of the offer. Define what each signal contributes and why.
Avoid assigning points simply because an event is easy to track. Measurement convenience is not the same as predictive value.
What can distort a score?
Incomplete information, duplicate records and unreliable tracking can produce misleading priorities. A large number can conceal weak inputs.
Review data quality before automating consequential follow-up. The system should make uncertainty visible rather than convert missing information into false precision.
What should a threshold do?
A threshold should trigger an appropriate review or action under the operating process. It needs an owner and a clear meaning.
Do not assume an arbitrary score should automatically move every prospect into an aggressive sales sequence. The action needs to fit the evidence.
How is the model evaluated?
Compare scores with actual qualification and outcomes over suitable periods. Review false positives and overlooked suitable prospects.
The purpose is to improve prioritisation, not to defend the initial model. Adjust criteria when experience shows they are unhelpful.
Should AI be involved?
Some systems use predictive methods, while others use explicit rules. Either approach needs evaluation and a clear understanding of its inputs.
A more complex model is worthwhile only if it improves the task sufficiently. Explainable operational rules can be more useful than sophistication without usable evidence.
How should teams maintain it?
Document the model, ownership and review cycle. Changes in the offer or audience can make older assumptions less relevant.
Use the score alongside the actual request and professional judgement. A number should help people understand a lead rather than prevent them from reading it.
Further reading
For reference, see the supporting documentation. Explore related explanations of Customer relationship management, Lead nurturing, Lead generation.
A practical example
Illustrative example: a company separates service fit from browsing activity. A highly active but unsuitable contact does not outrank a clear, relevant project request simply because more website events were recorded.
Why it matters
Lead scoring can organise attention when it reflects meaningful evidence. Clear criteria and ongoing evaluation prevent a convenient score from replacing qualification.
