Skip to content
Tutorials Center
My learning

Topics

Lesson 32 of 38 ยท Marketing Automation and AI

CRM, Lead Management and Lead Scoring

Learn how a CRM organizes contacts and deals, how marketing and sales share lead definitions (MQL, SQL), and how to build a transparent lead scoring model.

Marketing creates interest, but interest only becomes revenue if someone follows up with the right people at the right time. A CRM (customer relationship management) system is the shared database where that happens. This lesson explains how a CRM is organized, how marketing and sales agree on what a good lead is and how to build a simple lead score.

What a CRM stores

  • Contacts: people, with attributes (name, email, role, country, source) and a timeline of activity (emails opened, pages viewed, calls).
  • Companies / accounts: organizations contacts belong to (essential for B2B).
  • Deals / opportunities: potential sales, with a value, stage and expected close date.
  • Tasks and notes: follow-ups, call summaries.
  • Lifecycle stage: where a contact is in the journey.

Well-known tools include HubSpot, Salesforce, Pipedrive, Zoho, Freshsales and many others. The principles are the same.

Lifecycle stages and handoffs

From stranger to customer

Subscriber / visitor Known contact, not yet engaged
Lead Gave details or showed interest
MQL Marketing qualified: fits and engaged
SQL and opportunity Sales accepted and working it

A customer, then a repeat customer or advocate, comes next.

1 / 2

An MQL (marketing qualified lead) is a lead that marketing believes is ready to be passed to sales. An SQL (sales qualified lead) is one that sales has accepted after checking. Sales and marketing must write down these definitions together; disputes about "bad leads" almost always come from unclear definitions. A service level agreement (SLA) sets expectations: marketing delivers leads that meet the definition, sales contacts them within a set time and reports back on outcomes.

Fit and engagement

A useful score separates two ideas:

  • Fit (demographic or firmographic): how closely the lead matches your ideal customer: job title, company size, industry, country.
  • Engagement (behavioral): how interested they appear: visited the pricing page, requested a demo, opened emails, attended a webinar.

A student who downloads ten guides has high engagement but low fit. A perfect-fit buyer who never visits has high fit but no signal. You need both.

Build a lead score

Assign points to attributes and actions, and negative points for bad signs (a personal email address for a B2B product, unsubscribing, a job title of "student"). Set a threshold for MQL. Keep the model transparent so sales trusts it.

JS Lead scoring model

Press Run (or Ctrl + Enter) to see the console output here.

No output.

Showing the reasons next to the score makes it explainable, which builds trust with sales.

Calibrate the score

A score is a hypothesis. Check it against outcomes: do high scorers close more often? Compare conversion rates by score band and adjust points.

JS Does the score predict customers?

Press Run (or Ctrl + Enter) to see the console output here.

No output.

If rates do not rise across bands, the model is not helping; revisit the rules. Some teams later move to predictive scoring (machine learning on past outcomes); this needs enough closed deals and good data, and it should still be monitored. See AI in marketing.

Routing and follow-up

  • Speed: contact hot leads within minutes or hours, not days.
  • Routing rules: by region, language, account size or product.
  • Nurture the rest with an automated workflow until their score rises.
  • Feedback loop: sales records why a lead was rejected (wrong fit, no budget, no response), which marketing uses to improve targeting and scoring.

Data hygiene

Duplicates, missing fields and stale records ruin automation. Standardize fields, deduplicate on email or company domain, validate on entry and delete data you no longer need. Under privacy law (see privacy and compliance) you must also be able to find, export and delete a person's data on request.

Common mistakes

  • Scoring only behavior, so students and competitors flood sales.
  • Never decaying scores: a visit last year does not mean interest today. Reduce scores over time.
  • Hidden scoring logic that sales does not trust.
  • No agreed definitions of MQL and SQL.
  • Treating the CRM as a dumping ground without field standards.

Check your understanding

Quick check ยท 4 questions

of 4 answered You scored / 4 ยท perfect!

  1. 1. What is an MQL?

    MQL stands for marketing qualified lead.
  2. 2. Why does a good lead score combine fit and engagement?

    A perfect-fit person with no interest and an interested person who cannot buy are both poor leads.
  3. 3. Why apply score decay over time?

    Recent behavior is a stronger signal than activity from months ago.
  4. 4. How do you check whether your lead scoring works?

    A working score shows higher conversion in higher bands.

Summary

  • A CRM holds contacts, companies, deals and activity history.
  • Define lifecycle stages, MQL and SQL together with sales, with an SLA.
  • Score fit and engagement transparently, including negative points and decay.
  • Calibrate against real outcomes and keep data clean.

Next: AI in Marketing