Quick answer: Lead scoring gives each lead a number based on how well they match your ideal customer (fit) and how interested they seem (behaviour). Points are added for things like job title, company size, visiting the pricing page or requesting a demo, and sometimes removed for signs of a poor match. When a lead passes an agreed threshold, they are passed to sales. Start with a simple model built with your sales team, automate it in your marketing automation or CRM software, and review it against real sales results every few months.
When a business starts generating leads from its website, a new problem appears. Sales gets a list of names, some ready to buy, some students downloading an ebook for coursework, and no quick way to tell them apart. Calls go to the wrong people first, good leads cool off, and sales stops trusting marketing’s leads altogether. Lead scoring is a way to sort the list.
The two parts of a lead score
Most scoring models combine two kinds of information.
Fit: are they the right kind of customer?
Fit is about who the lead is. It usually comes from form fields, your CRM or research:
- Job title or role.
- Company size or turnover.
- Industry.
- Location, if you only serve certain areas.
- Budget or timeline, if you ask.
A managing director at a 50 person company in your target sector might score highly on fit. A student or a competitor would score low.
Behaviour: how interested are they?
Behaviour is about what the lead does:
- Pages visited, especially high intent pages such as pricing, case studies or contact.
- Content downloaded.
- Emails opened and links clicked.
- Webinars attended.
- Return visits within a short period.
- Direct requests, such as asking for a demo or a quote.
Someone who has viewed the pricing page three times this week is showing more interest than someone who downloaded one guide six months ago.
Why both matter
A perfect fit lead with no interest is not ready to buy. A very engaged lead who is not a fit will never buy. Looking at both helps you decide what to do next: call, nurture, or leave alone.
Some teams keep fit and behaviour as two separate scores, sometimes shown as a letter for fit and a number for behaviour (for example, A3 or C1). This makes it easier to see why a lead scored the way it did.
Building a simple scoring model
You do not need complex statistics to start. A points based model agreed with your sales team works well for most small and medium businesses.
Step 1: Look at past customers
Pull a list of customers who bought in the last year or two. What did they have in common? Which roles, company sizes and industries? What did they do on the site before getting in touch? Your CRM and analytics, including the goals you track in Google Analytics, are the starting point.
Also look at leads that did not buy. What separated them?
Step 2: Talk to sales
Ask your sales team what makes a lead worth calling, and what wastes their time. They often know signals that do not show up in the data, such as a certain job title that never has budget.
Step 3: Assign points
Give each attribute and action a value. For example:
Fit
- Target job title: +15
- Company in target size band: +10
- Target industry: +10
- Outside your service area: -20
- Student or personal email address: -10
Behaviour
- Requested a demo or quote: +30
- Visited pricing page: +10
- Downloaded a guide: +5
- Opened an email: +1
- Clicked an email link: +3
- No activity for 30 days: -10
These numbers are examples only. The right values depend on your business and should come from your own data and sales conversations.
Step 4: Set a threshold
Agree a score at which a lead becomes “sales ready” and is passed to a salesperson. Below that, leads stay in marketing nurture. Start with a sensible guess, then adjust once you see results.
Step 5: Decide what happens at each level
- Above the threshold: assign to sales with an alert, and set an expectation for how quickly they follow up.
- Middle scores: keep nurturing with relevant emails and content.
- Low scores or poor fit: keep on a general newsletter or remove from active campaigns.
Automating lead scoring
Scoring by hand does not scale. Most marketing automation platforms, such as HubSpot, Marketo, Pardot and Infusionsoft, include lead scoring. Several CRMs offer it too, either built in or through add-ons. Our introduction to marketing automation explains how these platforms fit together.
A typical automated setup:
- Forms on your site capture fit details and pass them to the platform.
- A tracking script records page visits for known leads.
- Email activity is recorded automatically.
- Scores update in real time as leads act.
- When a lead crosses the threshold, the platform assigns them in the CRM and notifies the salesperson by email or in a team chat tool.
You can also connect lighter tools with automation services if you do not use a full platform. Our guide to workflow automation covers simple ways to pass data between apps.
Score decay
Interest fades. A pricing page visit last week means more than one from last year. Many platforms let you reduce behaviour scores over time, or subtract points after a period of inactivity. Without decay, old leads build up high scores that no longer reflect reality.
Keeping fit data clean
Scores are only as good as the data behind them. A few habits help:
- Ask for the most useful fit questions on high value forms, such as demo requests.
- Use progressive profiling, where forms ask different questions on each visit, so you build a picture without long forms.
- Standardise fields with drop down menus rather than free text, so “MD”, “Managing Director” and “managing dir.” do not count as three different titles.
Agreeing the hand-off with sales
Lead scoring fails most often because of people, not software. Before launch, agree in writing:
- What counts as a sales ready lead.
- How fast sales will follow up.
- What sales does with leads that are not ready, such as returning them to nurture with a note.
- How results will be reported back.
This agreement is sometimes called a service level agreement between marketing and sales. It keeps both teams accountable.
Measuring and improving your model
Review the model every quarter:
- What share of sales ready leads became opportunities and customers?
- Are good customers coming from leads that scored low? If so, you are missing a signal.
- Are high scoring leads failing to buy? Some actions may be overvalued.
- What does sales say about lead quality now compared with before?
Adjust points and thresholds based on what you find. A model is never finished.
Common mistakes
- Too many rules at the start. Begin with ten or fifteen signals, not a hundred.
- Rewarding activity that does not predict buying. Email opens are easy to collect but weak on their own.
- No negative scoring. Without it, poor fit leads with lots of activity rise to the top.
- Ignoring decay. Old interest is not current interest.
- Setting it and forgetting it. Your market and content change, so your model should too.
FAQ
Do small businesses need lead scoring?
If you get a handful of leads a week, you can probably judge them by eye. Scoring becomes useful when volume makes it hard to give every lead attention quickly.
What is a good lead score threshold?
There is no standard number. Set one with your sales team, then adjust it based on how many scored leads turn into customers.
Can lead scoring work without marketing automation software?
A basic version can run in a spreadsheet or CRM, but automated tracking of behaviour is where scoring becomes practical.
Spend time on the leads that matter
Good scoring lets your sales team focus on the conversations most likely to lead to business, while automation keeps everyone else warm. We design scoring models and connect the tools that run them. See how we help businesses automate the busywork.
