To score leads, give points for how well a person fits your ideal customer and separate points for how engaged they are, then agree what total triggers a conversation with sales. Qualification is the human step that confirms the score: is there a real need, authority, a sensible timeframe and a budget range. For a small team, a simple two-axis model on a spreadsheet is usually enough.
Why separate fit from engagement
A single number hides two different questions. Someone can read every article you publish and never be a buyer, or be a perfect match who has not yet noticed you. HubSpot’s documentation on its lead scoring tool describes exactly this split: engagement scores track actions such as visiting a website, opening emails or clicking calls to action, while fit scores assess properties such as job title or company size. It also offers combined scores where the letter reflects fit and the number reflects engagement, so A1 is high fit and high engagement and C3 is low on both.
You do not need that software to use the idea. Two columns in a spreadsheet do the same job.
Build a simple model in five steps
- Define your best-fit customer. List the traits shared by customers you most want more of: type of organisation, size, location, the problem they describe.
- Choose five or six fit signals and give each a few points. For example, in a hypothetical web design studio: located in your service area, has a live website, role with decision authority.
- Choose five or six engagement signals. Visited the pricing or services page, submitted a form, replied to an email, returned within a week. Weight actions that show buying intent more than casual reading.
- Add negative points. HubSpot notes that rules can subtract points, for example for an unsubscribe. Also deduct for clear mismatches, such as an enquiry outside your service area.
- Set thresholds and actions. For example: high fit and high engagement gets a personal reply the same day; high fit and low engagement goes into a nurture sequence; low fit gets a polite, helpful standard response.
Cap the points any one group can contribute. HubSpot describes maximum limits per group so that, for instance, awareness activity cannot swamp conversion actions. Otherwise ten blog visits can outscore one quote request.
Let old activity fade
Interest expires. The same documentation describes score decay, where engagement points decrease over time at set intervals. A manual version is to count only the last 30 to 90 days of activity. This prevents a lead who looked at your site a year ago from sitting at the top of the list.
Qualify with a short conversation
Scores prioritise; they do not decide. Qualification questions confirm what the numbers suggest. Keep them to a handful:
- Need: what problem are you trying to solve, and what happens if you do nothing?
- Authority: who else is involved in the decision?
- Timeframe: when do you need this done?
- Budget range: do you have a range in mind?
- Fit: can we actually help with this?
Ask them as a conversation, not as a form. Record the answers in one place so the next person to speak to the lead does not repeat the questions. If you use AI to summarise enquiries or draft replies, see AI for sales enablement and keep a person responsible for the decision.
Test the model against reality
Every quarter, review recent customers: what scores did they have before they bought? And review high-scoring leads who never bought. If wins were scoring low, your rules are wrong. Our guide to measuring creative campaign performance covers the wider habit of checking measures against outcomes, and landing pages that convert campaign traffic covers capturing the right information at the form stage.
Limits and cautions
- With few leads, a score is a guess. Under a few dozen enquiries, a manual triage by a person is often more accurate. HubSpot’s own AI scoring option requires at least 50 contacts with 25 converted and 25 not converted, which gives a sense of how much data automated models need.
- Do not score on sensitive personal attributes. Use business-relevant signals.
- Scoring does not give you permission to contact someone. Marketing emails to people without consent can breach the Spam Act 2003, whose Federal Register text lists the rules on consent, sender identification and unsubscribe. Take professional advice on your obligations.
- Do not let a low score stop you replying courteously to a genuine enquiry.
Where Radiance Impact fits
Radiance Impact’s Growth team works on strategy and campaigns, and its Technology division can build forms and workflow automation that capture and route enquiries. See Growth and Technology, or contact us if you want help designing a simple scoring workflow.
Sources
- Understand the lead scoring tool, HubSpot Knowledge Base
- Spam Act 2003 (text), Federal Register of Legislation
