Most lead scoring models are educated guesses dressed up as data science. A marketing team picks some criteria, assigns arbitrary point values, and then wonders why the "qualified" leads sales receives never close.

The problem is not the effort. The problem is the framework. Arbitrary point allocations — 5 points for an email open, 10 for a demo request — do not correlate with actual buying intent. You end up with a volume play: lots of MQLs, few actual opportunities.

After building and iterating on lead scoring models across multiple B2B SaaS products, I landed on a structure that actually predicts pipeline quality. It is not complex. It has 16 criteria across four categories. It weights behavioral signals over demographic data. And it has a clear, defensible threshold for what qualifies as a sales-ready lead.

Below is the full template. Use it as-is, or adapt the weights to your specific sales cycle. Either way, stop guessing.

The Lead Scoring Model Template

This model scores leads on a 0–100 scale. Four categories. Sixteen total criteria. Each lead accumulates points across all categories until it hits (or fails to hit) the MQL threshold of 70.

MQL threshold: 70 points. Below 70, nurture and retarget. 70–90, route to SDR for qualification call. Above 90, same-day outreach from a sales rep.

How the Categories Break Down

The four categories are not arbitrary. Each one addresses a distinct dimension of buying intent.

Demographic Fit

Job title predicts influence in the buying process. A C-level executive who downloads your ROI calculator is different from an IC at the same company who did the same thing. The former is a champion building a business case. The latter is doing research. Both matter — but they require different follow-up sequences.

Firmographic Fit

Company size predicts deal complexity and ACV. A 10-person startup is not going to pay $50K/year for software, regardless of how engaged they are. Filter firmographic fit early so you are not chasing deals that will never close at your price point.

Behavioral Signals

These are your highest-intent indicators. Someone who visits your pricing page twice in a week is in active evaluation mode. They are not browsing — they are comparing. The pricing page visit gets the highest point value of any single criterion in the model because it is the most reliable predictor of commercial intent.

Budget & Timeline

These criteria are hardest to collect but most predictive of close rate. If a lead mentions a budget, a timeline, or a current vendor — even casually in a chat conversation — that signal is worth capturing. Weight these heavily because they separate people who are shopping from people who are buying.

Why Most Lead Scoring Models Fail

The standard failure mode looks like this: marketing declares any lead with a demo request an MQL. Sales gets 200 MQLs per month. 180 of them are not buyers. After six months, sales stops trusting marketing's lead quality, and the MQL definition quietly becomes "anyone who filled out a form."

This happens because the scoring model was built in a vacuum. Marketing picked criteria that were easy to track — form fills, email opens, page views — rather than criteria that actually correlate with purchase intent. The result is volume without quality.

The fix is straightforward: define your MQL threshold based on historical conversion data, not gut feel. Look at your last 100 closed-won deals. What was the average score of those leads at the time they were declared MQL? Use that number as your threshold. Then track your score-to-close rate by band and iterate.

How to Implement This Lead Scoring Model

If you are starting from scratch, here is the implementation sequence:

• Week 1 — Define your data sources. You need CRM data (job title, company size), marketing automation data (page visits, email engagement), and conversation data (chat transcripts, discovery call notes). Map each criterion above to a data source.
• Week 2 — Assign scoring weights in your CRM or MAP. HubSpot, Salesforce, and Marketo all support custom scoring models. Start with the weights in the table above. Do not adjust them until you have three months of conversion data.
• Week 3 — Set threshold alerts. Configure your system to notify sales when a lead hits 70+ points. Route it immediately. Speed of follow-up matters — a lead at 90 points who does not hear from sales for 48 hours is significantly less likely to convert.
• Month 2+ — Review band conversion rates. Pull conversion rates for each score band (0–30, 31–70, 71–90, 91–100). If your 71–90 band is converting at under 15%, your threshold is too low or your weights are wrong.

What to Do With Low-Scoring Leads

A lead at 30 points is not a dead end. They are a future opportunity. The key is to have a retargeting path that increases their score over time without sales involvement.

• Route them to a nurture sequence aligned to their role and company stage.
• Serve them content relevant to the problem their score indicates they are trying to solve.
• Trigger a chat interaction when they return to your site — a real conversation, not a chatbot, because you want to collect the behavioral and budget signals that your scoring model is missing.
• Re-score automatically when they engage. A lead who was at 30 and attended a live demo webinar jumps to 38. Two more visits to the pricing page and they are at 78 — MQL.

The AI Layer: Automate the Qualification Workload

Manual scoring and follow-up does not scale. If you have more than 500 inbound visitors per month, you cannot rely on SDRs to score and immediately respond to every lead that crosses the MQL threshold.

AI changes this. An AI sales agent can engage every website visitor in a real conversation, collect the behavioral and budget signals your form cannot capture, and score the lead automatically — all before a human is in the loop. When a lead hits 70 points, the AI has already delivered a personalized demo, answered objections, and booked a meeting with a qualified buyer.

That is what VendAItion does. It runs the top of your funnel 24/7, qualifies every visitor against this model (or your adapted version of it), and surfaces only the leads that sales should spend time on. At $149 per month, it costs less than a single SDR hour.

Frequently Asked Questions

Stop Guessing. Start Scoring.

Your best leads are already visiting your website. The problem is that your current system treats a first-time ebook download the same as a pricing page visitor who returned three times in a week. It is not.

Download this lead scoring model template. Map it to your CRM. Set your MQL threshold at 70. And watch what happens to pipeline quality when sales starts every week with leads that actually want to have a conversation.

If you want to see how AI automates the entire qualification and demo delivery process — scoring, routing, and meeting booking included — book a 30-minute demo.