Lead scoring statistics and benchmarks: the numbers you can trace to a source
Sourced statistics on lead scoring, funnel conversion, response time, buying groups, data quality, and AI, plus the popular figures nobody can trace.
Most lead scoring statistics you see quoted are old, unsourced, or both. This page collects the numbers that can be traced to the organization that ran the study, with the year, the sample, and a link to the source. It also lists the popular figures we could not trace, so you know which ones to stop quoting.
Figures were checked in October 2026. Many come from vendors with something to sell, and we say so where it applies. Benchmarks describe other companies. Use them to sanity-check your own numbers, not to set your targets.
How to read these numbers
- Vendor studies. A company that sells intent data will find that intent data matters. The figure can still be accurate; read it with the sponsor in mind.
- Self-reported surveys. Most of these numbers are what people said, not what a system measured.
- Benchmarks by stage. Definitions of MQL and SQL differ from company to company, so stage conversion rates are the least comparable numbers on this page.
- Dates. We give the year for each. Anything older than five years is marked as a classic.
Lead scoring adoption and satisfaction
There is no recent, independent survey on how many B2B companies use lead scoring. These are the traceable figures.
| Statistic | Source | Year and sample |
|---|---|---|
| 86% of marketers use lead scoring in their demand generation strategy. Fewer than 20% called their efforts highly effective, and 15% said sales would view scored leads as qualified | Demand Gen Report, Lead Scoring Survey | 2016. Sample not given. A classic |
| 44% of respondents use lead scoring, and 16% plan to within 12 months. Average self-rated effectiveness was 5 out of 10 | Decision Tree Labs, commissioned by Lattice Engines | 2013. Sample not given. Sponsored by a predictive scoring vendor. A classic |
| 55.3% of B2B marketers measure MQLs, down from 72% the year before | 6sense, Marketing Attribution and Contribution Benchmark | 2025. 716 B2B marketers. Vendor |
| Marketing and sales collaborate on only 3 of 15 commercial activities, and 90% of leaders say their functions’ priorities conflict | Gartner | Late 2023. 412 senior marketing and sales leaders |
| 68% of sales professionals say lead quality improved year over year | HubSpot, State of Sales | 2025. 1,000 sales professionals. Vendor |
What it means for scoring. The 2016 finding is still the most telling: nearly everyone scored leads, and few thought it worked. The gap between marketing and sales has not closed since.
Funnel conversion benchmarks
Stage rates depend on how each company defines a stage. The figures below come from one agency’s client data and from one revenue platform’s customer data, and they should be named as such when quoted.
| Statistic | Source | Year and sample |
|---|---|---|
| Lead to MQL averages 31%. By channel: client referrals 56%, SEO 41%, email 38%, PPC 29%, webinars 19% | First Page Sage | 2025. Agency client data, about 70% B2B |
| B2B SaaS funnel: lead to MQL 41%, MQL to SQL 39%, SQL to opportunity 42%, opportunity to close 36% | First Page Sage | 2025. More than 50 B2B SaaS clients |
| Lead to opportunity: B2B SaaS 6.2%, financial services 5.4%, manufacturing 4.9%, cybersecurity 4.1%, IT and managed services 3.0% | First Page Sage | 2025. Agency client data |
| New-logo win rate is 19%, and expansion win rate is 45% | Ebsta and Pavilion, GTM Benchmarks | 2025. 655,000 opportunities. Vendor data |
| Win rates are 3.1 times higher when opportunities match the ideal customer profile | Ebsta and Pavilion, H1 Update | 2024. 566 companies. Vendor data |
| Average close rate is 20% across industries: software 22%, finance 19% | HubSpot | 2024 survey. Sample not given. Vendor |
| 60% of SDRs are at quota, and the median quota is 10 opportunities a month | The Bridge Group, SDR Models and Metrics | 2025. 351 B2B companies |
A caution on MQL to SQL: First Page Sage publishes different rates on different pages, from about 13% to about 39% for B2B SaaS. Measure your own before you compare.
What it means for scoring. The ICP figure is the one to act on. If deals that match your ideal customer profile win three times as often, fit belongs at the center of the model.
Speed to lead
| Statistic | Source | Year and sample |
|---|---|---|
| Of 2,241 US companies audited, 37% answered a web lead within an hour, 24% took more than 24 hours, and 23% never responded | Oldroyd, McElheran, and Elkington, Harvard Business Review | 2011. A classic |
| Firms that made contact within an hour were nearly 7 times as likely to qualify the lead as those that waited an hour longer | Same study | 2011. 1.25 million leads |
| Of 1,000 B2B SaaS demo requests, 63.5% got no response. The average response time among the rest was more than 29 hours | RevenueHero | 2024. Vendor test |
| Of 114 B2B companies tested, the average email response took 11 hours 54 minutes, and one replied within 5 minutes | Workato | 2022. Vendor test |
What it means for scoring. A score sets the order of work. It should never slow the reply to someone who asked for a demo.
Buying groups and buyer behavior
| Statistic | Source | Year and sample |
|---|---|---|
| A typical buying decision involves 13 internal stakeholders and 9 external influencers | Forrester, State of Business Buying | 2026 |
| 86% of B2B purchases stall, and 89% involve more than one department | Forrester, State of Business Buying | 2024 |
| Buyers first contact sellers 61% of the way through the journey, and 79% of first contacts are started by the buyer | 6sense, Buyer Experience Report | 2025. Nearly 4,000 buyers. Vendor |
| The winning vendor was on the day-one shortlist 95% of the time | Same report | 2025. Vendor |
| 67% of B2B buyers prefer a rep-free experience, up from 61% a year earlier | Gartner | 2025 survey. 646 B2B buyers |
| 73% of buyers avoid suppliers who send irrelevant outreach | Gartner | 2024 survey. 632 B2B buyers |
| Sales reps spend 40% of their week selling | Salesforce, State of Sales | 2026. 4,050 sales professionals. Vendor |
What it means for scoring. With 13 people in a decision, a score on one lead misses most of the picture. Score the account and the people at it. And with reps selling for 40% of the week, the order they work in matters.
CRM data quality
| Statistic | Source | Year and sample |
|---|---|---|
| 76% of CRM users say less than half of their organization’s CRM data is accurate and complete | Validity, State of CRM Data Management | 2025. 602 CRM users and administrators. Vendor |
| 37% say their company loses revenue as a direct result of poor data quality | Same report | 2025. Vendor |
| Only 35% of sales professionals completely trust the accuracy of their organization’s data | Salesforce, State of Sales | 2024. 5,500 sales professionals. Vendor |
| 79% of high-performing sales teams prioritize data hygiene, against 54% of underperformers | Salesforce, State of Sales | 2026. 4,050 sales professionals. Vendor |
| Poor data quality costs organizations at least $12.9 million a year on average | Gartner | 2020 research. Company-wide, not CRM only |
What it means for scoring. A model can only score the fields that are filled in. Check how complete each field is before you give it weight.
Email opens as a signal
| Statistic | Source | Year and sample |
|---|---|---|
| Apple Mail accounted for 62.26% of tracked email opens in July 2026 | Litmus, Email Client Market Share | 2026. Over 1 billion opens. Updated monthly |
| Opens affected by Apple Mail Privacy Protection passed 50% of Apple opens by May 2022 | Litmus | 2022 |
What it means for scoring. Mail Privacy Protection loads tracking pixels whether or not a person reads the email. Score replies, clicks, and meetings instead of opens.
AI and predictive scoring
| Statistic | Source | Year and sample |
|---|---|---|
| 87% of sales organizations use some form of AI | Salesforce, State of Sales | 2026. 4,050 sales professionals. Vendor |
| Sellers who partner effectively with AI are 3.7 times more likely to meet quota | Gartner | 2024. 1,026 B2B sellers |
| 31% of chief sales officers name proving the return on AI tools as a top challenge | Gartner | 2025 survey. 227 chief sales officers |
| A review of 44 academic studies found decision trees and logistic regression were the most used methods in lead scoring models. It gives no single figure for lift | Wu, Andreev, and Benyoucef, Information Technology and Management | 2024. Peer reviewed |
We found no vendor-neutral study that gives one conversion lift figure for predictive scoring. Be wary of any that claims to.
How much data AI scoring needs, by vendor
| Product | Published minimum | Source |
|---|---|---|
| Salesforce Einstein Lead Scoring | 1,000 leads created and 120 converted in the last 200 days | Salesforce Help |
| Microsoft Dynamics 365 Sales | 40 qualified and 40 disqualified leads in the training window | Microsoft Learn |
| Zoho CRM Zia scores | 200 records to start, plus 75 ideal and 75 non-ideal records | Zoho Help |
| HubSpot, create score with AI | A sample of 50 contacts: 25 converted, 25 not | HubSpot Knowledge Base |
What it means for scoring. If you do not have the history, a model you set by hand and test against your own wins is the safer start.
Intent data
| Statistic | Source | Year and sample |
|---|---|---|
| 44% of B2B marketers who use intent data apply it to account or lead scoring, and 70% name data quality as a top challenge | Intentsify and Ascend2, State of Intent Data | 2024. 352 B2B marketers. Vendor |
| More than 85% of companies using intent data report business benefits, and more than 70% use several providers | Forrester | 2023 survey |
| 74.1% of B2B marketers measure opportunities from third-party intent signals, up from 61.6% | 6sense | 2025. 716 B2B marketers. Vendor |
What it means for scoring. Intent is one input. It usually names a company, not a person, so it belongs in the account score alongside fit.
Statistics to stop quoting
These appear in most roundups. We could not trace them to a primary source, or the source does not say what is claimed.
| Claim | Problem |
|---|---|
| “Qualification odds drop 400% between 5 and 10 minutes,” credited to Harvard Business Review | Not in the HBR article. It traces to a 2007 vendor study of six companies |
| “44% of companies use lead scoring,” presented as current | From a 2013 study sponsored by a scoring vendor |
| “Lead scoring increases lead generation ROI by 77%” | Usually credited to MarketingSherpa. No primary source found |
| “79% of marketing leads never convert to sales” | Usually credited to MarketingSherpa. No primary source found |
| “78% of customers buy from the first company to respond” | No primary source found |
| “B2B data decays 22.5% a year,” and the 30% and 70% versions | No primary source found |
| “Buyers are 57% of the way through before they contact sales” | From 2012, and replaced by newer research |
| “AI lead scoring lifts conversions by 30%,” or 50% | Credited to several firms. No study found |
| “MQLs fail 95% to 99% of the time” | Stated by a vendor without a named study |
What the numbers add up to
Read together, the traceable numbers point the same way. Buying is done by groups, so score accounts and the people in them. Buyers show up late, so fit and timing matter more than counting clicks. CRM data is patchy, so check the fields before you weight them. Email opens are unreliable. AI scoring needs history that many teams do not have, and nobody independent has measured its lift. And reps have limited time, so the order they work in is worth getting right.
Our five-step framework for building an ICP scoring model turns those points into a model. BoostIdeal runs that kind of model in HubSpot: you set the weights, and every score writes back to the record with its reasons.
Frequently asked questions
What percentage of companies use lead scoring?
There is no recent, independent figure. Demand Gen Report’s 2016 survey found 86% of marketers used lead scoring, and a 2013 study sponsored by a scoring vendor found 44%. Both are too old to describe today.
What is a good MQL to SQL conversion rate?
It depends on how you define each stage. One agency’s client data puts B2B SaaS at about 39%, while other pages from the same source show about 13%. Track your own rate over time instead of chasing a benchmark.
How fast should you respond to a lead?
As fast as you can. The Harvard Business Review study from 2011 found firms that made contact within an hour were nearly 7 times as likely to qualify the lead as those that waited an hour longer.
How many people are in a B2B buying group?
Forrester’s 2026 research puts a typical decision at 13 internal stakeholders and 9 external influencers. 6sense reports buying groups of ten or more.
Does predictive lead scoring work?
It can, with enough clean history. A 2024 academic review of 44 studies describes the methods in use but gives no single lift figure, and we found no vendor-neutral study that does.
Are email opens a reliable scoring signal?
No. Apple Mail Privacy Protection loads tracking pixels automatically, and Apple Mail accounts for most tracked opens. Use replies, clicks, and meetings.
The takeaway
Good statistics on lead scoring are rarer than the roundups suggest. The ones that hold up say the same thing from different sides: decisions are made by groups, buyers arrive late, data is incomplete, and reps are short on time. Use benchmarks to question your own numbers, quote them with their source and year, and build the model on your own wins.