Most lead stats that get passed around have no sample, year or named source. The few that trace back to a real study fall into two groups. The famous speed-to-lead multiplier comes from research published in 2011. The Gartner, McKinsey and Forrester numbers are recent, but they describe how buyers buy. None of them tells you how many of your leads will convert.
For outbound results I have not found an independent, current study with a published sample. Reply, meeting and close rates come from agencies and tool vendors reporting on their own campaigns, my team included. So I read every one of them, mine too, as one company's range.
Which lead stats can you actually trust?
Run every number through four tests before it goes into a plan or a board deck:
- Sample size: how many people or companies, stated in the source.
- Fieldwork year: when the data was collected (fieldwork means the survey or audit itself), which can be years before the article went up.
- B2B share: whether the sample is business buyers or mixed with consumers.
- Who ran it: an independent researcher, an analyst firm, or a vendor selling the answer.
Here are widely repeated lead stats, graded against their sources:
| Stat as usually quoted | Sample | Fieldwork | B2B share | Who ran it | Verdict |
|---|---|---|---|---|---|
| Contact a lead within an hour and it is almost 7 times likelier to qualify than an hour later (HBR, 2011) | Audit of 2,241 US companies, plus 1.25 million leads at 42 firms | 2011 or earlier | 13 of the 42 firms in the leads study | Researchers published in Harvard Business Review | Quote with caveat: old and consumer-heavy |
| The "call within five minutes" stats credited to the Lead Response Management Study, 2007 | Over 15,000 web leads and over 100,000 call attempts at 6 companies | 3 years of data, published 2007 | Not given | InsideSales.com and Dr. James Oldroyd (MIT) | Quote with caveat: old, and it measured contact and qualification, not closed sales |
| 67% of B2B buyers would rather buy without a rep (Gartner, 2026) | 646 B2B buyers | August to September 2025 | All B2B | Gartner | Quote as a stated preference, not behavior |
| Buyers use an average of ten channels (McKinsey 2026 Global B2B Pulse) | Nearly 4,000 decision-makers across 13 countries | Published 2026 | All B2B | McKinsey | Quote |
| A typical decision includes 13 internal people and 9 external influencers (Forrester, 2026) | Not given in the release | Published 2026 | Business buyers | Forrester | Quote with caveat: no sample |
| 86% of B2B purchases stalled (Forrester's 2024 report) | Not given in the release | Published December 2024; the 2026 release does not repeat it | Business buyers | Forrester | Quote with its year only |
| Agency reply-rate benchmarks, mine included | Each agency's own campaigns | Usually undated | B2B | The agency itself | Quote as one agency's range |
When the same stat shows up on two lists with two different values, open the original before you quote either one. The four tests work just as well on tool vendors. I ran them on AI SDR results, checked claim by claim.
What does a real lead funnel convert at, stage by stage?
Each stat above describes one stage at most. A lead is only worth what the whole funnel turns it into. So here are two of my team's campaigns, counted at every step. More funnels are on our case studies page.
Campaign A ran from a single LinkedIn account for 30 days. Campaign B ran for one month for a German software development company, with cold email from 20 addresses plus LinkedIn. Its 453 replies are email replies, while its 34 booked calls came from both channels.
| Stage | Campaign A | Ratio | Campaign B | Ratio |
|---|---|---|---|---|
| Contacted | 1,200 | 9,000 | ||
| Replies | 289 | 24% of contacted | 453 | 5% of contacted |
| Interested | 21 | 7% of replies | n/a | |
| Booked calls | 16 | 5.5% of replies | 34 (both channels) | 7.5% of email replies |
| Showed | 12 | 75% of booked | 28 | 82% of booked |
| Closed | 3 | 25% of showed | 4 contracts | 14% of showed |
Campaign B's 4 contracts were worth EUR 1,152,000. Campaign A went 1,200 contacts -> 3 closed deals in 30 days.
Watch the step right after the reply. In both campaigns, fewer than 8 in 100 replies became a booked call. That is the biggest drop in either funnel after the first reply, and none of the studies in the table covers it.
Reply rate is the number people brag about. On its own it says little about revenue. A reply can be a plain no or a "not this year." Founders look at the contracts that got signed. I look at the repliers who never made it to a call.
Campaign A ran on LinkedIn alone and replied at 24%. Campaign B's cold email replied at 5%, well above the 1% I plan around for email. A LinkedIn reply rate is no target for an email campaign, so plan each channel on its own numbers.
Even inside one channel the rate swings. Same message, two companies. One got 2% replies. The other got 7.5%. The difference was the sender.
Is speed to lead still the number to watch?
Only if you have inbound leads waiting for an answer. In the HBR audit, 37% of companies answered their test lead within an hour and 23% never answered at all.
Those that did reply within 30 days took 42 hours on average. Contact in the first hour made a lead over 60 times likelier to qualify than contact after 24 hours or more. Qualified meant "having a meaningful conversation with a key decision maker."
Newer data points the same way, though all of it comes from vendors auditing a market they sell into. A 2026 compilation of lead response studies cites a 2024 RevenueHero audit of 1,000 B2B software companies: 63.5% never replied to a demo request.
I could check that audit's method at the source. RevenueHero's own write-up of the audit explains the method. The team submitted a demo request, then took the timestamp of each company's first response. Companies that replied by hand averaged 2 days, 3 hours and 11 minutes. That is how long a warm lead waited.
The same compilation cites another vendor audit at 11 hours 54 minutes on average. I could only reach that one second-hand, so treat it as directional. Results depend on sample, channel and whether auto-replies count, so there is no single current average.
Use the HBR work to argue that slow answers lose leads. Do not use its multiplier as a forecast for a B2B pipeline today. Speed to lead is about answering a lead the buyer created. Outbound creates the lead itself, so other numbers apply:
| Stat family | Describes | Use it for |
|---|---|---|
| Speed to lead (HBR) | Responding to inbound leads | Routing and response rules |
| Buyer surveys (Gartner, McKinsey, Forrester) | How companies buy | Offer, proof and channel choices |
| Reply, meeting, show-up and close rates | Creating leads through outbound | Planning volume and fixing stages |
What do B2B buyer behavior stats mean for outbound?
They describe the person reading your message. In Gartner's release, 45% of buyers said they used AI during a recent purchase (survey fielded in 2025).
McKinsey's 2024 report shows the channel count doubled, from five in 2016 to ten. The 2026 Pulse splits interactions roughly in thirds between in-person, remote and digital.
Forrester's two 2026 counts add up to 22 people around one decision. One reply is a single voice in that group. Forrester does not connect this to stalled deals, but a group that size is a plausible reason a deal goes quiet after a good first call.
Nobody buys from an email. Purchases follow a check on the sender. By the time someone replies, they have usually already checked your profile, your site and your proof. That is why a weak profile sinks a perfect message.
What reply rate should a cold email campaign expect?
In my team's campaigns, cold email usually replies at about 1%. At thousands of emails a day it is often nearer 0.5%. Between 10% and 50% of replies usually turn into booked meetings. The two funnels above sit below that range because they count every reply, nos included. Count only interested replies and Campaign A booked 16 of 21.
Our tightest, heavily personalized campaigns have reached 15% to 17% replies, with roughly 75% of repliers interested. The rule underneath: the bigger the volume, the lower the averages.
As an outside view, Instantly's 2026 cold email benchmark report shows an average reply rate of 3.43%, a top quartile at 5.5% and elite campaigns above 10%. It is a vendor view, built on billions of interactions across thousands of workspaces on its own platform. It is still one company grading its own customers.
Set the month's goal at 50+ conversations started. In outbound, the call grows out of a conversation and the client grows out of the call. Which of these numbers belong on a weekly dashboard is a question for your lead generation KPIs.
How many contacts does it take to fill a sales calendar?
My team plans volume from two inputs before anything goes out: the reply rate and the share of replies that book.
My working rule is a list of 10,000 contacts or more each month to get 30 to 35 calls. It is a planning rule from our campaigns, and your market may behave differently. Which channels should create the leads at all is the question behind sales lead generation, and it comes before any of these ratios.
I add volume last, never first. A funnel that doesn't convert with small numbers will never convert with big ones. Find the stage that breaks and fix that first. Then add accounts and volume. That order is how we run B2B lead generation for clients.
Why do booked calls overstate lead results?
A booked call is a promise. In the two funnels above, between one booked call in four and one in six never happened. Booking a call is one job and closing the deal is another. If held calls are not closing now, a hundred more booked calls will only close at a lower rate.
My sanity line is about 20% of held calls closing for a high-ticket B2B service, and it moves by niche. Campaign A sat above it and Campaign B below.
If many booked calls never happen, fix your confirmations and what buyers find when they check you. If few held calls close, work on the sales call itself before buying more volume.
What to check in your own funnel this week
Pull your own counts for the last 90 days: calls booked, calls held, deals signed. Compute held divided by booked first, then signed divided by held. The lower ratio is your first job. Fix that stage before you add a single account or contact.
Not sure which of your own numbers to trust? Walk me through last quarter on a call and we can set it against the benchmarks.
Questions founders ask about lead stats
What does "lead data" mean?
Lead data is the record behind one lead: the person, company, title, where the lead came from, the date of first contact, and every stage it reached with a date on each. The dates matter most. Without them you cannot tell whether a ratio got better this month or was always that way.
How do you track leads from first contact to closed deal?
Give every funnel stage its own CRM field and record every reply, including the "not now" ones. Which CRM you use matters far less. Tag each lead with its sender and offer so you can split ratios later. Review stage ratios monthly, because one broken stage can hide inside a total that looks healthy.
How old can a lead statistic be before it stops being useful?
For anything in a forecast, prefer data collected in the last two years and print the fieldwork year next to the number. Publication date misleads: a survey released in 2026 may have been fielded a year earlier, so read the methodology note. An old number can show direction, but keep it out of your forecast math.
Should a small B2B team build its own lead benchmarks?
Yes, as soon as enough contacts have gone through one offer to show a pattern at every stage. Your own stage ratios beat any industry average because they reflect your list, your sender and your buyers. Keep one offer and one audience per test, or the numbers blend and say nothing useful about either.
