B2B Lead Generation Benchmarks for 2027: The Real Numbers to Plan Your Year
Short answer: in 2026, an average LinkedIn campaign gets slightly more than 1 in 4 people to accept the connection request and about 10% of messages answered; in cold email, the average reply rate is around 3.4%. Across the 9,193 LinkedIn conversations Vrand managed between June and November 2026, 52% of replies came only on the second message or later, 91% arrived Monday through Friday, and 15.4% of conversations with a reply became an opportunity in the pipeline. For 2027 planning, the number that matters most is not the reply rate but how many contacts you need to get one meeting.
Every December, sales teams build next year’s plan on numbers nobody can trace: “email gets 1% replies”, “LinkedIn converts better”, “you need seven follow-ups”. In this article we pulled together the benchmarks published in 2026 by studies with large samples, compared them with each other and added data from Vrand’s own operation. The goal is to help you build your 2027 plan on defensible ranges instead of repeated phrases.
A warning before we start: almost every prospecting benchmark comes from a company that sells prospecting tools and measures its own customers. They are useful as a reference, but they are not neutral and not necessarily comparable with each other. That is why we say where each number comes from, what sample it uses and what exactly it measures.
The numbers in one table
| Metric | Value | Source and sample |
|---|---|---|
| LinkedIn connection request acceptance | 28.5% | Expandi, 13.2 million invitations, May 2025 to April 2026 |
| LinkedIn message reply rate | 10.4% | Expandi, 6.7 million messages |
| Reply rate to the invitation note | 3.0% (fell from 3.5% to 2.2% over the year) | Expandi |
| LinkedIn connection to meeting | 1.3% | Belkins, 14,077 contacts across 34 projects, 2025 |
| Average cold email reply rate | 3.43% (top quartile: 5.5%+) | Instantly, 2026 report on 2025 data |
| LinkedIn replies that arrive on the 2nd message or later | 52.1% | Vrand, 6,639 conversations with a complete sequence |
| LinkedIn replies Monday to Friday | 91.5% | Vrand, 871 replies received in real time |
| Conversations with a reply that reach the pipeline | 15.4% | Vrand, 1,416 of 9,193 conversations |
| Software buyers who start research in an AI chatbot | 51% | G2, 1,076 buyers, March 2026 |
| US Google searches that end without a click | 68% | SparkToro with Similarweb data, Jan to Apr 2026 |
| Marketing budget as a share of revenue | 7.8% (7.7% in 2025) | Gartner CMO Spend Survey 2026, 401 CMOs |
1. What we saw in 9,193 LinkedIn conversations
Between June and November 2026 we managed 9,193 LinkedIn conversations in which the prospect replied, for 26 B2B companies and from 64 profiles. This is not survey data: these are real conversations recorded in our operations platform. Three findings seem useful to any team planning 2027.
More than half of replies come from the follow-up
We took the 6,639 conversations for which we have the complete message order and looked at how many of our messages came before the prospect’s first reply.
| Our messages before the first reply | Conversations | % |
|---|---|---|
| 1 | 3,166 | 47.7% |
| 2 | 2,043 | 30.8% |
| 3 | 1,054 | 15.9% |
| 4 or more | 358 | 5.4% |
| None (the prospect wrote first) | 18 | 0.3% |
Put differently: if a LinkedIn sequence stops after the first message, it leaves out more than half of the people who were going to reply. The third message still brings in one reply in six. From the fourth on, the contribution drops to 1 in 19, and that is where the cost of insisting (annoying people, wearing out the account) starts to outweigh the gain.
This is a different pattern from email. According to the coverage that cites it, Instantly’s report attributes 58% of replies to the first email. On LinkedIn, the first message accounts for less than half. One reasonable hypothesis is that on LinkedIn the first message lands right after the invitation is accepted, when the person still doesn’t know who you are, and the second one reaches someone who has already looked at your profile.
Replies cluster midweek and before 3 p.m.
To measure when people reply we used only the 871 replies we received in real time, because for those the recorded time is the time the prospect actually wrote.
| Day | % of replies |
|---|---|
| Monday | 16.4% |
| Tuesday | 17.7% |
| Wednesday | 25.1% |
| Thursday | 16.6% |
| Friday | 15.7% |
| Saturday and Sunday | 8.4% |
Wednesday accounts for one in four replies. By time of day (Buenos Aires time, UTC−3), 54% come in between 9 a.m. and 3 p.m., and 20% after 6 p.m. Two practical consequences: sending on weekends makes little sense, and a reply that arrives at 1 p.m. on a Wednesday deserves an answer the same day, because the prospect is still at their screen.
Be careful about generalizing: our clients prospect in several countries, and the time window shifts depending on where the buyer is. The day-of-week pattern is more stable than the hour pattern.
One in six or seven replies becomes an opportunity
Of the 9,193 conversations with a reply, 1,416 (15.4%) ended up linked to an opportunity in the pipeline. The rest are split between negative replies, “not now”, referrals to someone else and polite thank-yous with no interest.
That is the most useful number for planning, because it turns “replies” into something the sales team can actually work. If your 2027 plan counts replies as if they were opportunities, it is overestimating the pipeline about six times.
2. LinkedIn: stable acceptance, weaker first touch
The largest study published in 2026 is Expandi’s: 13.2 million invitations and 6.7 million messages sent from 13,302 accounts between May 2025 and April 2026. Its averages:
- Invitation acceptance: 28.5%. It stayed between 28% and 30% almost all year.
- Reply rate to the invitation note: 3.0%. This is the metric that fell the most: from 3.5% in May 2025 to 2.2% in April 2026, a 37% relative drop.
- Message reply rate: 10.4%. Stable, between 10% and 11%.
The decline in note replies matches what we see in the operation: pitching in the invitation works worse and worse. The conversation starts after you connect.
The Belkins study adds a data point that challenges a very common habit. Comparing invitations with and without a personalized note:
| Invitation | Acceptance | Reply |
|---|---|---|
| With a personalized note | 25.3% | 8.2% |
| Without a note | 27.6% | 5.3% |
The note lowers acceptance slightly and raises replies considerably. If your goal is to start conversations rather than collect contacts, write the note.
Belkins also publishes the full contact-level funnel, which is the closest thing to a planning number: out of every 100 invitations, 18.7 become connections; out of every 100 connections, 17.6 reply and 1.3 end in a meeting.
Differences by industry are large. In Expandi’s data, staffing and recruiting reach 36.5% acceptance and 18.9% message replies; software, 27.5% and 8.8%. If you sell to software companies, your benchmark is below average.
One practical limit to build into the plan: LinkedIn doesn’t publish an official figure, but the most cited reference is about 100 invitations per week per profile. That ceiling, more than the message copy, is what determines how much volume a single profile can move.
3. Cold email: the average is falling and deliverability rules
Instantly’s 2026 report, based on its users’ 2025 activity, puts the average reply rate at 3.43%. The top quartile exceeds 5.5% and the top 10% goes above 10.7%. Other studies with smaller samples report lower values: Visionary Marketing reports a 2.1% median in 2026, versus 4.7% in 2022, across 28,400 leads from 47 B2B programs.
Two caveats for reading those numbers correctly:
- Not everyone measures the same thing. Some count any reply, including automatic and negative ones; others count only human replies. Before comparing yourself, check what your tool counts.
- Deliverability matters more than copy. Since 2024 Google and Yahoo, and since May 2025 Microsoft, require senders of more than 5,000 emails a day to have SPF, DKIM and DMARC authentication, a spam complaint rate below 0.3% and one-click unsubscribe. A domain that doesn’t comply doesn’t reach the inbox, and no subject line fixes that.
4. The buyer: researches with AI and arrives with preferences
Prospecting numbers are moving because buyers changed how they research.
- 51% of B2B software buyers start their research in an AI chatbot and 71% use one at some point in the process, according to a G2 survey of 1,076 buyers in March 2026.
- 72% of buyers always or almost always verify what AI tells them, up from 58% a year earlier, according to TrustRadius’s 2026 report. They use AI to explore, not to decide.
- Buying groups are around 10 people and holding steady, according to coverage of 6sense’s 2026 report, which also argues that buyers arrive at evaluation with preferences already formed.
- 68% of Google searches in the United States ended without a click in the first four months of 2026, according to SparkToro. In 2019 it was about 49%. Ahrefs measured that when an AI overview appears, the top result gets 58% fewer clicks.
For prospecting this has a direct consequence: by the time your message arrives, part of the buying group has already asked an AI who solves their problem. If your company doesn’t show up in those answers, your first message has to do all the introduction work on its own. That is why visibility in search engines and AI assistants is now part of lead generation, not a separate topic.
5. Budgets: more AI, the same percentage
Gartner’s survey of 401 marketing leaders, carried out between January and March 2026, puts marketing budgets at 7.8% of revenue, almost the same as 7.7% in 2025. What changes is the mix: on average, 15.3% of the budget already goes to AI. And 56% of CMOs say they don’t have enough budget to execute their strategy.
The takeaway for 2027: don’t expect more money; expect to be asked what the money you already have is producing. A plan that can show how many contacts it takes to get a meeting, and what each meeting costs, has a better chance of defending its budget than one that reports messages sent.
How to use these numbers in your 2027 plan
Benchmarks are for working backwards: start from how many meetings sales needs and arrive at how much volume it takes. An example using the Belkins funnel:
- Goal: 20 meetings a month from LinkedIn.
- Connections needed: at 1.3% connection to meeting, about 1,540 connections a month.
- Invitations needed: at 18.7% acceptance, about 8,200 invitations a month, roughly 1,900 a week.
- Profiles needed: with a ceiling of about 100 invitations per week per profile, about 19 profiles.
It’s an example built on other people’s averages, not a prediction, but it shows something almost no plan accounts for: with a single LinkedIn profile and average conversion, 20 meetings a month is impossible. There are three levers, and it’s worth deciding which one you’ll pull before January:
- Better conversion: a better-segmented list and a message that explains why you are writing now raise every rate in the funnel. It’s the cheapest lever and the slowest.
- More profiles: adding team members or company spokespeople multiplies the invitation ceiling. It requires coordination so two profiles don’t write to the same person.
- More channels: combining LinkedIn with email spreads the volume and reaches people who don’t reply on one channel. It requires email deliverability to be in order.
The data in section 1 suggests three more adjustments that cost nothing: LinkedIn sequences of at least three messages, sending Monday through Friday with a focus on Tuesday, Wednesday and Thursday, and counting opportunities instead of replies. If you have stalled opportunities from this year, it’s also worth reviewing them before going after new ones: we cover that in how to reactivate B2B opportunities, and the broader trends for next year in B2B year-end 2026: trends for 2027.
Frequently asked questions
What is a good LinkedIn reply rate in 2026?
The average message reply rate is around 10% (Expandi, 2026) and invitation acceptance is between 25% and 29%. It varies a lot by industry: software is below average and staffing is well above it. Comparing yourself with your industry is more useful than comparing with the overall average.
And for cold email?
The most cited published average is 3.43% (Instantly, 2025 data), with the top quartile above 5.5%. Before comparing yourself, make sure your emails reach the inbox and that your tool counts replies the same way.
How many follow-ups should you send on LinkedIn?
In our conversations, 47.7% of replies came on the first message, 30.8% on the second, 15.9% on the third and 5.4% on the fourth or later. A three-message sequence captures more than 94% of replies; beyond that the gain is small and the risk of wearing out the account goes up.
Which day do prospects reply most on LinkedIn?
In our data, Wednesday accounts for 25.1% of replies and the weekend for only 8.4%. The 9 a.m. to 3 p.m. window (Buenos Aires time) gathers 54%, although the hour depends on the buyer’s country.
Do these benchmarks apply to Latin America?
The global studies have samples mostly from the United States and Europe; none of those cited publishes separate results for the region. Our own data comes from campaigns run from Latin America, but includes buyers in several countries. Use them as a starting range and measure your own funnel from the first month.
The number that matters in 2027
A benchmark won’t tell you how much you’ll sell, but it will tell you whether your plan is possible. If hitting the goal requires conversion rates nobody achieves or a volume your profiles can’t move, it’s better to find out in December than in March.
At Vrand we run B2B lead generation and appointment setting with data like the data in this article: sequences, profiles, follow-up and pipeline measured end to end. If you want to build your 2027 plan on your own numbers, let’s talk.
How we measured and sources
Vrand’s own data. LinkedIn conversations recorded in our operations platform between June 16 and November 13, 2026, for 26 B2B companies and 64 profiles, excluding Vrand’s own campaigns. They include only conversations in which the prospect replied, so they can’t be used to calculate a reply rate. The sequence analysis uses the 6,639 conversations with a complete message order. The day and time analysis uses the 871 replies received in real time, in Buenos Aires time. A conversation counts as an opportunity if it was linked to a pipeline record. The data is aggregated and doesn’t identify clients or people.
External sources (accessed November 13, 2026):
- Expandi, LinkedIn Outreach Benchmarks 2026: 13.2 million invitations, May 2025 to April 2026.
- Belkins, LinkedIn B2B outreach statistics 2026: contact-level funnel and comparison with and without a note.
- Instantly, Cold Email Benchmark Report 2026: average reply rate and distribution by sequence step.
- Visionary Marketing, B2B Cold Email Benchmarks Study.
- Google, email sender guidelines.
- Demand Gen Report on G2’s software buyer survey, 2026.
- TrustRadius, B2B Buying Disconnect 2026.
- 6sense, 2026 Global B2B Buyer Experience Report.
- Search Engine Land on SparkToro’s 2026 zero-click study.
- Ahrefs, AI Overviews and clicks.
- Gartner, 2026 CMO Spend Survey.
The prospecting studies cited come from companies that sell prospecting tools and measure their own users. The calculation examples combine published averages and are not a prediction of results for any company, industry or region.