Is prospecting the same as lead generation?

No, although they overlap for part of the motion. Lead generation is the whole system that turns a defined segment into a qualified conversation: the segment definition, the list, the message, the sending infrastructure, the qualification bar and the handover. Prospecting is one activity inside that system, the manual act of selecting a specific named account, finding the right person in it and reaching out to them. Every prospecting effort happens inside a lead generation system, whether or not anyone has designed one.

What is the difference between prospecting and lead generation?

Ownership and denominator. Prospecting is owned by an individual rep and measured per account: which accounts did you work, how well did you research them, did you follow up. Lead generation is owned by the business and measured per stage with a rate: accepted divided by sent, replies divided by messages sent, meetings held divided by meetings booked. If a metric has a person's name attached to it, it is probably a prospecting metric. If it has a denominator attached to it, it is probably a lead generation metric.

What is sales prospecting?

Sales prospecting is the deliberate selection of and outreach to named accounts that have not raised a hand. The rep chooses the account, identifies who inside it has the problem, finds a specific reason to reach out now, sends the first touch and keeps following up. The unit of work is one account. That is what separates it from a campaign, where the unit of work is a segment and the copy is written once for everyone in it.

Does prospecting come before or after lead generation?

Neither, because they are not sequential stages. Prospecting sits inside lead generation. The system defines the segment and builds the list before any prospecting happens, prospecting picks the accounts and opens the conversations, then the system takes over again at qualification and handover. Teams that treat prospecting as the first step usually skip the segment definition entirely, which is why their reps end up working a list nobody can defend.

Who should own prospecting, sales or marketing?

Sales owns prospecting because the unit of work is a named account and the judgement involved cannot be centralised. Marketing or growth normally owns the parts of lead generation that scale: the segment definition, the list, the sending infrastructure and the message testing. The failure mode we see most often is the reverse, where marketing hands over a list nobody in sales believes in, and sales is then measured on activity against it. Agree who owns which stage before you agree a target.

How should I measure prospecting and lead generation separately?

Measure prospecting on coverage and follow-through: accounts touched divided by accounts assigned, and accounts reopened divided by accounts parked last quarter. Measure lead generation on stage conversion with explicit denominators: accepted divided by sent, replies divided by messages sent, meetings held divided by meetings booked. Published reference points, reviewed September 2026, include acceptance of 28.5% across 13,218,869 requests (Expandi) and 26% (Belkins), and reply rates of 7.2% across 15.1 million contacts (Belkins) and 10.4% across 6,730,447 messages (Expandi). They come from different datasets with different denominators, so use them to sanity check one stage at a time and never multiply them together.

Definitions

Prospecting vs lead generation: where one ends and the other starts

Short answer

Lead generation is the system that turns a defined segment into a qualified conversation: the segment, the list, the message, the sending infrastructure, the qualification bar and the handover. Prospecting is one act inside that system, choosing a specific named account, finding the right person in it and reaching out. The useful split is not job titles, it is ownership and denominator: prospecting is owned by a rep and counted per account, lead generation is owned by the business and measured as a rate with a stated denominator. Confuse the two and you will report activity as performance, which hides the stage that is actually leaking.

Written by WeeGrow. Every benchmark on this page carries its publisher, its denominator and, where the publisher stated one, its sample size. They were last reviewed in September 2026 against the figures published across this site, rather than re-fetched from the publishers on that date. We publish no performance statistics of our own that we cannot evidence.

At a glance

Unit of work: One named accountOne segment
Owned by: The repThe business
Counted as: Accounts workedA rate, with a denominator
Fails because of: Coverage and follow-upTargeting and message

Left column is prospecting, right column is lead generation. Two of the ten stages further down belong to both, and those two are where most reporting goes wrong.

the part other pages leave out

“Requests sent is a prospecting number. It is not a lead generation result, and reporting it as one hides the leak.”

Activity counts have no denominator, so they cannot tell you which stage failed. Every metric on this page has one.

01The stage map

Ten stages, and which of the two owns each one

Most pages on this query swap two definitions, attach a job title to each and stop. That is not a usable answer, because the two words describe different scopes of the same motion rather than two competing activities. Here is the whole motion split into stages, with the stage marked as prospecting, lead generation or both, who normally owns it and what it is honestly measured on. Read the final column first.

Stage	Which is it	Who typically owns it	What it is honestly measured on
01. Define the segment you are going after	Lead generation	Founder, or whoever owns go to market. Never a junior rep.	Nothing numeric yet. The test is whether three existing customers recognise themselves in the sentence you wrote. If they do not, every number downstream is noise.
02. Build the list of accounts that fit	Lead generation	Growth, RevOps or the person who owns the data stack.	Coverage: accounts on the list divided by accounts that exist in the defined segment. A list of 200 inside a market of 4,000 is a 5% coverage problem, not a messaging problem.
03. Pick this account, and this person inside it	Prospecting	The rep, or the founder doing their own outbound.	Worked share: accounts actually touched divided by accounts assigned. This is the first place a good list quietly dies.
04. Research the account and find a reason to write	Prospecting	The rep. It cannot be bought as a service without losing the point of it.	Nothing. This is the one stage you assess by reading the output, not by counting it. Any metric you invent here turns into research theatre.
05. Write the message and the sequence around it	Lead generation	Whoever owns positioning, with the rep who reads the replies in the room.	Reply rate at sequence level: replies divided by messages sent. Judged per variant, never per rep.
06. Send the first touch, request, email or dial	Both	The rep executes. The system supplies the list, the copy and the sending infrastructure.	Delivery: first touches that actually landed divided by first touches attempted. Bounces, blocks and hit sending limits all show up here. Raw counts of requests sent, emails delivered and dials made carry no denominator, so they are a capacity planning input, not a measure of the stage, and they are the number most often reported as lead generation performance.
07. Earn permission to have the conversation	Lead generation	Nobody feels like they own it, which is why it leaks. In practice the system owns it: targeting, profile, deliverability.	Acceptance or connect rate: accepted divided by sent. Published LinkedIn acceptance sits at 28.5% across 13,218,869 requests (Expandi) and 26% (Belkins), with the range across industries running 17.5% to 40.1% (Expandi). Compare yourself with your own industry, not with the top of that range.
08. Handle the reply and qualify the person	Both	The rep, working to a qualification bar somebody senior wrote down.	Reply rate, replies divided by messages sent: 7.2% across 15.1 million contacts (Belkins) and 10.4% across 6,730,447 messages (Expandi). Then your own rate of qualified conversations divided by replies handled.
09. Book the meeting and hand it over	Lead generation	Sales leadership owns the definition of a qualified meeting. The rep owns the booking.	Qualified meetings divided by replies handled, and meetings held divided by meetings booked. The second one is the honest number and almost nobody reports it.
10. Reopen the accounts that said not now	Prospecting	The rep, on a named account basis, with the CRM doing the remembering.	Accounts reopened per quarter, divided by accounts parked last quarter. Measured quarterly. Measured weekly it becomes pestering.

Benchmarks in the table are published by the named vendor, quoted with the sample size wherever that vendor published one, and were last reviewed in September 2026. Belkins publishes no sample size for its 26% acceptance figure, so none is implied here. Belkins also reported 27.6% acceptance for requests sent without a note against 25.3% with one, across 11.5 million requests, which is a useful reminder that the personalisation you agonise over at stage 04 is not what moves stage 07.

The tell: can it be done twice the same way?

If a stage produces the same output whether it runs for ten accounts or ten thousand, it belongs to lead generation and should be built once, properly. Segment definition, list building, sequence design and sending infrastructure all pass that test. If a stage produces a different output for every account because a human made a judgement about that account, it is prospecting, and trying to systematise it strips out the only thing that made it work.

The two stages that belong to both

Stage 06, the first touch, and stage 08, handling the reply, are shared. The rep does the work, the system decides whether the work had a chance. Almost every argument between sales and marketing about lead quality is really an argument about who owns these two stages. Settle it on a whiteboard before you settle it in a quarterly review, and write the qualification bar down while nobody is defending a number.

02The expensive confusion

Reporting activity as performance hides the stage that is leaking

Here is the pattern, and it costs real money. A team measures what is easy to count, which is stage 06: requests sent, emails delivered, dials made. Those are prospecting activity counts with no denominator attached. At the end of the month they get reported upwards as lead generation performance, usually next to a meeting count. The board sees effort at one end and an outcome at the other, with eight stages in between reported as nothing at all.

So when the meeting count drops, nobody can say why, and the default diagnosis is that the reps need to do more. Volume goes up, the leaking stage stays exactly where it was, and the cost of the motion rises without the output moving. We have watched teams add a second sender to a campaign whose problem was that stage 01 had never been written down.

Three diagnoses, each pointing at a different stage. High activity, falling acceptanceis a stage 01 to 03 problem: you are contacting the wrong people, and no amount of extra sending will fix it. Published acceptance runs 28.5% across 13,218,869 requests (Expandi) and 26% (Belkins), and Expandi reports a range of 17.5% to 40.1% across industries. That band describes markets, not effort, so a low rate is not by itself a verdict on your targeting: the floor of the band is somebody’s perfectly normal industry. What does convict the targeting is movement. If acceptance falls as volume rises, or sits well below what comparable senders into your own industry see, the list and the selection are the problem, not the hours.

Good acceptance, almost no replies is a stage 05 problem. The people were right, the message was not. Published reply rates are 7.2% across 15.1 million contacts (Belkins) and 10.4% across 6,730,447 messages (Expandi), both counted as replies per message sent. And plenty of replies, no meetings is a stage 08 to 09 problem: either the qualification bar was never written down, or the offer does not survive the first real conversation.

Those figures come from different datasets with different denominators, so use them to sanity check one stage at a time and never multiply them into a forecast. If you want the full stage by stage arithmetic with every rate carrying its source and sample, that lives on our LinkedIn lead generation funnel breakdown, which exists precisely because blended benchmarks disagree so violently with each other.

Three questions that expose it

01Ask for one number with its denominator. If the answer to "what is our acceptance rate" is a count rather than a fraction, nobody is measuring the stage, they are measuring the effort.
02Ask which stage lost the most people last month. A team with a working model answers in seconds. A team reporting activity will restate the meeting count instead.
03Ask who owns stage 07. If the room goes quiet, you have found the leak, because permission to have the conversation is nobody's job title and everybody's problem.

Every benchmark quoted on this page is attributed to its publisher, with the sample size it was reported against wherever the publisher stated one. Where no published figure exists for a stage, this page says so rather than estimating one.

03The metric rule

The metric to ask for at each stage, and its denominator

One rule covers the whole page: never accept a number without its denominator. A count tells you somebody was busy. A rate tells you whether the stage worked. Each row is numbered back to the stage map above, so a rate that looks wrong here points at a named stage and a named owner. The final column is the one to memorise, because it is what a well-meaning report usually hands you instead.

Stage	The question to ask	The denominator	What you will be handed instead
02. Coverage	How much of the defined segment is on the list at all?	Accounts on the list, divided by accounts in the segment.	A number of contacts with no market size attached. 12,000 contacts means nothing until you know whether that is the whole market or 4% of it.
03. Worked share	How many accounts were genuinely worked, not just loaded?	Accounts touched, divided by accounts assigned.	Total requests sent, quoted on its own. That is a stage 06 activity count with no denominator: it tells you the team was busy, not how much of the assigned list anybody actually worked.
05. Message	What share of messages earned a reply, by variant?	Replies, divided by messages sent.	A reply rate quietly calculated against accepted connections instead. At published acceptance rates that denominator is about a quarter of the size, so the number looks roughly three to four times better and is comparable to nothing anyone has published.
06. Delivery	What share of the first touches we attempted actually landed?	First touches delivered, divided by first touches attempted.	A weekly send total. Sending limits, bounces and blocks are invisible in a total, and they are the cheapest thing on this page to fix.
07. Permission	What share of the people we contacted let us start a conversation?	Accepted or connected, divided by sent.	Acceptances quoted as a raw count. 400 acceptances from 3,000 requests and from 1,000 requests are two completely different businesses.
08. Qualification	What share of replies turned into a conversation worth having?	Qualified conversations, divided by replies handled.	Counting every positive reply as a lead. Interest is not qualification, and the gap between them is where forecasts go wrong.
09. Show rate	What share of booked meetings actually happened?	Meetings held, divided by meetings booked.	Meetings booked, reported alone, at the end of the month. The show rate is the number that decides whether the whole motion paid for itself.

Stages 01 and 04 are missing from that table on purpose: neither has an honest number, and inventing one is how research theatre starts. Stage 10 is missing because reopening parked accounts is measured quarterly, so it belongs in a quarterly review rather than a monthly report. Of the seven rates that are here, two have published reference points and five do not. Acceptance and reply rate can be sanity checked against the figures above. Coverage, worked share, delivery, qualification and show rate are specific to your market, your sending setup and your qualification bar, so the only honest benchmark is your own last quarter. Anyone quoting an industry standard for those five is quoting something nobody has measured at scale. None of these rates are an input to our ROI calculator, which runs the other way round: you give it a deal size, a close rate and a monthly budget, and it projects the meetings, pipeline and revenue that budget could produce. Treat it as a sizing exercise, and treat the rates above as the thing you instrument once the spend is live.

04Honest fit

Which of the two you should be fixing this quarter

The two failure modes look identical from the outside, an empty calendar, and have opposite fixes. Use deal size and the size of the market you can actually sell into to decide which one you are looking at before you spend anything. The two lists below meet at the same line, $25k deals into a market of a few hundred accounts, so every business lands in one of them. If you recognise yourself in both, fix the system first: worked share is worth nothing against a list nobody can defend.

Fix prospecting first if

Deals are worth $25k or more and the whole market you can sell into runs to a few hundred accounts, not thousands
The list is defensible but less than half of it has actually been touched this quarter
You have a rep or a founder who knows the market and is simply not getting through enough accounts
Accounts that said not now last quarter have never been reopened

At this shape the answer is usually fewer accounts worked far better, which is the logic behind account based marketing, not a bigger list.

Fix the lead generation system first if

You cannot write down in one sentence who you sell to and have three customers agree
Acceptance or connect rates fall as you send more, or sit well below what comparable senders into your own industry see
Replies arrive but nobody wrote down what makes a meeting worth taking
Deals are under $25k, or the market runs to thousands of accounts, so per account craft cannot pay for itself at that volume

This is the work we do with clients: the segment, the sequences and the measurement built once and handed over, starting with LinkedIn as the channel that opens the conversation.

Where we are the wrong answer

You want someone to do the prospecting for you, judgement included. We are done with you: your team still picks the accounts that need a human decision.
Your buyers are not reachable on LinkedIn, email or the phone. Field trades, walk-in retail and parts of public sector buying do not work this way.
You want a published rate card today. We scope and quote after a free audit, and if that is a dealbreaker, a done-for-you agency with public pricing is the better fit.
Your system already works and you only want more volume at the lowest unit cost. At that point an in-house team is usually the better economics.

If what you actually need is the system designed and measured properly, that is B2B lead generation. If the system is fine and you simply do not have the hours to work the accounts, an outsourced SDR team is the narrower brief. Most teams who arrive certain they need the second one turn out to need the first.

Get a free pipeline audit

05FAQ

Prospecting vs lead generation, answered

No, although they overlap for part of the motion. Lead generation is the whole system that turns a defined segment into a qualified conversation: the segment definition, the list, the message, the sending infrastructure, the qualification bar and the handover. Prospecting is one activity inside that system, the manual act of selecting a specific named account, finding the right person in it and reaching out to them. Every prospecting effort happens inside a lead generation system, whether or not anyone has designed one.
Ownership and denominator. Prospecting is owned by an individual rep and measured per account: which accounts did you work, how well did you research them, did you follow up. Lead generation is owned by the business and measured per stage with a rate: accepted divided by sent, replies divided by messages sent, meetings held divided by meetings booked. If a metric has a person's name attached to it, it is probably a prospecting metric. If it has a denominator attached to it, it is probably a lead generation metric.
Sales prospecting is the deliberate selection of and outreach to named accounts that have not raised a hand. The rep chooses the account, identifies who inside it has the problem, finds a specific reason to reach out now, sends the first touch and keeps following up. The unit of work is one account. That is what separates it from a campaign, where the unit of work is a segment and the copy is written once for everyone in it.
Neither, because they are not sequential stages. Prospecting sits inside lead generation. The system defines the segment and builds the list before any prospecting happens, prospecting picks the accounts and opens the conversations, then the system takes over again at qualification and handover. Teams that treat prospecting as the first step usually skip the segment definition entirely, which is why their reps end up working a list nobody can defend.
Sales owns prospecting because the unit of work is a named account and the judgement involved cannot be centralised. Marketing or growth normally owns the parts of lead generation that scale: the segment definition, the list, the sending infrastructure and the message testing. The failure mode we see most often is the reverse, where marketing hands over a list nobody in sales believes in, and sales is then measured on activity against it. Agree who owns which stage before you agree a target.
Measure prospecting on coverage and follow-through: accounts touched divided by accounts assigned, and accounts reopened divided by accounts parked last quarter. Measure lead generation on stage conversion with explicit denominators: accepted divided by sent, replies divided by messages sent, meetings held divided by meetings booked. Published reference points, reviewed September 2026, include acceptance of 28.5% across 13,218,869 requests (Expandi) and 26% (Belkins), and reply rates of 7.2% across 15.1 million contacts (Belkins) and 10.4% across 6,730,447 messages (Expandi). They come from different datasets with different denominators, so use them to sanity check one stage at a time and never multiply them together.