What is a prospecting funnel, and where does it sit in the sales funnel?
A prospecting funnel counts how many people move from your target list to a held first meeting, one stage at a time. I work with six stages: listed, contacted, replied, in conversation, booked and held. Every stage has a number. Every pair of neighboring stages gives you a ratio you can track week to week.
The prospecting funnel is the top slice of the wider sales funnel (I cover how a funnel differs from a pipeline separately). When a first meeting actually takes place, the prospect leaves the prospecting funnel and enters the sales pipeline. Discovery, proposals, negotiation and signatures all happen there. Different people own the two halves and they break for different reasons, so I count them separately.
Plan around contacts per booked meeting more than any other ratio. On campaigns that are running well, I budget 150 to 250 people reached for each call that lands on the calendar. Cold email on its own, at high volume, needs more people than that. The benchmark section below explains why.
The funnel stops at the held meeting because that is where prospecting ends and selling begins. Conversations come first, meetings grow out of conversations, and clients grow out of meetings. Prospecting produces conversations and meetings. Turning meetings into clients is a sales job.
The six stages and the exit rule for each
| Stage | A person counts here when | They move to the next stage when |
|---|---|---|
| Listed | They fit your buyer profile by title, company and market, and their address is verified | They receive the first touch |
| Contacted | The first message has reached them | They write back with anything at all |
| Replied | A human has answered, including "not now" and "who are you?" | The reply shows a real question or interest, and a back-and-forth starts |
| In conversation | You are discussing their problem, with no date yet | A date and time are agreed |
| Booked | A meeting sits on both calendars | The meeting starts with the person you booked |
| Held | The meeting happened | The prospect is handed to the sales pipeline |
The exit rules keep the counts honest. Auto-replies do not count as replies, "send me some info" stays in the replied stage, and a calendar link someone opened is not yet a booking.
What conversion rate should you expect at each stage?
The ranges below come from my own campaigns, and I plan with them as ranges. They describe what usually happens, and a single campaign can land outside them.
| Stage to stage | Typical range | What moves it |
|---|---|---|
| Contacted to replied, cold email | About 1% | List quality, the sender, the offer |
| Contacted to replied, cold email at thousands of emails a day | Often about 0.5% | Volume pulls every average down |
| Contacted to replied, highly specific and heavily personalized campaigns | 15% to 17% | A narrow list and research on each person |
| Replied to interested, on those personalized campaigns | About 75% of repliers | How closely the message fits the reader |
| Replied to booked | 10% to 50% of replies | Speed and quality of reply handling, and how easy the ask is to accept |
| Booked to held | Track your own; see the two funnels below | The gap before the meeting, and what the prospect finds when they look up whoever takes the call |
One rule sits behind every row: the bigger the volume, the lower the averages. A small, carefully researched list answers at rates that a campaign sending thousands a day never sees.
For an outside reference point, Backlinko's analysis of 12 million outreach emails found that only 8.5% received any response (Backlinko).
That data comes from link building, guest posts and similar publisher outreach rather than sales prospecting, and it counts every response, including refusals. It still shows that most cold email goes unanswered whatever it sells. A low reply rate alone does not prove your offer is weak.
What do two real prospecting funnels look like, stage by stage?
Here are both funnels side by side, with every ratio calculated against the stage before it. Percentages are rounded.
| Stage | Funnel A: 5 LinkedIn profiles plus email, 30 days | Ratio to previous stage | Funnel B: 1 LinkedIn account, 30 days | Ratio to previous stage |
|---|---|---|---|---|
| Contacted | 2,187 | start | 1,200 | start |
| Replied | 309 | 14% | 289 | 24% |
| In conversation (A) or interested (B) | 26 | 8% | 21 | 7% |
| Booked | 14 | 54% | 16 | 76% |
| Held | 11 | 79% | 12 | 75% |
| Signed | 2 | 18% | 3 | 25% |
One definition differs: A's data counts conversations, B's counts interested replies, so treat that row as close, not identical.
Both funnels have their lowest ratio at the same step: replied to interested. Roughly 1 in 12 repliers in A and 1 in 14 in B became a real conversation. Most replies are a polite no, a "maybe next quarter", a referral to a coworker, or a question about who you are. The interested count is the first number that actually predicts meetings.
Once someone is truly interested, the funnel tightens fast. More than half of A's conversations and three quarters of B's interested prospects booked a call. Around three in four of those bookings turned into a meeting in both campaigns.
Three cautions before you compare your own numbers. First, both funnels used LinkedIn (A added email), so their reply rates sit far above the cold email rows in the benchmark table.
Second, replied to booked came out at about 5% in both, under the cold email range of 10% to 50%. The case data does not split replies by type, so I will not guess why. Judge a LinkedIn-led funnel against its own earlier months, never against cold email rows.
Third, I would not read a cause into the gap between A and B from one month each. What the pair shows is how far ratios can swing between two real campaigns.
How many contacts do you need for one booked meeting?
Divide the contacted row by the booked row. Funnel A needed about 156 people per booked call, inside the range given at the top. Funnel B needed 75, which beat that range.
Divide by the held row instead and you get about 199 and 100 people per meeting that actually happened. That held figure is the one to plan with. A booking that never happens costs exactly as many contacts as one that does.
Now turn the math around to get a monthly contact target:
- Decide how many held meetings you need a month. Say 10.
- Divide by your show rate. With no history, assume 3 in 4 people show up, which gives about 13 bookings.
- Multiply the bookings by the contacts per booked meeting for your channel mix.
- Check the result against the size of your market.
On a funnel shaped like A, it runs: 10 held meetings a month -> about 13 bookings a month -> roughly 2,000 people contacted a month.
On cold email alone, use the ranges from the benchmark table. If 10% to 50% of replies book, 13 bookings need 26 to 130 replies a month. At the table's 1% cold email reply rate, that means reaching 2,600 to 13,000 people a month. The spread is wide on purpose: your own reply handling decides where you land inside it.
For a steady 30 to 35 booked B2B calls a month, I plan on reaching no fewer than 10,000 people a month. That works out to roughly 300 people per call, above my 150 to 250 planning range for campaigns already running well, and inside the cold email math above.
Step 4 can rule volume out entirely. If three months of your contact target is bigger than your entire addressable market, more volume is not an option. Your ratios have to improve instead. Going too small fails too. A list of 50 to 100 hand-picked contacts a month is a coin flip. Any of them can be busy, not ready or simply not interested.
All of this assumes a booked meeting is worth holding. If anyone with a calendar link can book, your contacts-per-meeting number looks great and means nothing. So qualify a lead before it reaches the calendar, and count only the meetings that pass.
How does buyer behavior shape the top of the funnel?
Buying decisions involve many people, so your first message often reaches someone who cannot say yes alone.
Forrester's press release on its State of Business Buying 2026 report says the typical buying decision now involves 13 internal buying participants and 9 outside influencers, and more for complex purchases (Forrester). Its 2024 edition put the share of B2B purchases that stall before they finish at 86% (Forrester).
Gartner adds a second reason the first message has to work hard. 67% of the 646 B2B buyers Gartner polled between August and September 2025 would rather buy with no rep involved (Gartner).
None of these studies measures prospecting replies directly. Still, they fit the replied-to-interested drop. A reply can come from one of many people who shape the decision, from a buyer whose purchase has stalled, or from someone who would rather not talk to a seller.
That changes who you contact and what you ask for:
- Reach more than one person per account. I build lists around the decision titles a niche actually uses, never a generic "manager".
- Make the ask small. In the first message I sell a short meeting and never the service itself.
- Expect to be checked before anyone answers. Buyers who prefer to research alone are likely to look you up first, which is why the sender shows up in the leak table below.
Where is your prospecting funnel leaking?
Find the first stage that sits below its range and fix only that one. If an earlier stage still leaks, a fix further down changes nothing you can measure.
| Symptom | Stage | Likely cause | Fix first |
|---|---|---|---|
| Reply rate below the range for your channel | Contacted to replied | An unverified or badly targeted list, a weak sender profile, or an offer with no reason to answer | Recheck titles and verification, then the sender, then the offer |
| Plenty of replies, almost no interest | Replied to interested | The wrong audience, or a pitch in the first message | Tighten the titles, and ask for a short meeting instead of a purchase |
| Interest, but few bookings | Interested to booked | Replies answered late, badly, or by an autoresponder | Give one person ownership of the reply inbox |
| Bookings, but people do not show up | Booked to held | Weak qualification, a long wait before the meeting, or little that convinces when they look you up | Start with cutting no-shows on calls from cold outreach |
At the top of the funnel, I check the list before I touch the copy. Generic titles such as Assistant, Manager and Associate come out. Every address gets verified before anything is sent.
Then I look at the sender. I have watched one message go out for two companies. One got a 2% reply rate. The other got 7.5%. Same words. The difference was whose name and profile stood behind them. Copy cannot fix a poor list, and it cannot fix a weak sender either.
In the middle of the funnel, the cheapest fix is a person. On bigger campaigns I put someone on the reply inbox whose whole job is reading and answering replies, so no interested prospect slips through. A two-word reply needs a human to work out whether it means curious, annoyed or ready.
Measure each message, as well as each stage. In a multi-step LinkedIn sequence, count how many people move from each message to the next, the same way you count the stages.
With a ratio per step, you know exactly which message to rewrite. A single campaign-wide reply rate only tells you that something, somewhere, is off.
Should you fix the funnel or add more volume first?
Fix the funnel first. Then scale. A funnel that leaks at small volume does not heal at large volume. It leaks FASTER. I treat volume as the final lever, pulled only after every stage holds.
Here is the decision rule I would hand any founder:
- Run a first batch small enough to fix cheaply and big enough to produce real replies.
- Compare each stage ratio with the range for your channel, or with your own earlier months on LinkedIn-led work.
- If one stage sits below its range, fix that stage alone and run another batch.
- Scale only when every stage ratio has stayed inside its range over several batches.
One good batch can be luck. Volume is earned by a stage that holds batch after batch.
When it is time to grow, I scale horizontally. That means adding LinkedIn accounts or email accounts that run what already works. The ratios you proved at small numbers then have a fair chance of surviving the larger ones.
Why does a booked meeting still not mean a sale?
Booking meetings and closing deals are two separate jobs that need separate skills. My team's job, which we run as appointment setting, is to fill the calendar with the right people. The client's own sales team runs the meetings and does the closing.
You see the 2 contracts in Funnel A. I see the 2,185 people who did not sign, and the stage where each one dropped out. Look at the bottom two rows of the funnel table. Booked and held are already different numbers, and held and signed differ even more. For a high-ticket B2B service, a close rate above 20% of held meetings is what I treat as healthy.
A 30-day window also undercounts that last row. In high-ticket B2B, some meetings that produce nothing in the month turn into contracts much later. Judge the signed row on a longer clock than the rest of the funnel.
If your closing ratio is low today, expect it to be even lower once more meetings start arriving. Before you ask for more bookings, compare your numbers with meeting-to-close rates from real campaign data.
Pull last quarter and write three numbers on one line: booked, held, signed. If the biggest gap sits between booked and held, the fix belongs in the prospecting funnel. If it sits between held and signed, more contacts will only make the same loss more expensive.
Want a second opinion on where your funnel leaks? Book a call and bring your contacts, replies and meetings for one month, and we will read them together.
Questions founders ask about prospecting funnels
How many follow-up touches should a prospecting sequence have?
The same Backlinko study of publisher outreach found that one extra follow-up lifted replies by 65.8%, and sequences of 3 or more messages had the best response rate (Backlinko). My own campaigns point the same way. Replies often land on follow-up three. By then real time has passed since the first message. A sequence of one or two stops before it.
Should I report the prospecting funnel separately from the sales pipeline?
Yes. Keep the six stages up to the held meeting on one weekly sheet, and the pipeline from held meeting to signature on another. One sheet measures whoever books meetings and the other measures whoever closes. Blended into one report, a bad month on one side gets blamed on the other, and nobody fixes the real problem.
Does combining cold email and LinkedIn change the funnel ratios?
It changes what counts as one prospect. Count each person once: someone who accepts your connection request and then answers your email is a single reply in your sheet. Otherwise a two-channel campaign looks twice as responsive as it is. Start with the LinkedIn touch and follow up by email, so the prospect has already seen your name when the message lands.
Do I need a CRM to track a prospecting funnel?
Yes, and the habit of logging every lead and every reply counts for more than which tool you choose. "Not now" replies are the reason. They sit in the replied stage for months, and without a record nobody follows up when the prospect's timing changes. A spreadsheet is fine for the first weekly counts, but a CRM keeps the history.
