Shruti Kaushik

LinkedIn26 min read

How to Get Your First 10 B2B Customers: Data and a Method

A 4-ring method for your first 10 B2B customers, built on 2023 to 2026 outbound and founder data, with worked examples, mistakes and a ledger template.

By

Chapters
  1. 01What page one says, and what it leaves out
  2. 02The data on how first customers actually arrive
  3. 03The data on outbound reply rates
  4. 04The data on channels beyond email
  5. 05The data on deliverability and list quality
  6. 06The data on how many conversations you need
  7. 07Compiled dataset: 17 sources
  8. 08What the data disagrees on
  9. 09Shruti's analysis
  10. 10The Ten-Door Method
  11. 11Worked examples (hypothetical)
  12. 12Mistakes that cost founders their first 10
  13. 13A 90-day checklist
  14. 14How Shruti does this for clients
  15. 15Methodology
  16. 16Sources

To get your first 10 B2B customers, work in four rings. Sell to people who already trust you, ask them for introductions, run outbound on a short list of named accounts, then publish proof so buyers find you. Expect cold email to be the weakest ring: median reply rates in large datasets sit below 1%.

Most guides on this topic are lists of tactics from one founder's memory. This one starts from the data: 17 sources, with sample sizes and dates, where they agree, where they conflict, and a ledger you can use to find out which ring works for your product. The method is mine. The numbers are not, and each one links to the page where I read it.

#What page one says, and what it leaves out

Before writing this I searched the target query and read the pages that came up. The results fall into four groups.

  • Founder essays. Lenny Rachitsky's piece built on interviews with 24 contributors, Steli Efti's Close post from 2016, and a Product Hunt story by Chris Bakke. Useful, but each rests on stories, and the stories skew toward companies that went on to succeed.
  • Vendor guides. Prospeo, Boomerang and similar pages give timelines and funnel ratios, with a product to sell at the end. Some of their figures have no source.
  • Community reposts. The SaaSBoomi essay republished on eChai gives sensible principles and no statistics.
  • Thin or off-topic pages. At least one result in the first page was a general file-sync blog that does not answer the question.

What none of them did is put the outbound datasets next to the founder stories and say which to believe. That is what this article does. If you only read one section, read "What the data disagrees on".

#The data on how first customers actually arrive

Founder accounts and vendor models agree that the first customers come mostly from warm sources. Lenny's interviews rank personal network first and press sixth. Boomerang's enterprise model expects 0 to 2 of the first 10 from cold or content-led inbound. Treat both as practitioner evidence, not controlled studies.

Lenny Rachitsky's series is the closest thing the founder community has to a reference. The article does not report a survey sample or the share of founders using each channel, so I use it for named examples only.

  • 24 contributors Thanked in Lenny Rachitsky's series on winning the first 10 B2B customers, with seven channels ranked in recommended order, starting with personal network and ending with "just launch" (Lenny's Newsletter, 5 Sep 2023).
  • 15 to 18 New IT contacts per month was Okta's target for six months, and its founder said they reached 85% or more of that quota (Lenny's Newsletter, 5 Sep 2023).
  • Half Of Census's first 10 customers came through co-founder Sean Lynch's contacts and the other half through Boris Jabes's (Lenny's Newsletter, 5 Sep 2023).
  • 80 employees Was the size of Zip's first customer, Deserve, won through LinkedIn messages that began as requests for advice (Lenny's Newsletter, 5 Sep 2023).
  • 6 companies Shared a single Crunchbase account at Retool, which used it to filter startups by funding and vertical and then emailed CTOs and VPs of Operations (Lenny's Newsletter, 5 Sep 2023).
  • 8,000 People requested a Slack invitation on day one, rising to 15,000 two weeks later, which shows how rare a pull-driven start is (Lenny's Newsletter, 5 Sep 2023).

The vendor model is more specific, and less independent. Boomerang sells warm-introduction software, so read its split with that interest in mind.

  • 2 to 4 Of the first 10 enterprise customers come from the direct founder network, 2 to 4 from investor and advisor introductions, 2 to 4 from early-customer referrals and 0 to 2 from cold or content-led inbound, for contracts of $30k to $200k (Boomerang, undated).
  • 12 to 18 months To reach customer 10 in that model, with 4 to 6 signed by month 6 and 8 to 12 by month 9 for founders with strong networks (Boomerang, undated).

The page is undated, so I cannot tell you how current it is.

One more founder data point matters because it runs the other way. Steli Efti's Close post has not changed its advice since 2016, and the advice is blunt.

  • 10 people Is the minimum to ask for a purchase within 24 hours, regardless of how they respond, and the post tells founders to multiply their first price guess by 3 (Close, 3 Feb 2016, updated 28 May 2026).

#The data on outbound reply rates

Reply rates differ by a factor of ten across datasets, from 0.2% to 8.5%. The gap comes from definitions: replies per email, replies per contact, auto-replies included or not, and who is on the platform. Plan around the low end and treat the high end as a ceiling.

If your first 10 customers include any cold outreach, you need a realistic denominator. These are the largest datasets I could read in full or through a source that states its sample.

Now the sales-side evidence on what changes those rates.

  • 344 Cold emails for the average rep to land one meeting, and top performers book 8.1 times more meetings, in an analysis the article describes as 28 million or more emails (Gong, 24 Jul 2025, updated 27 May 2026).
  • 57% Is the maximum reply-rate loss Gong associates with pitching in the first email (Gong, 24 Jul 2025, updated 27 May 2026).
  • 30.2% Is how much less likely C-level executives are to reply to cold email than non-executives, in an analysis of 1M+ executive sales cycles (Gong, 29 Jan 2026).
  • 100 words Is the point past which reply rates drop sharply, with the best emails between 50 and 100 words (Gong, 29 Jan 2026).
  • 85 million Cold emails analyzed in Gong's guide, which claims the top 10% of senders book 8 times more meetings, with the benchmark tables behind a form (Gong, undated).

#The data on channels beyond email

Phone, LinkedIn and referrals each have measured conversion figures, but sample quality varies. Cognism's 55,701 dials produced a 4.82% dial-to-meeting rate in 2024, and Expandi's 20 million LinkedIn attempts showed 29.61% connection acceptance. Neither is a reply rate, so do not compare them to email directly.

Reply rates are the number everyone quotes, and they are the wrong unit for comparing channels. A connection accepted is not a conversation, and a dial answered is not a meeting. The figures below come from a roundup by LeadHaste, an outbound agency, which reports them with the original publisher named. I could not open the original pages, so treat these as secondary.

Two practical points follow. First, even the cold-calling data, which looks best, is a measure for sellers who dial all day, not for a founder with a few spare hours. Second, the multi-touch figure of 8 against 5 suggests the difference between a good and average seller is not volume but the quality of each touch, which matches Gong's finding that pitching in the first email costs replies.

#The data on deliverability and list quality

Outbound at the first-10 stage fails quietly when the mail never arrives. Platform rules set the ceiling.

I could not open Google's own support page in this run, so I rely on two secondary citations of it. Check the current rules there before you send.

#The data on how many conversations you need

Estimates for the funnel are vendor-built. Prospeo's guide says 30 to 50 qualified conversations produce 10 customers, with about 1 in 5 serious conversations converting early. Chris Bakke reports 2 to 6 qualified meetings in 30 days from a list of 20 named companies. Neither is a benchmark.

  • 30 to 50 Qualified conversations are needed to close 10 customers, with about 1 in 5 serious conversations converting at the early stage (Prospeo, 2026, undated).
  • 50 messages To a personal network yield 10 to 15 replies, 3 to 5 conversations and 1 to 3 customers, in the same guide's funnel math, against 100 cold emails yielding 2 to 14 replies and 1 to 3 converting conversations (Prospeo, 2026, undated).
  • 2 to 6 Qualified meetings within 30 days from a list of 10 to 20 target companies, with one example founder booking 7 meetings in 17 days from 20 companies (Chris Bakke, Product Hunt Stories, 5 Oct 2024).
  • 13.6% Reply rate reached by one founder in the Prospeo guide's example across 400 or more outreaches, up from 1% to 2% (Prospeo, 2026, undated).

Prospeo sells contact data, and its funnel figures carry no stated sample. I use them as a starting range for the ledger below and nothing more.

#Compiled dataset: 17 sources

The table lists every source behind a number in this article, with sample size and date as stated on the page I read. Where a source was read through another publisher, the table says so. Four of the 17 state no sample size at all, and three more state one only through a second publisher.

Compiled dataset: 17 sources
# Source What it measures Sample size Date
1 Lenny's Newsletter Founder-reported first-10 channels 24 contributors thanked, no survey n 5 Sep 2023
2 Close (Steli Efti) Founder advice on first 10 Not stated 3 Feb 2016, updated 28 May 2026
3 Chris Bakke, Product Hunt One cold campaign and named-list method 7,000 emails, one campaign 5 Oct 2024
4 Boomerang Enterprise first-10 source split and timeline Not stated Undated
5 Prospeo Funnel ratios for first 10 Not stated 2026, undated
6 Smartlead Reply rate by percentile 850M+ emails Undated, likely 2026 H1
7 Instantly Platform reply rate and first-email share Billions of interactions Data 1 Jan to 18 Dec 2025, page 16 Sep 2026
8 Busyseed Reply baseline across sending groups 287,790 emails, 45 groups Sent 26 Jan to 14 Aug 2026
9 Sales.co (via Konabayev) Any-reply, positive-reply, reply type 1,369,823 contacts, 161 campaigns Verified 20 Aug 2026
10 Pitchbox and Backlinko (via Instantly) Average reply, follow-up lift 12M emails Undated
11 Saleshandy (via Lemlist) Average reply rate 53M emails Undated
12 Belkins (via Lemlist) Reply rate excluding auto-replies 7.5M emails Undated
13 Gong, cold email study Emails per meeting, pitching penalty 28M+ emails per body text 24 Jul 2025, updated 27 May 2026
14 Gong, executive study Executive reply gap, length effect 1M+ executive sales cycles 29 Jan 2026
15 Cognism (via LeadHaste) Cold call connect and meeting rates 55,701 dials 2024
16 Expandi (via LeadHaste) LinkedIn connection acceptance 20M+ attempts Undated
17 RAIN Group (via LeadHaste) Touches to first meeting 489 sellers Undated

Count the gaps in that table. The sources with the largest samples (Smartlead, Instantly, Saleshandy) tell you least about who was emailing whom. The sources that name the audience (Bakke, Busyseed) have the smallest samples. Nobody has published a clean, audited study of how B2B startups win their first 10 customers, which is why founders pass stories around instead.

#What the data disagrees on

Five conflicts matter. Reply rates range from 0.2% to 8.5%, founders disagree on whether cold outreach works at all, seniority effects point in opposite directions, the share of replies from follow-ups varies from 21% to 42%, and timelines run from 6 to 18 months. Each has a reconcilable cause.

1. Is the reply rate 0.2%, 0.74%, 1.2%, 3.43% or 8.5%?

All of them are real, and none is wrong. The differences come from four places.

Denominator. Smartlead measures replies per contact. Instantly measures replies divided by total emails sent. Sales.co divides replies by contacts in the funnel. Pitchbox and Backlinko's study does not state its audience in the page that cites it, so I cannot say whether it resembles B2B software buyers.

Auto-replies. In the Sales.co set, 45.1% of classified replies were auto-replies and 14.1% were positive. Belkins excluded auto-replies and landed at 0.45%, while the Lemlist roundup does not say whether Saleshandy's 3.7% did. A dataset that counts out-of-office messages will read several times higher than one that does not.

Who is on the platform. Instantly's 3.43% is an average across senders who chose a tool built for cold email at scale, and its top 10% figure is 10.7% or more. Smartlead's median is 0.74% and its top 5% is 3.85%. Same activity, two platforms, and the Smartlead top 5% sits near the Instantly average. That gap is consistent with platform mix and measurement, not with one platform's customers being better writers.

List quality and targeting. Bakke's 0.2% came from 7,000 emails to a broad list. His 2 to 6 meetings in 30 days came from 10 to 20 named companies with several contacts each. The first number divides by volume. The second does not.

What to do with this: plan on a reply rate of about 1% for a cold list you have not tested, since the two largest neutral-looking samples sit at 0.74% and 1.2%. Treat 3% or more as something your own ledger has to earn.

2. Does cold outreach work for the first 10, or not?

Boomerang says "no cold-outbound shortcut exists at this stage" and expects 0 to 2 of the first 10 from cold or content-led inbound. Lenny's interviews list Zip winning its first customer, Deserve, through LinkedIn, Retool reaching DoorDash, Rappi and Brex through emails to CTOs and VPs of Operations, and Ramp getting Truebill from a cold email in its first 50.

Both can be true. Look at who won through cold: Zip's founders wanted a customer who did not know them, because they wanted a clean read on whether the idea was good. Retool picked a vertical, filtered by recent funding and emailed a named role. Those are targeted outbound motions, closer to Bakke's list of 20 than to a 7,000-email blast. Boomerang's claim fits blast-style cold. The Lenny examples fit named-account cold. The method below separates them for that reason.

3. Do executives reply more or less?

Gong found C-level executives 30.2% less likely to reply than non-executives. Sales.co's data, read through Konabayev, found C-level contacts had a positive share of 14.16%, about 3.3 times the manager result, while directors replied at 2.46% against 1.48% for junior contacts.

These are different questions. Gong measures whether anyone replies. Sales.co measures what share of replies are positive. Executives reply less often, and when they do, the reply is more likely to be a yes or a referral. For a founder with limited hours, that favours fewer, better-researched messages to senior buyers over volume to junior ones. Neither dataset isolates causation, and Sales.co's cohort was not shown to be the same as its funnel.

4. How much do follow-ups matter?

Instantly says 58% of replies come from the first email and 42% from follow-ups. Sales.co says initial outreach produced 79.4% of replies and follow-ups 20.6%. Pitchbox and Backlinko report a 65.8% response lift from one follow-up.

The three can sit together. A lift measures the effect of adding a follow-up to a sequence, while a share measures where replies landed, and shares depend on how many follow-ups a sender uses. Sales.co also found follow-ups had a lower reply rate (1.22%) than initial outreach (1.83%), which is the normal pattern: each extra touch reaches a smaller, already-uninterested pool. Send the follow-up, but do not expect it to rescue a weak first message.

5. How long does it take?

Okta targeted 15 to 18 new IT contacts per month for six months. Boomerang's model puts customer 10 at 12 to 18 months for contracts of $30k to $200k. Prospeo's guide describes the full B2B buying process as 3 to 6 months.

Deal size and network explain the spread. Boomerang's range is for enterprise-sized contracts, and its faster path assumes a strong network. Okta sold to IT contacts the founders already knew from a previous employer. If your price is low and your buyers know you, plan on quarters. If your price is high and your network is thin, plan on more than a year, and do not mistake that for failure.

#Shruti's analysis

Across the sources, one pattern holds: the closer the first contact is to existing trust, the less the channel matters. Warm networks beat cold lists, named lists beat large ones, and public proof lowers the cost of every other ring. The first 10 reward focus on who you ask, not on message tricks.

Reading all of this together, three conclusions hold up against the conflicts above.

First, the disagreement between "cold works" and "cold does not work" is mostly a disagreement about what counts as cold. Blast outreach to a purchased list and targeted outreach to a researched name are different activities with different results, and most of the published reply-rate data is about the first. A founder who reads the low median and concludes outbound is dead has measured the wrong thing. A founder who reads the top-tier figure and plans around it has measured the wrong thing in the other direction.

Second, the unit of progress at this stage is a conversation with a person who has the problem, not a sent email. Every dataset above counts messages, and none counts what the founder learned. That is why the ledger in the method below records the trigger, the pain and the objection for each conversation. The named examples in Lenny's interviews, from Census to Gusto, mostly start with a pre-existing relationship, which the datasets cannot show.

Third, public proof is the ring that compounds. A named customer, a published result or a visible founder makes the next introduction easier and the next cold message less cold. It also decides whether a buyer or an AI answer engine names your company when someone asks who solves the problem. That is a separate surface from email and LinkedIn, and it is the one most first-time founders ignore until after the first 10.

I hold the least confidence in the timelines. They come from vendor models and founder recall, and they vary with price and network more than any other number here.

#The Ten-Door Method

The Ten-Door Method is a four-ring plan for your first 10 B2B customers. You list named accounts (doors), work them ring by ring from warm to cold, and record every conversation in a ledger. The counts and thresholds are working rules from my practice, not industry benchmarks.

A door is one named account where you can say who the buyer is, what trigger makes now the right time, and how you would reach them. The method is built from the data above: warm first, named lists over blasts, one ledger to replace borrowed benchmarks.

The Ten-Door Method
Ring Doors How you reach them What the data supports
1. Trust People who know you personally Direct message or call, with a price Lenny's examples (Census, Gong, Gusto, Notion); Boomerang's 2 to 4 of 10 from network
2. Introduction People your Ring 1 contacts and investors know Ask for one named introduction each Boomerang's 2 to 4 of 10 from investors and advisors
3. Named list Up to 20 researched accounts Multi-channel, several contacts per account Bakke's list method; Retool and Zip examples
4. Proof Buyers who find you Publish customer results and founder writing Boomerang's 0 to 2 from inbound; the ring that compounds

The working rules

  1. Cap the named list at 20 doors per ring pass. This follows Bakke's 10 to 20. Past that, you cannot research each door properly.
  2. Ask for money in the first conversation. Follow Efti: ask at least 10 people within 24 hours, and price at roughly three times your first instinct. A free pilot is a different product.
  3. Log every conversation. One row: door, ring, trigger, pain in the buyer's words, objection, outcome, and the date.
  4. Review after 20 logged conversations. Replace Prospeo's estimate of 30 to 50 conversations per 10 customers with your own conversion number.
  5. Do not open Ring 4 as a substitute for Rings 1 to 3. Content does not close a first customer by itself. It makes the closing conversation shorter.
  6. Stop and rewrite the pitch if five conversations in a row end in the same objection.

The ledger template

TEN-DOOR LEDGER

Door (company, role):
Ring: 1 Trust / 2 Intro / 3 Named list / 4 Proof
Source of the door: network / intro from ___ / list / inbound
Trigger (why now):
Pain in buyer's words:
Who else is involved (champion, signer, blocker):
Price quoted:
Objection:
Outcome: no reply / conversation / paid pilot / customer / lost
Next step and date:

WEEKLY REVIEW
Doors opened this week:
Conversations held:
Paid yes:
Doors I should have skipped, and why:
Pattern forming (same buyer, trigger, pain)?

#Worked examples (hypothetical)

These three examples are illustrative composites. They are not real clients, and the numbers are arithmetic on the sourced rates above, not results anyone achieved.

Example 1: Ring 1 and 2 for a $20k product (hypothetical). A founder selling workflow software to logistics operations lists 30 people she has worked with. She messages all 30 with a price and gets conversations with 6. She asks each of the 6 for one introduction and gets 4 more doors. Using Prospeo's 1 in 5 rate as a rough guide, 10 conversations might give 2 customers. She does not stop there. She logs all 10 and finds that 7 of the 10 buyers were operations directors who started looking after a carrier failure. That trigger becomes the first line of her Ring 3 outreach.

Example 2: Ring 3 arithmetic for the same founder (hypothetical). She builds a named list of 20 logistics companies that had a public carrier disruption in the last year, with three contacts each, so 60 people. At Smartlead's median of 0.74% per contact, 60 contacts would give under one reply by email alone. That is the point of the arithmetic: a cold email to 60 people is too small a sample to learn anything. So she adds LinkedIn messages and two phone calls per account and sends a short note referencing the disruption. If the targeted version performs near Instantly's 3.43% average, 60 contacts give about two replies. Her ledger, not the benchmark, tells her which number she is closer to after the first 20 conversations.

Example 3: Why blast fails at the first-10 stage (hypothetical). A second founder sends 7,000 emails because a tool makes it cheap. At Bakke's 0.2%, that is 14 responses. At Smartlead's 0.74%, it is about 52. Even at the better figure, he now has dozens of replies to read, no research behind any of them, and a sender reputation to protect against the 0.3% complaint line. A founder with 20 researched doors and a calendar slot per door learns more from the first 5 conversations than he does from all 52 replies.

#Mistakes that cost founders their first 10

The common mistakes are skipping Ring 1 because it feels like favours, blasting a purchased list, running free pilots, leaving the ledger empty, pitching in the first email and going quiet publicly once outreach starts. Each has support in the data or the founder accounts above.

  1. Skipping the network because it feels like favours. Gong's founders, Census's founders and Notion's founders all started with people they knew. A paid yes from a trusted contact is a legitimate sale and the fastest way to learn what the product is worth.
  2. Blasting a purchased list. The data above says the typical result is a reply rate near or below 1% and a deliverability risk. The method caps lists at 20 doors for a reason.
  3. Running free pilots with no price. The reply tells you nothing about willingness to pay. Efti's rule of asking within 24 hours exists to avoid that trap.
  4. Pitching in the first email. Gong ties pitching to a reply-rate loss of up to 57%. Lead with the buyer's trigger and a question.
  5. Writing long emails. Gong's 100-word line and Instantly's elite campaigns under 80 words point the same way.
  6. Ignoring auto-replies in your own numbers. If 45.1% of replies in a large dataset are out-of-office messages, your reply rate probably is not what the dashboard says. Log positive and negative replies separately.
  7. Treating the first 10 as one pattern. Ten customers from five unrelated deal types is a consulting business. The ledger's trigger and pain fields show you whether the 10 rhyme. See the founder-led sales strategy guide for what to do once they do.
  8. Doing outreach in silence. A message from a founder whose profile has no recent writing, no customer proof and no presence in AI answers is colder than it needs to be. This is my judgment from practice, not a measured effect.
  9. Hiring before the pattern exists. Boomerang's model suggests no SDR or AE before customer 8. Hiring earlier asks someone else to learn the pattern you have not finished learning.

#A 90-day checklist

FIRST 10 B2B CUSTOMERS: 90-DAY CHECKLIST

WEEK 1 (setup)
[ ] Write one sentence: who has the problem, what triggers it, what it costs
[ ] List 30 people who know you and could buy or introduce
[ ] Set up the Ten-Door Ledger (sheet or CRM)
[ ] Decide a price, then check it against "multiply my first instinct by 3"

WEEKS 1 TO 3 (Ring 1)
[ ] Message every person on the list with a price, not a pilot
[ ] Ask at least 10 people to buy within 24 hours of deciding
[ ] Log every reply, including no's and the reason

WEEKS 3 TO 6 (Ring 2)
[ ] Ask each Ring 1 contact for one named introduction
[ ] Ask investors and advisors for 3 named introductions each
[ ] Review the ledger: same buyer, trigger, pain showing up?

WEEKS 5 TO 10 (Ring 3)
[ ] Build a list of up to 20 named accounts that match the pattern
[ ] Find 3 contacts per account, at least one senior
[ ] Write each opening line from a public trigger (funding, hire, incident, post)
[ ] Reach each account on two channels
[ ] Keep each email short and free of pitch in the first message
[ ] Check domain health and complaint rates before sending at all

WEEKS 1 TO 12 (Ring 4, in parallel, 2 hours a week)
[ ] Publish one customer result or lesson a week
[ ] Make sure the company name appears where buyers look
[ ] Check how AI answer engines describe you

WEEK 12 REVIEW
[ ] Count paid customers
[ ] Count conversations per customer from the ledger
[ ] Replace borrowed benchmarks with your own numbers
[ ] Decide: stay in Ring 1 to 3 or start writing the sales motion down

For the AI answer engine check, use the free AI visibility check, read how AI assistants decide which companies to recommend, and see the guide on founder AI visibility. If a buyer is comparing you against incumbents, your page for alternatives is part of Ring 4.

#How Shruti does this for clients

I work as a founder-led go-to-market operator. I build inbound systems, and the AI behind them, for B2B companies. That is the demand side of this article: Ring 4 is where most of my work sits, and it is the ring that makes Rings 1 to 3 easier.

My background is in running businesses as well as advising them. I founded Wise Mango in 2015 and Klype AI in 2025. I have written 3,600+ LinkedIn posts for SaaS and AI founders, and my clients include Sprinklr, Sony, Zee Entertainment, Array Networks and SugarBox. I was named in Forbes Game-Changing Leaders 2025 and I am the author of The LinkedIn Blackbook.

What I do with a founder at the first-10 stage is narrower than a campaign. I look at the ledger, find the buyer, trigger and pain that repeat, and build the public proof and inbound path around that pattern, so a message to a named account lands on a profile and a site that back it up. If you want the full playbook I use with founders, join my private letters.

#Methodology

I searched the target query "how to get your first 10 B2B customers" on 5 Nov 2026 and read the pages that ranked: Lenny Rachitsky's essay, Close, Chris Bakke's Product Hunt story, Prospeo, Boomerang, the SaaSBoomi repost on eChai and others. I then looked for primary data in four groups: founder accounts, outbound reply-rate datasets, sales-engagement research and deliverability guidance.

Source selection rules: prefer a page with a stated sample size and date; prefer the original publisher; mark anything read through another publisher; drop any number I could not see on a page in this run. Dated sources span 2016 to 2026. Only the Close post predates 2023, and it was updated in May 2026.

Limits you should know about. Several primary sources I wanted were not available to me in this run, including 6sense, Gartner, the Bridge Group, Carta, Y Combinator, First Round, CB Insights, Belkins and Google's own support page, because the page requests were not approved. I did not substitute figures from memory. That is why Saleshandy, Belkins, Pitchbox and Backlinko, Cognism, Expandi, RAIN Group and Sales.co appear as "via" entries, read through Lemlist, Instantly, LeadHaste or Konabayev, and why the dataset has gaps in buyer research and venture data. Those secondary figures are as reported and may differ from the original. The Smartlead page, the Boomerang page, the Prospeo page and Gong's 85 million email guide carry no explicit date. Gong's article gives 28M+ emails in the body and 25 million in its description. Instantly's page gives two slightly different date ranges for its data. Reply rates across vendors use different formulas and should not be averaged.

The Ten-Door Method, the ring structure, the working rules, the ledger, the checklist and the worked examples are my own. The worked examples are hypothetical, and their arithmetic multiplies sourced rates by round contact counts. No client results appear in this article, and no pricing for my own services.

#Sources

Quick questions

How do you get your first 10 B2B customers?

Start with people who already trust you, then move to a short named list of target accounts, then publish proof. Lenny Rachitsky's founder interviews rank personal network first and cold outbound second. Boomerang's enterprise model expects 2 to 4 of the first 10 from the founder's network.

How long does it take to get the first 10 B2B customers?

It depends on deal size. Okta's founders targeted 15 to 18 new IT contacts a month for six months. Boomerang's model for $30k to $200k contracts puts customer 10 at 12 to 18 months, with 8 to 12 signed by month 9 at the faster end.

Does cold email work for the first 10 customers?

It works, but weakly. Smartlead's data on 850 million emails shows a median reply rate of 0.74%, while Instantly's platform average is 3.43%. Chris Bakke reported 0.2% from 7,000 generic emails, then 2 to 6 meetings from a list of 20 named companies.

How many conversations do I need to close 10 customers?

Prospeo's 2026 guide estimates 30 to 50 qualified conversations, with about 1 in 5 serious conversations converting early on. It is a vendor estimate, not a benchmark, so treat it as a planning range and replace it with your own ledger after 20 conversations.

Should I charge my first B2B customers?

Yes. A free pilot only proves the buyer will accept free software. Steli Efti of Close advises asking at least 10 people to buy within 24 hours and multiplying your first price instinct by 3, so the price is tested before the product is built out.

When should I hire a salesperson?

After the founder has closed a pattern, not a handful of favours. Boomerang's model suggests no SDR or AE before customer 8 and a first AE at months 10 to 12. Read the founder-led sales strategy guide on this site for the handoff signals.

Is LinkedIn or email better for B2B outreach?

The datasets measure different things, so neither wins outright. LeadHaste's roundup cites a 29.61% LinkedIn connection acceptance rate from Expandi and a 3.43% email reply rate from Instantly. Acceptance is not a reply, so use both for a named list and compare in your own ledger.

About the author

runs Wise Mango, a LinkedIn ghostwriting studio for SaaS and AI founders, and Klype (klype.ai), an AI platform that writes in a founder’s voice. She wrote The LinkedIn Blackbook and was named one of Forbes Game-Changing Leaders in 2025. Every figure on this page links to its source.