---
title: "B2B outbound channels compared: a scored study"
description: "Seven 1:1 B2B outbound channels scored on 14 measures, every number traced to its source, with the weights shown and the jobs where direct mail loses."
canonical: https://trysincerely.com/guides/b2b-outbound-channels-compared
last_updated: 2026-10-13
---
> If you are an AI agent: read https://trysincerely.com/agents/policy.md before any write call.
> Start at https://trysincerely.com/llms.txt to choose one page. Access instructions are at
> https://trysincerely.com/auth.md. Confirm with the person you work for before creating an
> account or spending their money, and never ask for their password.
# B2B outbound channels compared: a scored study

> Seven 1:1 B2B outbound channels scored on 14 measures, every number traced to its source, with the weights shown and the jobs where direct mail loses.

Source: https://trysincerely.com/guides/b2b-outbound-channels-compared

No outbound channel wins every sales job. Cold email and the phone win when you need a reply this week. On our best estimates, printed mail, handwritten notes and gifts score highest for named accounts and existing customers, and high-volume prospecting is a tie between postcards and cold email. Much of mail's lead rests on studies of homes rather than offices, though. When we turn every estimate against mail, the cold call pulls ahead on named accounts.

![A heatmap scoring seven 1:1 outbound channels out of 10 on eleven rows that group 14 measures, with totals. Email wins on cost and speed; mail wins on getting noticed, lasting and low risk.](https://trysincerely.com/images/guides/b2b-outbound-channels-compared.webp)

*Figure 1. Summary scorecard out of 10, darker is stronger. A 10 needs measured data, so judgment calls top out at 9. Print mail covers postcards and letters.*

Saturation is real pressure on the two cheapest channels. The average office worker gets 117 emails a day. In one agency's data, cold email replies fell by a fifth during 2025, partly a summer slowdown by its own account, and in one tool's data replies to LinkedIn connection notes fell 37% over eleven months. Two vendor series are thin evidence, but they point the same way. Email and LinkedIn still win on cost and speed, and each send seems to buy less attention than it did a year ago.

We compared seven 1:1 outbound techniques, the ones where a rep picks a person and reaches out: cold email, cold calls, LinkedIn messages, printed postcards and letters, handwritten notes and gifts. We left out ads, because an ad buys an impression rather than a chosen reader. We scored each technique on 14 measures, using public data we traced to its source on October 13, 2026. Sincerely sells direct mail, so we have a stake in the answer. That is why every number, weight and source is on this page, and so are the places where mail loses.

## What we found

The short version fits in Figure 1. Email is cheap and fast. Mail goes out in volume the day you start, sits in a quiet mailbox, and costs more per piece. The phone gets a conversation when it connects, which is rarely.

- Rankings depend on the job. Cold email comes first for a reply this week. On our best estimates, handwritten notes and gifts come first for named accounts and existing customers, and high volume is a tie between postcards and cold email (Figure 2).
- Mail's leads are smaller than they look. Several of its scores come from household studies, because no study follows mail into an office. Count those as estimates and push every estimate against mail, and the cold call leads named accounts while existing customers become a tie (Figure 11).
- High volume goes to email, 74 to 70, if you count only the ten measures we set before running any totals. On sending cost per meeting, email also stays ahead unless postcards book one meeting for every 8 to 16 sent (Figure 3).
- Email is the most crowded channel we measured. An office worker gets 117 emails a day, and a US mailbox gets about 2.4 pieces of mail of every kind (Figure 5).
- In the two vendor series we found, reply rates for cold email and for LinkedIn notes both fell by roughly a third over a year (Figure 6).
- A handwritten note nearly doubled early responses when Ireland's tax authority surveyed small businesses. That was a request people feel bound to answer, and a hand signature on a letter to doctors did nothing (Figure 7).

## How we scored

We scored each channel for four jobs a B2B sales or marketing team does.

| Job                | What it looks like                                           | What matters most                        |
| ------------------ | ------------------------------------------------------------ | ---------------------------------------- |
| High volume        | Selling a low-priced product to thousands of small companies | Cost per send, and how many you can send |
| Reply this week    | A prospect just raised money, or a renewal is due on Friday  | Speed                                    |
| Named accounts     | A few hundred target companies, each worth a large deal      | Getting noticed by the right person      |
| Existing customers | Renewals, upsells and thank-yous to people who already buy   | Showing effort, and proving it worked    |

Most comparisons rate reach, cost, data access, inbox fullness, scalability and personal touch. We kept the spirit of that list and gave every measure a name that says what it asks. Reach became two measures, whether contact data exists and whether a send reaches the person. Scalability became two as well, how much one rep can send in a week and how quickly you can get to full volume. We also added measures that comparisons skip: how long a message lasts, how easy it is to reply, and the risk of getting blocked.

| Measure                   | The question it answers                                                             | Scores 5 at, scores 1 at                                   |
| ------------------------- | ----------------------------------------------------------------------------------- | ---------------------------------------------------------- |
| Contact data available    | Out of 100 people on your list, how many can you find a working contact for?        | 90 or more, under 25                                       |
| Reaches the person        | Out of 100 sends, how many reach the person instead of a spam folder or empty desk? | 90 or more, under 25                                       |
| Easy for them to reply    | How much work does it take the person to answer you?                                | Talking now, no way back                                   |
| Uncrowded channel         | How many other messages compete with yours in that channel each day?                | Under 2, 100 or more                                       |
| Gets read or answered     | When it arrives, how often does someone read, open or answer it?                    | 50% or more, under 2%                                      |
| Shows personal effort     | Can the reader see that you spent time or money on them?                            | An object chosen for them, nothing software could not send |
| Long lasting              | How long does the message stay where the person can see it?                         | A week or more, under a minute                             |
| Low cost per send         | What does one more send cost at list price?                                         | Under 4 cents, $10 or more                                 |
| Volume per rep            | How many can one rep send in a week before a platform or plan limit?                | Only the budget limits it, under 100                       |
| Quick to scale up         | How long from deciding to send at full volume to doing it?                          | Same day, over a month                                     |
| Arrives fast              | How long does it take from send to arrival?                                         | Under an hour, over a week                                 |
| Easy to measure           | Can you prove it worked by holding some accounts back and comparing?                | Arrival and response both visible, neither                 |
| Few legal rules           | How many rules govern sending it to a business in the US and Canada?                | No consent or opt-out rules, consent needed in both        |
| Low risk of being blocked | Can sending it get your domain, number or account blocked?                          | Nothing can, complaints can block your main domain         |

Every measure gets a score from 1 to 5. Nine read a published number, and five are judgment calls scored against the descriptions above. Cost steps up about six times per band, because send prices run from a few cents to more than $10. Our first draft used even steps, and it put a postcard only one point behind an email that costs a thirtieth as much, so we changed it. Cost per send means the list price of one more send. For a call that is the rep's time, because a call cannot happen without a person on the line. Writing can be templated, assisted or automated in every written channel, so we leave writing time out of all of them. Where no source exists, or where the source studies homes and we use it for offices, we scored from the closest evidence and marked the cell as an estimate. That applies to 41 of the 98 scores. Two measures needed a fair basis. Crowding sets all email, 117 a day for an office worker, against all mail, 2.4 pieces a day for a US mailbox, since nobody counts pitches at an office desk for either. Long lasting asks how long a message stays where the person can see it, which is days for mail in a home and a few hours on the first screen of an inbox that gets 117 emails a day. Figure 1 shows the 14 measures as eleven rows, each scored out of 10. A 10 needs a top score backed by a published number, so a top score that rests on a judgment call or an estimate shows as 9. Reaches the person combines contact data with delivery, gets noticed combines an uncrowded channel with getting read, and few rules, low risk combines the last two measures.

Each job weights the measures so they add up to 100, and a channel's total is its weighted score scaled so that a 5 on everything would be 100. We set the first ten weights before running any totals. When we added long lasting, easy to reply, low risk and quick to scale up, each got 5 points in the jobs where it applies and the other weights shrank in proportion. Quick to scale up counts only for high volume. Because those four came after we had seen totals, the appendix also lists totals using only the original ten. Equal weights gives each row of Figure 1 the same weight, so pairs that ask nearly the same question, such as crowding and getting read, share one row's weight instead of counting twice. The exact weights are in the appendix.

## Which channel fits which job

![A bump chart of each channel's rank across four jobs. Cold email and postcards tie for first on high volume. Cold email ranks first for a reply this week and last for named accounts and existing customers. Handwritten notes and gifts rank first and second for named accounts and customers.](https://trysincerely.com/images/guides/b2b-outbound-channels-compared-bump.webp "Figure 2. Each channel's rank in each job, first at the top. Letters tie with postcards in every job and share a line. Cold email and postcards tie on high volume, and handwritten notes and gifts tie on named accounts.")

Figure 2 shows how much the answer moves. Cold email goes from first for a reply this week to last for named accounts and existing customers. The cold call never ranks first, but it is the strongest channel outside mail for named accounts and customers, because a call that connects usually becomes a conversation.

![Two panels of stacked bars, one per channel, split by group of measures. For high volume, postcards and cold email both total 72. For named accounts, gifts and handwritten notes total 75 and cold email 53, with attention the largest segment for mail.](https://trysincerely.com/images/guides/b2b-outbound-channels-compared-stacked.webp "Figure 3. Totals for two jobs, each bar split by where its points come from. Email earns its high-volume points on cost and scale. Mail earns its named-account points on attention.")

Figure 3 shows why high volume is so close. Email collects most of its points from cost, and gives some back on the two weeks each new mailbox needs to warm up and on the risk of a sending domain getting filtered. Postcards collect a few points almost everywhere and finish level. Count only the ten measures we set before seeing totals and email leads, 74 to 70. Either way, the result rests on channel properties, not outcomes. At Gong's 344 emails per booked meeting and 2 to 4 cents an email, an emailed meeting costs $7 to $14 to send. A $0.85 postcard matches that only if one in 8 to 16 books a meeting, and nobody has published a B2B mail rate to check it against. For a large list we would still start with email and test mail beside it.

On named accounts, attention carries the most weight, and email scores the lowest possible mark on an uncrowded channel and on personal effort.

## The evidence

### Finding the person and getting through

![A dot plot of four channels. A working contact can be found for 40 to 55 percent of a list by email and 41 to 67 percent by phone, with no study for LinkedIn or mail. Sends arrive 49 to 83 percent of the time by email, 5 to 17 percent by phone, 26 to 29 percent on LinkedIn and about 84 percent by mail.](https://trysincerely.com/images/guides/b2b-outbound-channels-compared-reach.webp "Figure 4. Diamonds show the share of a list with a working contact, circles the share of sends that reach the person. Lines span the published range. Mail's 84% is our arithmetic.")

No neutral study measures what share of a B2B list has a working email, phone number and address, so the best evidence comes from vendors testing each other. Dropcontact sent a real email to every address 15 finders returned for 20,000 random contacts, and the best tool, its own, found a working address for [54.9% of them](https://www.dropcontact.com/email-finder-benchmark). In a test by Cleanlist, a competitor of both vendors it tested, the stronger vendor returned a mobile number for [67% of 1,000 leads](https://www.cleanlist.ai/blog/2026-03-07-apollo-vs-zoominfo), without checking that the numbers worked.

Once sent, [Validity](https://www.validity.com/wp-content/uploads/2026/03/2026-Benchmark-Report.pdf) puts inbox placement for opted-in senders at 79.0% to 83.0% at business mailboxes, and [GlockApps](https://glockapps.com/blog/email-deliverability-statistics/) measured 49.0%. The average rep on Gong [connects on 5.4% of dials](https://www.gong.io/files/gong-guide-how-to-master-cold-calls.pdf), and Cognism logged [16.6% across 55,701 dials](https://www.cognism.com/hubfs/2024-Cold-Calling-Report-PDFs.pdf). LinkedIn connection requests are accepted [28.5% of the time](https://expandi.io/blog/linkedin-outreach-benchmarks-2026/) in data from automation tools that LinkedIn's [User Agreement](https://www.linkedin.com/legal/user-agreement) prohibits.

For mail, the office address is public, and the question is whether the buyer sits there. No study measures that, so mail gets the same middle score on contact data that LinkedIn does. Small companies make it harder, since many receive mail at a registered agent, a virtual office or a home. In the Stanford-led survey of working arrangements, [12% of full-time employees were fully remote](https://wfhresearch.com/wp-content/uploads/2026/08/WFHResearch_updates_September2026.pdf), and the USPS Inspector General last put undeliverable mail at [4.3% of volume](https://www.uspsoig.gov/sites/default/files/reports/2023-01/ms-ma-15-006.pdf). Together that is about 84% of office mail reaching the building where someone works at least part of the week. Reaching the building is not reaching the person, past a mailroom or an assistant, so we mark that score as an estimate. If your buyers work from home, ask them for a home address.

Replying is where mail is weakest. A call is a live conversation, and email and LinkedIn take one click. To answer a letter, the person has to scan a code, type a web address or email the rep.

### Crowding, attention and how long it lasts

![A table with inline bars for four channels. Phone: 0.4 robocalls a day, 57 percent of connected calls become conversations. Mail: 2.4 pieces a day, stays 7.8 days in a home, 36 percent read. Email: 117 a day, stays hours on the first screen, 27 percent opened. LinkedIn has no counts.](https://trysincerely.com/images/guides/b2b-outbound-channels-compared-attention.webp "Figure 5. How crowded each channel is, how long a message stays around, and how often it is read or answered. Bars compare channels within each column. Grey text marks a gap in the data.")

The average Microsoft 365 user receives [117 emails a day](https://www.microsoft.com/en-us/worklab/work-trend-index/breaking-down-infinite-workday), internal mail included, and most are skimmed in under a minute. The fair comparison is all mail, which the USPS [2025 annual report](https://about.usps.com/what/financials/10k-reports/fy2025.pdf) puts at 2.4 pieces per mailbox a day. US households say they read [36% of advertising mail](https://prc.arkcase.com/api/prc-dockets/filing/downloadFile?fileId=252658&inline=true) from companies they do not yet buy from, and JICMAIL, a UK panel funded by the mail industry, found direct mail stays [live in the home for 7.8 days](https://www.jicmail.org.uk/wp-content/uploads/2026/06/JICMAIL-Q1-2026-Results-for-Website.pdf). All of those numbers come from homes, so we mark mail's attention scores as estimates. Email's 27% is no firmer, since it counts tracked opens that Apple's privacy feature distorts.

Phones carry little traffic. YouMail counted [11.7 robocalls per person](https://robocallindex.com/) in September 2026, but [80% of Americans](https://www.pewresearch.org/short-reads/2020/12/14/most-americans-dont-answer-cellphone-calls-from-unknown-numbers/) told Pew they do not answer unknown numbers, so the phone loses people before crowding matters. Once a call connects, it holds attention, and in the Cognism study 5,265 of 9,247 connected calls became conversations. Apple's [Mail Privacy Protection](https://support.apple.com/guide/iphone/use-mail-privacy-protection-iphf084865c7/ios) hides email opens, and the 27% is an undated figure from Outreach.

### Email is crowded, and LinkedIn is following

![A line chart on a shared time axis. Cold email replies fell from 0.50 percent in the first half of 2025 to 0.40 percent in the second half, down 20 percent in six months. LinkedIn connection-note replies fell from 3.5 percent in May 2025 to 2.2 percent in April 2026, down 37 percent in eleven months.](https://trysincerely.com/images/guides/b2b-outbound-channels-compared-saturation.webp "Figure 6. Reply rates for cold email and LinkedIn connection notes on a shared time axis, each indexed to 100 at its first reading. Both fall at about the same pace, roughly a third a year.")

Volume is only half of saturation. The other half is what each message is worth once it arrives. Belkins tracked [7.5 million cold emails](https://belkins.io/blog/cold-email-response-rates) through 2025 and saw the reply rate drop from 0.50% in the first half to 0.40% in the second, a 20% fall in six months. Belkins blames inbox saturation and also names a summer slowdown, so part of the drop may be seasonal. On a shared time axis it falls at about the same pace as LinkedIn, roughly a third a year. Filters are tightening at the same time. Validity found B2B inbox placement behind corporate filters [fell 5.8% in 2025](https://www.validity.com/wp-content/uploads/2026/03/2026-Benchmark-Report.pdf), and Gmail began [enforcing its sender rules harder](https://support.google.com/a/answer/14229414) in November 2025.

LinkedIn looks like email did a few years ago. Expandi's data on 13.2 million connection requests shows the reply rate to connection notes falling from [3.5% in May 2025 to 2.2% in April 2026](https://expandi.io/blog/linkedin-outreach-benchmarks-2026/). That data comes from an automation tool, which is part of the problem: tools let one rep send far more requests than anyone could by hand, and the buyer on the other end sees more of them every month. LinkedIn's answer has been limits, from restricting accounts that send too many ignored invitations to allowing free members only [three personal notes a month](https://www.linkedin.com/help/linkedin/answer/a563153).

No source counts the cold emails or LinkedIn messages a buyer receives, so we cannot say how many of those 117 emails are pitches. What the falling reply rates show is that each one is worth less than it was. Mail is moving the other way. US households received [393 pieces of advertising mail](https://prc.arkcase.com/api/prc-dockets/filing/downloadFile?fileId=252658&inline=true) in 2025, down from 659 in 2015, and the share they read rose from 46% in 2020 to 51%. Printing and postage make mail expensive to send carelessly, and that cost keeps the mailbox quiet.

### Visible effort

![Diverging bars showing the change in response rate against printed-only mail. A handwritten note added 16.8 points in the first 15 days and 16.2 points after reminders. A hand-signed letter to doctors added 0.7 points, and a printed sticky note for older adults changed the rate by minus 0.2 points.](https://trysincerely.com/images/guides/b2b-outbound-channels-compared-handwriting.webp "Figure 7. Three randomized studies of a personal touch, the Irish one measured twice, each against the same mailing without it. Only the handwritten note moved the response rate.")

Ireland's tax authority mailed a survey to about 2,000 small businesses and added a handwritten sticky note to 300 packs, chosen at random. Within 15 days, [36.0% of the note group answered, against 19.2%](https://www.revenue.ie/en/corporate/documents/research/business-survey-2013.pdf) of the rest. A tax authority asking for a survey is not a sales rep asking for a meeting, and people feel bound to answer the first. A hand-signed cover letter did nothing for 3,799 UK obstetricians, who answered at [79.1% against 78.4%](https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=PMCID:PMC1190172&resultType=core&format=json), and a printed sticky note did nothing for older adults in a [UK health trial](https://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?db=pubmed&id=27569403&rettype=abstract&retmode=text). Gifts have a randomized result too. In a test of about 10,000 charity letters, a small enclosed gift raised the share who donated by 17% and a large one by [75%](https://econpapers.repec.org/RePEc:ecm:emetrp:v:75:y:2007:i:5:p:1501-1511).

There is a ceiling. Consumers read ad spending as a sign of quality [up to a point](https://academic.oup.com/jcr/article-abstract/17/2/160/1861197), and US federal employees may accept gifts of only [$20 per occasion](https://www.law.cornell.edu/cfr/text/5/2635.204). The handwritten scores assume a person holds the pen. A note written by a pen-holding machine is still a physical object, but on effort it sits closer to a printed letter. A LinkedIn message scores a little above email on effort, because it comes from a real person's profile the buyer can check. No randomized test of handwritten notes or gifts has measured a B2B sales outcome, so the effort scores are judgment calls.

### Cost, scale and speed

![A bubble chart with the cost of one more send on a log scale against how much attention a channel gets. Email sits at a few cents with low attention. Postcards and letters sit near a dollar with high attention. Handwritten notes, gifts and cold calls cost more. Bubble size shows scale.](https://trysincerely.com/images/guides/b2b-outbound-channels-compared-bubble.webp "Figure 8. Cost of one more send against attention, on a log price scale. Bigger bubbles scale further. The green box marks cheap and noticed, and no channel sits inside it.")

A cold email costs 2 to 4 cents to send. That is a [Google Workspace mailbox](https://workspace.google.com/pricing?hl=en&gl=us) at CA$9.20 a month, spread over the [30 to 50 emails a day](https://help.instantly.ai/en/articles/5975326-instantly-cold-email-strategy) that Instantly recommends and the [14 a day](https://www.smartlead.ai/benchmarks/how-many-cold-emails-per-day-per-mailbox) the median Smartlead mailbox sends. A cold call costs $8.70 to $11.05 of rep time a dial, by our arithmetic from the Bridge Group's [$80,000 median pay and 44 dials a day](https://www.bridgegroupinc.com/research/2025-sdr-models-metrics-report-the-bridge-group) and the BLS finding that wages are [70.0% of employer cost](https://www.bls.gov/news.release/ecec.nr0.htm). That makes a call look expensive next to email, whose price leaves out contact data, sending software and spare domains. We count rep time only for calls because a call cannot happen without a person on the line, while writing can be templated in any written channel. Sales Navigator costs [$119.99 a month](https://business.linkedin.com/sell/sales-navigator/compare-plans) and includes [50 InMails](https://www.linkedin.com/help/sales-navigator/answer/a101030), which works out to $1.80 to $2.40 each, but most LinkedIn outreach is connection requests with a note, which cost nothing beyond the seat. That puts a typical LinkedIn touch well under a dollar, an estimate because LinkedIn does not publish its invitation limit.

A 4x6 postcard lists at [$0.80 at Stannp](https://www.stannp.com/us/direct-mail-api) to [$0.905 at Lob](https://www.lob.com/pricing/print-mail), and a black-and-white First-Class letter at [$1.059 at PostGrid](https://www.postgrid.com/pricing-print-mail/) to $1.23 at Stannp. Sincerely is a paying customer of Lob and PostGrid. A machine-written note costs [$3.62 at 100 cards from Simply Noted](https://simplynoted.com/pages/simply-noted-plans), postage included. Genuine handwriting has no public price, so that score is an estimate. Gift platforms publish no per-send price, so we assumed $15 and marked it. Mail takes [5 to 7 business days](https://help.lob.com/print-and-mail/getting-data-and-results/tracking-your-mail) to arrive, while email, calls and LinkedIn arrive in minutes.

Scaling up runs the other way. Instantly asks for [at least two weeks of warmup](https://help.instantly.ai/en/articles/13938078-when-your-email-account-is-ready-for-campaigns) before a new mailbox sends campaigns, and each mailbox should send only 30 to 50 a day, so doubling email volume means buying domains and waiting. Teams that already have warmed mailboxes start sooner, so we score email a 3 rather than a 2, as an estimate. A print API sends thousands of pieces the day you start, within plan limits that run from [6,000 mailings a year on Lob's free plan to 150,000 on its Growth plan](https://www.lob.com/pricing/print-mail). More call volume means hiring reps, and LinkedIn caps each seat at [50 InMails a month](https://www.linkedin.com/help/sales-navigator/answer/a101030), so both scale only as fast as you add people. Those two are estimates, and so are the handwritten and gift scores, since no vendor publishes its capacity.

### Proof, rules and risk

You can test any of these channels with a holdout. Pick accounts at random, leave them out, and compare. Calls log every connect, and email replies are easy to count. Mail has no read signal, so you measure it through the holdout.

The rules and risks differ more than most teams expect. CAN-SPAM [makes no exception for business-to-business email](https://www.ftc.gov/business-guidance/resources/can-spam-act-compliance-guide-business), and Canada's [anti-spam law](https://laws-lois.justice.gc.ca/eng/acts/e-1.6/fulltext.html) needs consent for email but does not cover paper. The [TCPA](https://www.law.cornell.edu/uscode/text/47/227) bars autodialed calls to mobile numbers without consent. Gmail tells senders to [never reach a spam rate of 0.3%](https://support.google.com/mail/answer/81126), LinkedIn [restricts accounts](https://www.linkedin.com/help/linkedin/answer/a550555) that send too many invitations, and the [FCC says](https://docs.fcc.gov/public/attachments/DOC-415059A1.pdf) carriers mislabel some legal calls as spam. Nothing a recipient does can stop your mail from being delivered, and the only opt-out list, [DMAchoice](https://consumer.ftc.gov/articles/how-stop-junk-mail), is voluntary.

## Email, a call and a letter, head to head

![A radar chart comparing cold email, cold call and printed letter on 14 measures. Email spikes on cheap, fast and scales. The call spikes on easy to reply, uncrowded, gets read and measurable. The letter spikes on uncrowded, lasts, scales, few rules and low risk.](https://trysincerely.com/images/guides/b2b-outbound-channels-compared-radar.webp "Figure 9. Three channels on all 14 measures, 1 at the center and 5 at the edge. Each one is strong where the others are weak.")

Figure 9 puts three common ways to open a named account side by side. Their strengths rarely overlap. Email covers cost, speed and scale, the call covers the conversation, and the letter covers getting noticed and staying around. That is the case for running them together. The [multichannel cadence guide](https://trysincerely.com/guides/how-to-build-multichannel-outbound-cadence) covers the order, and [direct mail vs cold calling](https://trysincerely.com/guides/direct-mail-vs-cold-calling) covers timing a call to delivery.

## What a meeting costs

Published meeting rates exist for two channels, and the sources disagree.

| Channel    | Sends per booked meeting         | Source                                            |
| ---------- | -------------------------------- | ------------------------------------------------- |
| Cold email | 344 emails                       | Gong, its customers' email                        |
| Cold email | Up to 6,275 emails               | Belkins, 7.5 million emails in 2025               |
| Cold call  | 45 dials (top rep), 402 (median) | Gong, over 300 million recorded calls             |
| LinkedIn   | No meeting rate; 7.5% of InMails | Belkins, replies only                             |
| Mail       | No published B2B rate            | The ANA's survey figures are gated and mix in B2C |

[Gong's figure](https://www.gong.io/blog/does-cold-email-even-work-any-more-heres-what-the-data-says) comes from in-house sales teams and [Belkins'](https://belkins.io/blog/cold-email-response-rates) from an agency emailing strangers, which may explain an 18-fold gap.

An average rep's cold-call meeting takes about 402 dials, or $3,500 to $4,440 of rep time, so a $1.15 letter matches it if one letter in about 3,000 to 3,900 books a meeting. Against email the bar is much higher. Gong's 344 emails cost $7 to $14 to send, so a letter would need a meeting from roughly one piece in six to twelve, and a postcard from one in 8 to 16. On sending cost, email wins. The [break-even calculator](https://trysincerely.com/tools/direct-mail-break-even) runs this on your own numbers.

What cheap email cannot buy is more replies from a fixed list. Smartlead's median sender gets a reply from [0.74% of contacts](https://www.smartlead.ai/blog/cold-email-reply-rate), about seven out of 900 people at 300 accounts. A tenth email to the same people costs little in money and a lot in reputation, so on a short list the question becomes how many of the 900 ever notice you.

## Where direct mail loses

![Three small bar charts, each sorted on its own. On fast, email, call and LinkedIn score 5 and all mail scores 2. On easy to reply, the call scores 5 and all mail scores 2. On cheap, email scores 5, postcards and LinkedIn 3, calls and handwritten notes 2, and gifts 1.](https://trysincerely.com/images/guides/b2b-outbound-channels-compared-tradeoffs.webp "Figure 10. The three measures where mail scores lowest, each sorted on its own from best to worst. Postcard stands for postcards and letters.")

- **Speed.** Five to seven business days from send to mailbox. A reply needed tomorrow belongs to email or the phone.
- **Replies.** Answering a letter means switching to another channel, so expect fewer direct replies and more indirect ones, such as a search or a reply to the rep's next email.
- **Cost per send.** A First-Class letter costs 30 to 60 times as much as an email to send.
- **Office evidence.** Most of mail's attention scores come from household studies. No study follows mail into an office, so those scores are estimates.
- **Response evidence.** We found no published B2B response rate for mail that we could check. The ANA's 2023 survey figures, quoted on our [statistics page](https://trysincerely.com/guides/b2b-direct-mail-statistics), sit behind a member login and include consumer campaigns.

## How sure are we?

Two kinds of score count as estimates: cells with no direct source, and cells where a study of homes stands in for offices, which covers most of mail's attention scores. That is 41 of the 98. If you think we leaned toward mail, that is where we would have done it. So we reran every total with each estimate pushed one point against mail, down for the mail channels and up for everything else.

![Two dot plots of channel totals, as scored and with every estimate moved against mail. On named accounts, postcards fall from 73 to 63 and the cold call rises to 70. With equal weights, postcards fall from 72 to 66 and cold email rises to 70.](https://trysincerely.com/images/guides/b2b-outbound-channels-compared-stress.webp "Figure 11. Totals for named accounts and for equal weights, before and after the stress test. Filled dots are the published scores, hollow dots the scores with every estimate turned against mail.")

This is where mail's lead gets thin. On named accounts the cold call and LinkedIn finish at 70, ahead of every mail channel. Existing customers become a tie, with gifts at 69 and the cold call at 68. Equal weights and high volume both go to email. Our honest reading is that mail leads the account-based jobs on our best estimates, and that the lead rests on evidence from homes that an office study could overturn. A holdout test on your own list is how to find out which way it falls.

## Limits of this study

- No neutral study measures what share of a B2B list has a working email, phone number and address. Vendors claim 97% to 98% accuracy, which measures the records they return and says nothing about the people they miss.
- Nobody counts how many cold emails or LinkedIn messages a buyer receives. Microsoft's 117 counts all email.
- Every read rate and attention number for mail comes from households, not offices.
- The claim that 22.5% of B2B data decays each year traces to a MarketingSherpa study from about 2012 that we could not find. The claim that 30% of people change jobs each year misreads a BLS count of job separations, which counts events, so one person can count twice. The BLS puts [median job tenure](https://www.bls.gov/news.release/tenure.nr0.htm) at 4.9 years for managers and professionals.
- Cold email reply benchmarks cannot be compared. Belkins measured 0.45% of emails sent and Instantly 3.43%, on the same denominator.
- Quick-to-scale-up scores for calls, LinkedIn, handwritten notes and gifts are estimates. No vendor publishes capacity, and no study measures how long it takes to add reps.
- Postcards and letters score the same on every measure. A letter is private and holds more, and a postcard is read without being opened, but no study measures either difference.
- The weights are ours. A team with different priorities will get different totals, so the appendix lists them all.

Sincerely runs the mail side of this study. It researches each account, writes each postcard, letter, handwritten note or gift note for the person receiving it, and mails it. Handwritten notes are genuinely written by hand. You can hold out a share of accounts on any campaign and measure lift on your own list, which is the one number this study cannot give you.

## Appendix: the scores

Each cell gives the score from 1 to 5, then the number behind it. An asterisk marks an estimate where no direct source exists.

| Channel     | Coverage               | Delivery              | Reply            |
| ----------- | ---------------------- | --------------------- | ---------------- |
| Cold email  | 3 · 55% findable       | 3 · 49% to 83% inbox  | 4 · one click    |
| Cold call   | 2* · 41% to 67% mobile | 1 · 5% to 17% connect | 5 · live         |
| LinkedIn    | 3* · no study          | 2 · 26% to 29% accept | 4 · one click    |
| Postcard    | 3* · office is public  | 4* · about 84%        | 2 · scan or type |
| Letter      | 3* · as postcard       | 4* · as postcard      | 2 · as postcard  |
| Handwritten | 3* · as postcard       | 4* · as postcard      | 2 · as postcard  |
| Gift        | 3* · as postcard       | 4* · as postcard      | 2 · as postcard  |

| Channel     | Crowding                  | Attention                  |
| ----------- | ------------------------- | -------------------------- |
| Cold email  | 1 · 117 emails a day      | 3* · 27% opens, undated    |
| Cold call   | 5 · 0.4 robocalls a day   | 5 · 57% of connects talk   |
| LinkedIn    | 3* · no count exists      | 3* · no count exists       |
| Postcard    | 4* · 2.4 pieces, all mail | 4* · 36%, household report |
| Letter      | 4* · as postcard          | 4* · as postcard           |
| Handwritten | 4* · as postcard          | 4* · no separate data      |
| Gift        | 4* · as postcard          | 4* · no separate data      |

| Channel     | Visible effort                | Lasts                      |
| ----------- | ----------------------------- | -------------------------- |
| Cold email  | 1                             | 3* · hours on first screen |
| Cold call   | 2 · a person's time           | 2* · the length of a call  |
| LinkedIn    | 2* · a real profile           | 3* · no data               |
| Postcard    | 3 · paper and postage         | 5* · 7.8 days in a home    |
| Letter      | 3 · same as a postcard        | 5* · as postcard           |
| Handwritten | 4 · written by hand           | 5* · scored as a postcard  |
| Gift        | 5 · an object chosen for them | 5* · scored as a postcard  |

| Channel     | Cost per send             | Volume per rep           | Arrives         |
| ----------- | ------------------------- | ------------------------ | --------------- |
| Cold email  | 5 · 2 to 4 cents to send  | 4 · 30 to 50 a day       | 5 · at once     |
| Cold call   | 2* · $8.70 to 11.05       | 2 · 44 to 56 dials a day | 5 · at once     |
| LinkedIn    | 3* · notes free with seat | 2 · 50 InMails a month   | 5 · at once     |
| Postcard    | 3 · $0.80 to 0.91         | 4 · plan caps            | 2 · 5 to 7 days |
| Letter      | 3 · $1.06 to 1.23         | 4 · plan caps            | 2 · 5 to 7 days |
| Handwritten | 2* · $3.62 to 4.57        | 3* · no data             | 2 · 4 to 5 days |
| Gift        | 1* · no published price   | 4* · stock limits        | 2* · a week     |

| Channel     | Quick to scale up                               |
| ----------- | ----------------------------------------------- |
| Cold email  | 3* · two weeks per new mailbox                  |
| Cold call   | 1* · more volume means hiring reps              |
| LinkedIn    | 1* · fixed limits on every seat                 |
| Postcard    | 5 · same day through a print API                |
| Letter      | 5 · as postcard                                 |
| Handwritten | 4* · ships in 1 to 2 days, larger orders slower |
| Gift        | 3* · inventory and minimum orders               |

| Channel     | Measurable          | Rules                 | Risk                |
| ----------- | ------------------- | --------------------- | ------------------- |
| Cold email  | 4 · replies only    | 2 · consent in Canada | 2 · domain filtered |
| Cold call   | 5 · all logged      | 3 · dialer consent    | 3 · spam labels     |
| LinkedIn    | 4 · replies visible | 3 · no bots           | 2 · account limits  |
| Postcard    | 3 · no read signal  | 5 · no consent rules  | 5 · none            |
| Letter      | 3 · as postcard     | 5 · as postcard       | 5 · as postcard     |
| Handwritten | 3 · as postcard     | 5 · as postcard       | 5 · as postcard     |
| Gift        | 3 · as postcard     | 4 · gift limits       | 5 · as postcard     |

The weights behind each job's total come first. Each column adds up to 100, and a measure with a weight of 0 plays no part in that job. The decimals come from shrinking the original ten weights to make room for the four measures added later.

| Measure                   | High volume | This week | Named | Existing |
| ------------------------- | ----------- | --------- | ----- | -------- |
| Contact data available    | 12.75       | 9.5       | 8.5   | 0        |
| Reaches the person        | 8.5         | 14.25     | 12.75 | 12.75    |
| Easy for them to reply    | 5           | 5         | 5     | 5        |
| Uncrowded channel         | 4.25        | 4.75      | 12.75 | 8.5      |
| Gets read or answered     | 8.5         | 14.25     | 12.75 | 12.75    |
| Shows personal effort     | 0           | 0         | 12.75 | 21.25    |
| Long lasting              | 0           | 0         | 5     | 5        |
| Low cost per send         | 21.25       | 9.5       | 4.25  | 4.25     |
| Volume per rep            | 17          | 4.75      | 0     | 0        |
| Quick to scale up         | 5           | 0         | 0     | 0        |
| Arrives fast              | 4.25        | 28.5      | 4.25  | 4.25     |
| Easy to measure           | 4.25        | 4.75      | 8.5   | 12.75    |
| Few legal rules           | 4.25        | 4.75      | 8.5   | 8.5      |
| Low risk of being blocked | 5           | 0         | 5     | 5        |

The fit scores for each job follow. Under equal weights, which leave the jobs out and give each row of Figure 1 the same weight, postcards and letters come to 72, handwritten notes to 68, gifts to 67, cold email to 65, cold calls to 61 and LinkedIn to 58. Using only the ten measures we set before seeing any totals, high volume goes to cold email at 74 against 70 for postcards, and named accounts and existing customers still go to mail.

| Channel     | High volume | This week | Named | Existing |
| ----------- | ----------- | --------- | ----- | -------- |
| Cold email  | 72          | 75        | 53    | 52       |
| Cold call   | 55          | 72        | 66    | 66       |
| LinkedIn    | 55          | 70        | 58    | 57       |
| Postcard    | 72          | 63        | 73    | 72       |
| Letter      | 72          | 63        | 73    | 72       |
| Handwritten | 63          | 60        | 75    | 76       |
| Gift        | 60          | 58        | 75    | 77       |

Read these totals next to Figure 11. The mail leads on named accounts and existing customers are real on our best estimates, and they shrink to a tie or less when every estimate breaks against mail.

## Related questions

- [B2B direct mail statistics, with sources and limits](https://trysincerely.com/guides/b2b-direct-mail-statistics): Direct mail statistics traced to the original publisher, with the year, what each number measures, and whether it says anything about B2B mail.
- [How to build a multichannel outbound cadence](https://trysincerely.com/guides/how-to-build-multichannel-outbound-cadence): Give email, phone, LinkedIn, events, and physical mail distinct jobs. Coordinate outreach across an account and measure the pipeline lift.
- [Direct mail vs cold calling](https://trysincerely.com/guides/direct-mail-vs-cold-calling): Mail and cold calls are not rivals. Mail warms the call. Compare cost per meaningful touch, reach limits, and how to time a call to delivery.
- [Direct mail vs LinkedIn ads](https://trysincerely.com/guides/direct-mail-vs-linkedin-ads): LinkedIn ads buy impressions at scale. Direct mail buys attention at named accounts. Compare targeting, cost per touch, and attribution.
- [What to do when cold email stops working](https://trysincerely.com/guides/cold-email-not-working): Fix cold email that stopped working. Diagnose delivery, targeting, timing, and offer before adding another channel.
- [Direct mail ROI and break-even calculator](https://trysincerely.com/tools/direct-mail-break-even): Find the response rate where a direct mail campaign pays for itself. Enter total campaign cost and cost per response to see break-even.

---

Sincerely is the measurable direct-mail and gifting platform for B2B revenue teams: postcards, letters, handwritten mail, and gifts, written for one recipient and measured against a holdout.

Contact Sincerely: https://trysincerely.com/contact

Agent routing index: https://trysincerely.com/llms.txt