How to build an email marketing report that counts replies, meetings and pipeline.

A 40% open rate next to zero closed deals is a common line in a monthly email marketing report, and the usual response is to rewrite the subject lines. Both numbers can be wrong. The open rate counts a lot of machines, and the zero can be off because the CRM never connected a closed deal to the email that started it.

The next monthly report is the moment to change what it counts.

Why the open rate in an email marketing report counts machines.

Apple's support page says Mail Privacy Protection hides a recipient's opens from senders. Platforms record an open when the tracking image loads, and with the feature on, it can load without a person seeing the message.

Litmus, which sells email analytics, counted 76.56% of the 1.5 billion opens it tracked in September 2026 as Apple, Mail Privacy Protection included. That describes tracked opens across Litmus's customers, not readers, and Litmus notes every sender's mix differs. Check your own platform's email client report first. Classic Outlook blocks automatic picture downloads by default, which hides some B2B readers too.

Three in four opens Litmus tracked came from Apple.

Share of tracked email opens by email client, September 2026

Three in four opens Litmus tracked came from Apple.

Item	Value
Apple (includes Mail Privacy Protection)	76.56%
Gmail	16.10%
Outlook	3.48%
Yahoo Mail	2.35%
Other top-10 clients	1.45%

"An 'open' is just someone (or something) fetching a remote image." Steve Atkins, Word to the Wise, 2024

Opens still have one job. Laura Atkins of Word to the Wise pointed out in April 2026 that mail in the spam folder doesn't load images, which makes a sudden drop at one mailbox provider a useful delivery warning.

Clicks need a filter too.

Clicks have a separate problem. Constant Contact's help center explains that security filters often follow every link before a message reaches the person, and Mailchimp's internal testing, published in October 2025 (method and sample not published), found about 1 in 5 clicks are bots with basic detection, rising to close to half with third-party data added.

Check these settings before the next report:

• HubSpot: bot filtering is on by default, covering security scanners, Apple's Mail Privacy Protection and corporate screeners such as Mimecast. The switch is under Settings, Tracking & Analytics, Tracking Code, Advanced Tracking, Email Bot filtering.
• Mailchimp: bot filtering is on by default for Mail Privacy Protection opens and bot clicks, under Analytics, Analytics settings, Bot activity. Filtered clicks start August 10, 2025.
• Constant Contact: bot click filtering is a beta, not yet open to every account and off by default, under account settings, Advanced settings.
• Salesforce Marketing Cloud Account Engagement or any other platform: ask the admin what it filters, then spot-check a send. Several links clicked at the same moment often point to a security program.

At a larger company, print each system's filter in the report footer, which keeps filtered and unfiltered numbers from being mixed.

Why the zero can be wrong too.

The lag. Dreamdata's 2026 benchmark, from 3.5 million customer journeys at thousands of B2B companies (Dreamdata sells attribution software), put the B2B customer journey at 272 days, up from 211, and the first 220 came before a deal entered the pipeline. A 30-day report window sees the first month of a nine-month process.

Most of a B2B journey happens before a deal exists.

Days in the B2B customer journey, Dreamdata 2026

Most of a B2B journey happens before a deal exists.

Item	Value
Before a deal enters the pipeline	220 days
Whole journey	272 days

The buying group. Forrester's State of Business Buying, 2026, based on a survey of more than 17,500 business decision makers, found the typical buying decision includes 13 people inside the company and nine outside influencers. HubSpot's attribution credits email only through a marketing email click by a contact on the deal, and deal and revenue attribution need Marketing Hub Enterprise. A reply logged in the inbox is credited to a conversation or a sales email rather than to the newsletter, and readers not on the deal earn the email nothing.

The few companies shopping now. John Dawes of the Ehrenberg-Bass Institute wrote in 2021 that up to 95% of business clients aren't in the market for many services at any one time, with providers like a main bank or law firm changed about once every five years. His reasoning gives a quick estimate:

• Share of accounts in the market this month ≈ 1 ÷ months between purchases in your category
• Accounts likely to choose a provider this month ≈ accounts on the list × that share

Eight checks before anyone rewrites a subject line.

Confirm a clear ask, a clean list and good delivery first.

1. What was the email built to do?

Emails that keep existing clients support renewals and repeat orders, and they rarely show up as new deals. Name each send with its job and year, like "2026-11-newsletter-existing-clients" and "2026-11-newsletter-new-business", and report them apart.

2. How many replies and form fills did it bring in?

Send replies to a monitored inbox and count them per 1,000 delivered. Tag every link (utm_source=newsletter, utm_medium=email, utm_campaign=2026-11-newsletter-new-business). Google Analytics files a session under Email when the source or medium is "email", and a hidden form field carries the campaign into the CRM. With no CRM, use a spreadsheet with one row per reply.

At a larger company, replies go to a shared queue with a named owner, logged on the contact record.

3. How fast does someone answer a reply?

A reply that waits a few days usually goes cold. Set a same-business-day rule and report the median hours to first answer, with routing from the form-to-first-call playbook.

4. Can the CRM connect a closed deal to an email?

Without a field or tag, a deal that started with a reply or a forward usually gives email no credit, even when it helped. Add the required deal field from the second asset below. In HubSpot, it's a custom deal property made required by a conditional stage property on the second stage (Settings, Data Management, Objects, Deals, Pipelines). In Salesforce, it's an opportunity picklist with a validation rule once the deal moves past the first stage, and with no CRM, a column in the deals sheet.

5. Does anyone ask new clients what they read?

Many mention an email the CRM never connected to the deal. Ask on the first call and log the answer word for word.

6. Does the list reach the decision makers?

Count how many people from each account in open pipeline are on the list. One contact against a group of 13 leaves most of the group unreached.

When Stephen Mather, a former borax advertising manager, mailed the National Parks Portfolio in 1916, he drew names from the rolls of clubs, professional societies, universities and chambers of commerce, and added every member of Congress, the people who voted to create the National Park Service that August. The list was built around the decision.

7. Does the content connect to what the business sells?

Give each edition a question or link back to a service you sell.

8. How long does a deal take to close, and where does silence go?

An email sent in March can help close a deal in September, long after the March report showed zero. Read results as monthly groups of engaged contacts, kept open for your usual time from first contact to a closed deal (longer than a deal's time in the pipeline), the cohort view from the one-page marketing report.

A prospect who came inbound or took a call and then stopped replying hasn't said no, and The JOLT Effect (Dixon and McKenna, 2022, built on more than 2.5 million sales conversations) treats indecision as its own outcome. Move those deals to Closed No Decision, not Closed Lost, with a date to try again and a monthly email with no sales ask, until their timing changes or a year passes with no reply or website visit. In HubSpot, make it a Lost stage at 0% and filter it out of win-rate reports.

A worked example.

Example numbers: a firm sends a monthly newsletter to 4,000 contacts at 600 companies, and its clients switch providers about every four years.

• One month's report shows 3,900 delivered, 1,600 opens (41%) and zero deals.
• The people line reads 9 replies (9 ÷ 3,900 × 1,000 = 2.3 per 1,000 delivered), 4 form fills from email links and 3 booked meetings.
• In-market estimate: 600 companies × 1 ÷ 48 months ≈ 12 or 13 companies shopping in a given month, a small pool, which is why the month is judged on replies and meetings rather than deals.
• With nine months from first contact to a closed deal, the month's group stays open for nine months, and its zero deals say nothing yet.

The one-page email report, ready to copy.

Common mistakes, and what a specialist will push back on.

"Turn on multi-touch attribution." Models only split credit among touches the system recorded, and a forward to a colleague isn't one of them. A 2019 Marketing Science paper by Gordon and colleagues found that observational methods often failed to recover the effects measured in large randomized ad experiments at Facebook (ads shown to Facebook users, not B2B email). Treat attribution as one view next to the deal field and the first-call answer.

"Self-reported answers are unreliable." Read them as patterns across a quarter, and they catch the forwarded newsletter tracking misses.

A four-edition test to start this month.

The four-edition reply test, for most lists under about 32,000 contacts (the holdout below needs about 15,900 per group).

1. Pick one change that gives readers a reason to answer, like building each edition around a question a client asked that week and ending on one direct question.
2. Send the next four editions with the change to the same list, at the same day and time.
3. Measure replies per 1,000 delivered and meetings booked within 30 days of each send, against the last four editions.
4. Count it as a result when replies per 1,000 delivered at least double for all four, with no rise in unsubscribes or spam rate. Check the same months last year too, if you have them.

A holdout, for large lists. Hold back a random part of the list for 90 days and compare meetings per contact. Lewis and Rao's 2015 Quarterly Journal of Economics study of 25 large digital ad experiments with US retailers and brokerages (consumer sales, not B2B) found the median confidence interval on return on investment was over 100 percentage points wide. Size it with Lehr's crude rule (Statistics in Medicine, 1992), which assumes 80% power and 5% significance: contacts per group ≈ 16 × p × (1 − p) ÷ d², where p is the average of the current and target meeting rates and d is the lift to detect.

Example numbers for a larger list: a 0.5% meeting rate over 90 days and a target of 0.75% gives an average p of 0.625%, and 16 × 0.00625 × 0.99375 ÷ 0.0025² ≈ 15,900 contacts per group. A list of 4,000 can't run it, and the reply test and the deal field will teach more. Work out what a lead is worth before you judge either one.

Frequently asked questions.

What should an email marketing report include?

A delivery check (bounces, complaints, opens by provider), what people did (replies, form fills from email links), what it led to (meetings, deals that started with an email, new clients who named one), a lag view by month of engagement, and one decision for next month.

How do you attribute revenue to email marketing?

Combine three sources. Tag every link with UTM parameters captured in a hidden form field, add a required deal field for where the first conversation came from, and ask new clients what they read before reaching out. In HubSpot, email attribution counts only clicks, and deal and revenue attribution need Marketing Hub Enterprise.

How long does email marketing take to show results in B2B?

Longer than a monthly report. Dreamdata's 2026 benchmarks put the B2B customer journey at 272 days. Read email results as monthly groups of engaged contacts, kept open for the usual time from first contact to a closed deal, and judge the first months by replies and meetings.

What is a good reply rate for a B2B newsletter?

I haven't found a credible published benchmark for newsletter replies, and vendor averages mix cold email with newsletters. Set your own baseline from the last four editions as replies per 1,000 delivered. In the reply test above, a change counts when that rate at least doubles for four editions in a row.

Sources.

• Apple, Use Mail Privacy Protection on iPhone, iPhone User Guide, 2026. support.apple.com
• Litmus, Email Client Market Share, September 2026 data. litmus.com
• Steve Atkins, Prefetches and Proxies, Word to the Wise, 2024. wordtothewise.com
• Laura Atkins, The spam folder and opens, Word to the Wise, 2026. wordtothewise.com
• Constant Contact, Preventing bot clicks from being represented in your email reporting, 2025. knowledgebase.constantcontact.com
• Mailchimp, Unmasking Bot Clicks in Email Marketing Analytics, 2025. mailchimp.com
• Mailchimp, About Bot Activity and Bot Filtering, 2026. mailchimp.com
• HubSpot, Understand bot filtering in marketing email analytics, 2026. knowledge.hubspot.com
• HubSpot, Understand attribution reporting, 2026. knowledge.hubspot.com
• HubSpot, Set up and manage object pipelines, 2026. knowledge.hubspot.com
• Microsoft, Block or unblock automatic picture downloads in classic Outlook email messages, 2026. support.microsoft.com
• Salesforce Trailhead, Create Validation Rules, 2026. trailhead.salesforce.com
• Dreamdata, LinkedIn Ads Benchmarks Report 2026 (press release with methodology), 2026. prnewswire.com
• Forrester, The State Of Business Buying, 2026 (press release), 2026. forrester.com
• Forrester, Buyer Insights 2026 (survey base), 2026. forrester.com
• John Dawes, Ehrenberg-Bass Institute, Advertising effectiveness and the 95-5 rule, 2021. marketingscience.info
• Google, Email sender guidelines, 2026. support.google.com
• Google, Set up Postmaster Tools, Gmail Help, 2026. support.google.com
• Google, [GA4] Default channel group, 2026. support.google.com
• Matthew Dixon and Ted McKenna, The JOLT Effect, Portfolio, 2022. penguinrandomhouse.com
• Brett Gordon, Florian Zettelmeyer, Neha Bhargava and Dan Chapsky, A Comparison of Approaches to Advertising Measurement: Evidence from Big Field Experiments at Facebook, Marketing Science, 2019. doi.org
• Randall Lewis and Justin Rao, The Unfavorable Economics of Measuring the Returns to Advertising, Quarterly Journal of Economics, 2015. doi.org
• Robert Lehr, Sixteen S-squared over D-squared: a relation for crude sample size estimates, Statistics in Medicine, 1992. doi.org
• National Park Service, 50 Nifty Finds #35: On the Same Track, 2023. nps.gov