# HubSpot Email Tracking: 8 Reasons Opens, Clicks and Logs Disagree

HubSpot email tracking measures supported open and click events; email logging stores an activity in the CRM. Neither proves that a particular person read, understood or wanted the message. When the numbers disagree, identify the email type, sending route, tracking policy and report population before changing the setup. Use replies and completed business actions to judge the next sales step, with tracking as supporting evidence.

Canonical URL: https://www.martechdevs.com/blog/hubspot-email-tracking

By [Jafrin Akter Alo](https://www.martechdevs.com/authors/jafrin-akter-alo)

Data Engineer and Co-founder at [Martechdevs](https://www.martechdevs.com) and [Growth and Analytics Partners](https://www.growthandanalytics.com)

[LinkedIn](https://linkedin.com/in/jafrinalo)

Last updated: 2026-10-25

Next step: [Salesforce to HubSpot migration: plan, field mapping and cutover checklist](https://www.martechdevs.com/blog/salesforce-to-hubspot-migration)

## What is the difference between tracking, logging and delivery?

Treat tracking, logging and delivery as separate outcomes. A message can arrive and be logged while producing no measurable open, or generate a tracking notification without appearing in the report you expected.

| Evidence | What it establishes | What it does not establish | Where to investigate |
| --- | --- | --- | --- |
| Mail sending or delivery evidence | The sending system processed the message or the destination received it | Human attention or purchase intent | Mailbox and receiving-system evidence |
| CRM email activity | A supported capture route stored an activity | That an open-tracking pixel existed | Record timeline and logging rules |
| Open event | A qualifying image load or inferred open was recorded | Which person read every word | Tracking method and event context |
| Click event | A tracked link interaction was recorded | That it was the intended buyer making a decision | Recipient attribution and automation filtering |
| Reply or completed action | A response or business event occurred | That every reply is positive or every action belongs to this email | Message content and business records |

HubSpot provides separate one-to-one tracking and logging controls. Its account setting can prohibit new tracking while leaving logged email activity available. Changes apply to subsequent sends, not previously captured history. [One-to-one tracking and logging controls](https://knowledge.hubspot.com/connected-email/turn-off-one-to-one-email-tracking-or-attachment-logging)

An administrator should be able to state the intended outcome for a message before testing it. “Log the account conversation but do not track recipient engagement” is a valid policy, not a broken configuration.

For installation and mailbox compatibility, use the [HubSpot Outlook integration guide](https://www.martechdevs.com/blog/hubspot-outlook-integration). This article focuses on interpreting and diagnosing the resulting evidence. Our [CRM and lifecycle work](https://www.martechdevs.com/crm-lifecycle) connects that evidence to sensible sales and service actions.

## Which eight checks explain mismatched evidence?

Use the symptom to choose the next check rather than reinstalling the email integration immediately. These eight diagnostic categories separate configuration, measurement and reporting problems.

| Check | Typical observation | Evidence needed | First action |
| --- | --- | --- | --- |
| 1. Tracking versus logging | Email body exists but no engagement appears | Actual Log/Track choice and account policy | Confirm whether tracking was intended |
| 2. Sending route and seat | One sender gets clicks while another does not | Mail path, active add-in and assigned entitlement | Test the same supported path |
| 3. Image and privacy behavior | Recipient replies despite no pixel open | Client behavior and received message | Keep absent measurement distinct from absent reading |
| 4. Inferred opens | First open has the reply or click timestamp | Prior open count and first engagement | Recognize the inference |
| 5. Recipient identity | A Someone notification or repeated opens | To/CC/BCC and forwarded-message context | Avoid assigning an unidentified event to a person |
| 6. Link format | One link records clicks and another does not | Received HTML and visible link text | Check one-to-one link requirements |
| 7. Filtering and self-activity | Instant or missing test notifications | Bot-filter status and sender's own activity | Separate test interference from recipient evidence |
| 8. Report scope | Timeline, sales report and marketing report disagree | Source, filters, cohort and metric denominator | Reconcile like-for-like populations |

Retain the message reference, sender, account, sending route and relevant settings. A screenshot of a missing notification rarely contains enough context to distinguish these cases.

## Which settings should you inspect for a one-to-one email?

Inspect the account's permission to track, the actual message choice and the user's click-tracking entitlement. A default preference and an account prohibition are different controls.

In Settings → Data Management → Objects → Activities, review the account's Email Log & Track settings. Account access permission is required to change them. Account-wide defaults affect the Sales Chrome extension and Office 365 add-in; the legacy Outlook desktop add-in does not use those defaults. [One-to-one account settings](https://knowledge.hubspot.com/connected-email/turn-off-one-to-one-email-tracking-or-attachment-logging)

For one-to-one clicks, HubSpot requires Sales Hub Starter, Professional or Enterprise and an assigned Sales Hub seat. [Sales email tracking requirements](https://knowledge.hubspot.com/connected-email/understand-hubspot-sales-email-open-and-click-tracking)

Sending from a CRM record also has its own prerequisites: an individual work email connection requires an eligible Core, Sales Hub or Service Hub seat, and users need access to the contact. A View-Only seat cannot connect that mailbox. [One-to-one sending requirements](https://knowledge.hubspot.com/one-to-one-email/send-and-reply-to-one-to-one-emails)

For a controlled test, choose an existing contact whose tracking is permitted, use a new message and record the exact Log and Track choices. Test the real client used by the salesperson. Do not assume a successful CRM send proves that an Outlook or Gmail compose path has identical controls.

Keep recipient eligibility intact. An account's stored processing basis is not a field to fabricate because someone wants an open notification. Apply the organization's approved tracking policy; a legitimate restriction can mean the correct result is no tracking.

## Why does a logged email have no open or click data?

Logging can succeed without tracking, so first establish how the activity entered the CRM. Retrospective capture does not manufacture past engagement evidence.

Manual use of HubSpot's BCC address records email content but does not add open or click tracking. Its permitted sender, access, recipient and size rules differ from a tracked send; the total message limit is 40 MB including headers and attachments. [BCC and forwarding rules](https://knowledge.hubspot.com/connected-email/log-email-in-your-crm-with-the-bcc-or-forwarding-address)

The account's incoming-email rule also matters. Log replies only captures replies to a logged thread. Log all emails from known contacts covers correspondence with existing contacts through connected individual work emails. A separate beta can create contacts from incoming messages. [Contact email logging rules](https://knowledge.hubspot.com/connected-email/manage-contact-email-logging-rules)

If the activity exists but the body is redacted, check the account's email-body redaction beta. It preserves sender, recipients, subject and attachments while removing the body from newly logged messages. Turning it off cannot restore those bodies. Confirm the intended policy before treating the missing text as a sync failure. [Email-body redaction behavior](https://knowledge.hubspot.com/connected-email/manage-contact-email-logging-rules)

Do not change a whole account's capture policy to repair one missing event. First decide whether the message was supposed to be stored and where. A broad logging change can capture correspondence that was intentionally outside the CRM.

Never Log is another boundary. It excludes specified contacts from future logging and contact creation, but activity can still appear in the activity feed. It does not remove historical activities or govern third-party integration email. [Never Log scope](https://knowledge.hubspot.com/connected-email/add-emails-and-domains-to-the-never-log-list)

That distinction explains why “I can see a notification” and “the contact has no email activity” are not automatically contradictory. Review the correct surface and the intended policy before deleting records or weakening an exclusion.

## Why can a recipient reply without a recorded open?

A person can read an email while its tracking image remains blocked. The absence of a pixel event is missing measurement, not a reliable statement about attention.

Blocked images or plain-text reading can prevent HubSpot's tracking pixel from loading. When an email has no recorded opens, a subsequent tracked click or reply can create one inferred open at that engagement's timestamp; later interactions do not repeat that initial inference. [Open tracking and inference](https://knowledge.hubspot.com/connected-email/understand-hubspot-sales-email-open-and-click-tracking)

A hypothetical sequence makes the distinction clear:

| Time | Actual action | Available evidence | Interpretation |
| --- | --- | --- | --- |
| 09:00 | A permitted tracked email is sent | Sent activity | No recipient engagement established |
| 09:05 | Recipient reads with images blocked | No pixel load | Reading may remain unmeasured |
| 09:12 | Recipient replies | Reply plus possible first inferred open | The open timestamp reflects inference, not necessarily first reading |
| 09:20 | Recipient replies again | Another reply | Do not infer another first-open event from this rule |

A sales process should react to the reply's meaning. An out-of-office response, an objection, a requested meeting and a request to stop need different actions, even if each appears as a reply event.

If the message never arrived, investigate the actual sending or bounce evidence separately. The [bulk-sender requirements guide](https://www.martechdevs.com/blog/bulk-sender-requirements) covers authentication and delivery obligations; changing an SPF record is not a repair for an intentionally blocked image.

## Why are there repeated opens or an anonymous Someone event?

One tracking image can be observed more than once or by more than one recipient. Do not convert that count into a claim about a particular buyer's interest.

Multi-recipient sales emails share a tracking image. Unattributed Someone events stay outside email activity reports. Forwarding and reused pixels can also produce repeated activity. [Recipient attribution limits](https://knowledge.hubspot.com/connected-email/understand-hubspot-sales-email-open-and-click-tracking)

Use a fresh compose or approved template when testing, rather than copying an already sent message. Keep group emails out of a person-level interpretation until attribution is established. A large account with several stakeholders especially needs the distinction between “someone interacted with this message” and “the budget owner engaged.”

Inspect your own activity as well. HubSpot documents self-open suppression for the Sales extension and legacy Outlook add-in; configuration conflicts can interfere, and an excluded mobile IP can change. A test from the sender's Sent folder is therefore a poor stand-in for an independent recipient. [Self-open handling](https://knowledge.hubspot.com/connected-email/block-self-open-notifications-with-the-hubspot-sales-outlook-add-in-or-chrome-extension)

For a meaningful diagnostic, use a separate controlled recipient and record what that recipient actually did. A clean test should establish the supported mechanism, not attempt to make every notification counter increase.

## Why does one link track while another does not?

For a one-to-one email, check the received link's format before blaming the destination website. HubSpot does not track every visual representation of a URL in the same way.

One-to-one click tracking requires hyperlinked non-URL text, not a displayed URL linked to itself. It also requires HTML email. [One-to-one link rules](https://knowledge.hubspot.com/connected-email/understand-hubspot-sales-email-open-and-click-tracking)

Use a descriptive label as the controlled HTML specimen:

```html
<a href="https://example.com/guide">Read the implementation guide</a>
```

The reserved example domain is a placeholder. Insert your real test destination using the email editor's hyperlink control, then inspect the received message and final landing URL. Preserve any intended query parameters. Test from a separate controlled recipient environment, and check the redirect and the reported event independently.

Do not apply the sales-email plain-text restriction to every HubSpot email type. Marketing email has separate HTML and plain-text click-tracking controls. Use the configuration for the message you actually sent. [Marketing click controls](https://knowledge.hubspot.com/marketing-email/manage-marketing-email-open-and-click-tracking)

## How do marketing-email settings change the diagnosis?

Check the marketing email's own tracking settings and reporting population. Sales-email permissions, marketing tracking and website source attribution answer different questions.

Marketing tracking is configured under Settings → Marketing → Email → Tracking. Turning off open tracking also disables HTML and plain-text click tracking. Per-email disabling requires account tracking to be on and the source-tag option set to add tags to all URLs. These changes apply only to unsent messages. [Marketing tracking settings](https://knowledge.hubspot.com/marketing-email/manage-marketing-email-open-and-click-tracking)

Review source tagging independently. HubSpot offers adding tags to every URL, adding them only where absent, or not adding them; the first option overwrites existing custom tags. Identity tracking is another control. [Source and identity settings](https://knowledge.hubspot.com/marketing-email/manage-marketing-email-open-and-click-tracking)

For a campaign test, inspect the final landing URL and confirm that your intended UTM values survive. The [UTM naming-convention guide](https://www.martechdevs.com/blog/utm-naming-convention) helps keep campaign naming consistent. A working email click counter does not prove your analytics system assigns the same source or campaign.

Forwarding adds an identity complication: a tracked marketing-email click from a forwarded copy can remain attributed to the original contact. Do not treat the cookie association as proof of the browser user's identity. [Forwarded marketing-email attribution](https://knowledge.hubspot.com/marketing-email/manage-marketing-email-open-and-click-tracking)

This matters when a salesperson forwards a campaign to a colleague or a buyer shares a link internally. Keep personalization and customer-specific actions behind their proper identity checks rather than trusting the original recipient label alone.

## How should privacy protection and bot filtering affect your conclusion?

Treat image loads and link requests as measured events with known limits. Review filtering status before interpreting a sudden change as better or worse content.

Apple's Protect Mail Activity downloads remote content in the background regardless of engagement and hides the user's network identity through relays. An image request therefore need not correspond to the moment a person read the email. [Apple Mail Privacy Protection](https://www.apple.com/legal/privacy/data/en/mail-privacy-protection/)

HubSpot's marketing bot filter excludes suspected scanner, privacy-filter and corporate-screening activity. It is enabled by default, and the sent email's performance page shows its status. The setting is configurable, so record the actual state for the send being compared. [Marketing email bot filtering](https://knowledge.hubspot.com/marketing-email/understand-bot-filtering-in-marketing-email-analytics)

Do not turn filtering off merely to restore a larger open rate. Compare delivery, replies, useful website actions and complaints alongside the measurement change. A filter is also not a guarantee that every remaining event is a human decision.

Keep the approved privacy policy separate from troubleshooting. If tracking is intentionally disabled for a recipient or message, the useful work is to make reporting and follow-up tolerate that missing signal. Do not ask recipients to weaken their privacy settings to make the dashboard complete.

## Why can click-through rate improve when clicks stay unchanged?

A rate can rise because its denominator fell. Confirm the product's metric definition before presenting that movement as stronger engagement.

For marketing email, HubSpot defines open rate as openers divided by recipients who received the email, click rate as clickers divided by those recipients, and click-through rate as clickers divided by openers. Its HTML click map also groups links to the same URL even when UTM parameters differ. [Marketing email metric definitions](https://knowledge.hubspot.com/marketing-email/analyze-your-marketing-email-campaign-performance)

In this hypothetical same-send reconciliation, keep 1,000 deliveries and sixty unique clicking recipients fixed. Only the reported opener population changes:

| Metric | View A: 240 openers | View B: 120 openers | What changed |
| --- | ---: | ---: | --- |
| Open rate | 24% | 12% | Reported opener count changed |
| Click rate | 6% | 6% | No change in clickers or deliveries |
| HubSpot click-through rate | 25% | 50% | Fewer reported openers, not more clicks |

The arithmetic is:

```text
Open rate = unique openers / delivered recipients
Click rate = unique clickers / delivered recipients
Click-through rate = unique clickers / unique openers
```

These are definitions for a consistent marketing-email cohort, not an instruction to combine a sales-email activity count with a campaign-delivery count. If the denominator is zero, report the rate as undefined; do not label zero openers and positive clicks a zero-percent result.

Retain raw counts next to rates. Record the email, recipient population, reporting window and filtering configuration. When comparing campaigns, also examine audience changes and send volume before attributing the difference to a subject line or CTA.

## Why can sales-email reports exclude activity visible elsewhere?

A CRM timeline is not the same reporting population as sales-email analytics. Compare the source of each activity before treating a lower report count as data loss.

HubSpot excludes suspected non-human activity, retroactively logged messages, forwarding-address messages and manually logged email from sales-email analytics. Logged but untracked one-to-one email does not supply open or click information. [Sales-email reporting exclusions](https://knowledge.hubspot.com/one-to-one-email/send-and-reply-to-one-to-one-emails)

Create a short reconciliation sample. For each message, record the activity ID, original sending route, capture route, logging/tracking state, associated contact and report filters. Separate excluded activity from a genuinely missing record.

If one email appears on a contact, company and deal, compare the activity identifier before calling it three sends. Association fan-out can be intentional; a report that counts associations rather than distinct email activities can inflate workload without any duplicate delivery.

The [HubSpot sequences versus workflows guide](https://www.martechdevs.com/blog/hubspot-sequences-vs-workflows) helps when engagement feeds an automated next step. Require the action's own eligibility and stop conditions. A missing open should not keep an already answered prospect in a reminder loop.

## What should a controlled tracking test prove?

Test the configured mechanism and the downstream decision separately. A passing pixel test does not validate a lead score or a sales follow-up.

1. Select one email type, sender and supported client route.
2. Record the account, recipient policy, relevant seat and message settings.
3. Compose a fresh message with one descriptive hyperlink and harmless content.
4. Send it once to a controlled recipient and retain delivery evidence.
5. Inspect the CRM activity, associations and tracking events separately.
6. Repeat a deliberately untracked or image-blocked case where appropriate.
7. Compare the resulting report population and rate definitions.
8. Confirm the workflow uses the evidence according to the approved policy.

![Identify the email route and intended evidence. If tracking is not permitted, retain that policy and adjust the expectation. Otherwise verify settings and a controlled send, repairing mismatched mechanisms or reporting. Both paths require approved action eligibility and stopping rules before the workflow runs.](https://www.martechdevs.com/assets/blog/hubspot-email-tracking.svg)

Diagram connections:

- Identify email type sending route and intended evidence to Tracking is permitted for this message?. 
- Tracking is permitted for this message? to Keep the approved policy and correct the measurement expectation: No. 
- Tracking is permitted for this message? to Verify settings and inspect a controlled send: Yes. 
- Verify settings and inspect a controlled send to Mechanism and reporting scope match the test?. 
- Mechanism and reporting scope match the test? to Repair the failed setting route format or report definition: No. 
- Repair the failed setting route format or report definition to Identify email type sending route and intended evidence. 
- Mechanism and reporting scope match the test? to Review the proposed sales action with stronger outcome evidence: Yes. 
- Keep the approved policy and correct the measurement expectation to Review the proposed sales action with stronger outcome evidence. 
- Review the proposed sales action with stronger outcome evidence to Action eligibility and stopping rules pass?. 
- Action eligibility and stopping rules pass? to Hold the action and repair its decision rules: No. 
- Hold the action and repair its decision rules to Review the proposed sales action with stronger outcome evidence. 
- Action eligibility and stopping rules pass? to Use the bounded workflow and monitor exceptions: Yes

## Checklist

- [ ] Tracking and logging have separate approved expectations.
- [ ] The sending route, user seat and message controls are known.
- [ ] Privacy restrictions are preserved during testing.
- [ ] The one-to-one link uses the supported HTML format.
- [ ] Self-activity and recipient activity are distinguished.
- [ ] Anonymous and inferred events are interpreted correctly.
- [ ] Marketing bot filtering and source-tag settings are recorded.
- [ ] Reports use consistent populations and named denominators.
- [ ] Replies and completed actions are reviewed for meaning.
- [ ] Follow-up stops when the customer has answered or opted out.

If the evidence still disagrees, bring one redacted message and its activity/report references to the [$99 martech audit](https://www.martechdevs.com/pricing). A useful repair identifies the broken boundary and leaves the next operator able to explain the result.

## Frequently asked questions

### Can a tracking notification prove that the decision-maker read a proposal?

No. A notification does not establish who personally read the proposal, how much they understood or whether they intend to buy. Multi-recipient mail, forwarding and automated retrieval complicate that interpretation. Use the notification as context for a relevant conversation, then rely on explicit responses, confirmed meetings and agreed next steps when updating the commercial status of the opportunity.

### Should a salesperson mention the exact time an email appeared to be opened?

Usually that detail adds little value and can create an inaccurate impression of surveillance. The timestamp may reflect an image load or inferred engagement rather than the recipient's reading time. Follow up with a useful answer, relevant information or the next agreed action. Let the conversation establish interest instead of presenting a technical tracking event as personal knowledge.

### Can you compare two representatives fairly if one tracks fewer emails?

Not by treating their raw open counts as comparable performance. First compare sending routes, tracking policy, recipient mix and the eligible report population. Keep sent activity, delivery, replies and outcomes separate. Where tracking differs legitimately, evaluate measures both representatives can influence and observe consistently, such as completed follow-ups or qualified meetings, rather than penalizing missing measurement.

### What should an email-tracking incident record avoid collecting?

Avoid copying unrelated message bodies, unnecessary attachments or personal browsing details into a troubleshooting document. Use controlled messages, redacted identifiers and the minimum settings needed to reproduce the issue. Retain the original evidence through the approved access process when necessary. The incident record should explain the sending and measurement path without creating another uncontrolled repository of customer correspondence.

### How should you evaluate an email experiment when opens are unreliable?

Choose an outcome closer to the experiment's purpose, such as a meaningful reply or completed registration, and use a consistent eligible audience and observation window. Retain delivery and tracking diagnostics so missing measurement is visible. Evaluate the treatment against that outcome rather than switching the winning metric after seeing results or assuming a higher pixel-open rate proves a better message.

## Sources

1. [HubSpot one-to-one tracking and logging controls](https://knowledge.hubspot.com/connected-email/turn-off-one-to-one-email-tracking-or-attachment-logging)

2. [HubSpot email tracking mechanics and eligibility](https://knowledge.hubspot.com/connected-email/understand-hubspot-sales-email-open-and-click-tracking)

3. [HubSpot one-to-one sending and reporting](https://knowledge.hubspot.com/one-to-one-email/send-and-reply-to-one-to-one-emails)

4. [HubSpot BCC and forwarding logging](https://knowledge.hubspot.com/connected-email/log-email-in-your-crm-with-the-bcc-or-forwarding-address)

5. [HubSpot contact email logging rules](https://knowledge.hubspot.com/connected-email/manage-contact-email-logging-rules)

6. [HubSpot Never Log behavior](https://knowledge.hubspot.com/connected-email/add-emails-and-domains-to-the-never-log-list)

7. [HubSpot self-open notification handling](https://knowledge.hubspot.com/connected-email/block-self-open-notifications-with-the-hubspot-sales-outlook-add-in-or-chrome-extension)

8. [HubSpot marketing tracking settings](https://knowledge.hubspot.com/marketing-email/manage-marketing-email-open-and-click-tracking)

9. [Apple Mail Privacy Protection](https://www.apple.com/legal/privacy/data/en/mail-privacy-protection/)

10. [HubSpot marketing email bot filtering](https://knowledge.hubspot.com/marketing-email/understand-bot-filtering-in-marketing-email-analytics)

11. [HubSpot marketing email metric definitions](https://knowledge.hubspot.com/marketing-email/analyze-your-marketing-email-campaign-performance)