Yes, mostly. A click takes a decision that an open never needed: somebody has to see a link and choose it. That is why clicks became the signal people trusted once opens got noisy. The catch is that security software clicks too, often before anyone reads the message, and the share of clicks it produces has climbed sharply this year. So the useful answer for a Gmail sender is a little narrower. A click is better evidence than an open, and a click with one other piece of evidence beside it is close to an answer.
Why clicks earned their reputation
An open is an image request. Whether it fires depends on the reader's software, so opens go missing when images are off and appear from nowhere when something loads images on the reader's behalf. Omnivery, an email infrastructure company that also sells bot detection, published first party figures in August 2026 drawn from billions of events across 126 mailbox provider families. In its consumer data, roughly half of all opens have been machine generated since 2023. Clicks held out for much longer: through 2023, only about 2 percent of consumer clicks were bots. That gap is where the reputation comes from, and it describes the past better than the present.
Who clicks before your reader does
Corporate mail security is built to visit links. Microsoft's Safe Links documentation says that with protection on, URLs are scanned before the message is delivered, links without a valid reputation are detonated in the background, and rewritten links are checked again at the moment the user clicks. The recommended policy holds the message until that scan finishes. Each of those visits reaches your tracking link looking exactly like a click.
Mailchimp's guide to bot clicks notes that scanners click every link, including the logo and the unsubscribe button. Klaviyo's help centre adds mobile carriers to the list and says plainly that there is no way to prevent bot clicks by inbox providers entirely.
How big the share has become
Omnivery breaks its numbers out by where the mail was read, which is the detail most write-ups skip:
- Consumer mail overall: about 2 percent of clicks were bots through 2023, rising to roughly 16 percent by 2026.
- Outlook and Hotmail: around 30 percent, which Omnivery attributes to Microsoft's link scanning.
- Gmail: about 6 percent through 2025 and early 2026, then a jump to about 22 percent in late February 2026 that has held every week since. Omnivery reads the clean step as a change on Gmail's side, most likely more aggressive link prefetching or scanning. That is its inference from the data.
- Corporate inboxes behind security gateways such as Proofpoint, Mimecast and Barracuda: 80 to 95 percent of clicks are automated, and have been since 2020. That is where many of the buyers a salesperson most wants to reach read their mail.
Mailchimp's own testing points the same way. It finds roughly one click in five is a bot with basic detection, and close to half once third party intelligence is added. HubSpot filters suspected bot activity by default, using IP blocklists and how quickly interactions happen, and warns its numbers will usually look lower than platforms that count everything.
Reading one click on one email
When you sent one email to one buyer, you are reading a single event. Four questions settle most cases.
- When did it happen? A click within seconds of sending, before anyone could plausibly have read the subject line, is a scanner until something says otherwise.
- How many links? A person clicks one or two. Every link at once, signature links and unsubscribe included, is a machine checking them.
- Where did it land? A work address at a large company is the most scanned inbox there is, so a lone click there is weaker evidence than the same click from a personal address.
- What came next? A scanner visits and leaves. A person usually does something else: replies, books the meeting, or opens the email again later on a phone.
So link to places where the next step is visible, like a booking page. If the click is your only evidence, follow up as you would after an unconfirmed open: briefly, with one easy reason to reply.
What Fast Mail Tracker counts, and what still gets through
Every tracked link goes through a redirect that records the visit and forwards it to the real page, so nothing breaks for the reader. A click is then kept out of your counts, your alerts and the signal ramp if its user agent names a known scanner or security product such as Proofpoint, Mimecast, Barracuda, Symantec or Trend Micro, if it carries no browser or operating system at all, or if it arrives within three seconds of the send. Those hits stay on the record with a label rather than being deleted. Clicks that pass are tied to the exact link, and on a group send each recipient gets their own links, so a click names the person.
The limit is the one every tracker shares. Omnivery's report points out that a user agent is trivially forged. A scanner that presents a full browser identity, waits more than a few seconds and visits one link will be counted as a click by us, as it would be by most tools. That is why the ramp ranks a reply above a click, and why the four questions above matter more than any single counter.
A scanner visits and leaves. A person does something next.
We learned the cost of scanner clicks on our own unsubscribe link. Campaign sends carry one, and its first version acted on a plain visit, so a scanner checking links could have unsubscribed someone who never saw the email. RFC 8058, the standard behind one click unsubscribe, describes the same problem for the unsubscribe link in a message's headers: anti-spam software often fetches it automatically, and a sender has no mechanical way to tell that request from a person's. The link now opens a confirmation page, and only the button on that page, or a mail client's own one click request, records the opt out.
Fast Mail Tracker is in early access while Google's security review runs, so places are limited. If you want clicks counted with that kind of care on your own Gmail sends, use the install button at the end of this article.
Frequently asked questions
Is click tracking more accurate than open tracking?
Yes, but less so than it used to be. An open is an image request that can be blocked or triggered by software, while a click usually needs a person to choose a link. Omnivery's 2026 report, based on billions of email events, found roughly half of consumer opens have been machine generated since 2023, while only about 2 percent of consumer clicks were bots through 2023. That click figure rose to roughly 16 percent by 2026, and to 80 to 95 percent at corporate inboxes behind security gateways, so a click is stronger evidence than an open without being proof on its own.
Why do I see clicks a few seconds after sending an email?
Almost always a security scanner. Microsoft's Safe Links documentation says URLs are scanned before a message is delivered, and its recommended setting holds the message until that scan completes, so the scan's visit reaches your tracking link before any person sees the email. Gateways from vendors such as Proofpoint, Mimecast and Barracuda do the same. A click that arrives within seconds of sending, especially on every link at once, should be treated as automated.
How can I tell if a click on my email came from a bot?
Check four things. Timing: a click within seconds of sending is suspect. Breadth: people click one or two links, scanners click all of them, including unsubscribe links. Destination: work addresses at large companies are scanned most heavily. Follow through: a person usually does something next, such as replying, booking a meeting or opening the email again later, while a scanner visits and leaves. A click with one of those follow up signals is strong evidence; a lone click is a lead.
Can a security scanner unsubscribe someone from my emails?
It can if the unsubscribe link acts on a plain visit. RFC 8058, the standard for one click unsubscribe, notes that anti-spam software often fetches the links in a message's header fields automatically, and that a sender has no mechanical way to tell that request from a person's. The safe pattern is a link that opens a confirmation page, with the opt out recorded only when the person presses a button there or their mail client sends the one click request the standard defines.
Sources
- Omnivery, The State of Email Bots 2026
- Microsoft Learn, Safe Links in Microsoft Defender for Office 365
- Microsoft Learn, Set up Safe Links policies in Microsoft Defender for Office 365
- Mailchimp, Unmasking Bot Clicks in Email Marketing Analytics
- Klaviyo Help Center, Understanding bot clicks
- HubSpot Knowledge Base, Understand bot filtering in marketing email analytics
- RFC 8058, Signaling One-Click Functionality for List Email Headers (Levine and Herkula, 2017)