Tracking & reports
Numbers that survive being questioned.
Open and click rates stopped meaning what they used to mean in 2021, and most reporting has not caught up. A tool that shows you a 62% open rate without telling you how much of that was Apple fetching images on a customer's behalf is not being generous — it is handing you a decision you will make wrongly, with confidence.
Tracking lives on its own host.
Opens and clicks are handled by t.clicksomo.com, and by a per-customer alias if you want your own. That separation is not cosmetic. Tracking domains get blocklisted — sometimes for what somebody else did — and a blocklisted tracking host must never be the same host that serves your application and your published pages.
It also means an open pixel carries no application cookie, because it is not on the application's origin and physically cannot.
A click redirect is signed and validated
Every tracked link carries a signed token, and the redirect target is validated against the link set of the message it belongs to. A redirector that will forward to any URL handed to it is an open redirect, and an open redirect on a domain that sends mail will be found and used, usually within weeks.
Unsubscribe in one click
Every bulk message carries List-Unsubscribe and List-Unsubscribe-Post as RFC 8058 describes, which is what puts the native unsubscribe button in the recipient's own mail client and what the large inbox providers now require. It is honoured immediately, and honoured again at send.
- 1TokenSigned, naming the message and the link. Not an encoded URL.
- 2ValidatedAgainst this message's own link set. Anything else is refused rather than followed.
- 3RecordedA click event on the contact — which is itself a trigger another funnel can start on.
- 4RedirectedTo your link. No cookie set, no application origin involved.
- iTokenless by designNobody is signed in when a pixel loads, so this path is deliberately narrow and reads nothing it does not need.
Separate hostSigned and validated
Bots
Stored raw, reported filtered,
always labelled.
Apple Mail Privacy Protection fetches every image in every message it receives, whether or not anybody looked at it. Outlook's SafeLinks visits every link to check it is safe. Both produce events that are indistinguishable from a person, unless you look at them properly.
Raw is kept
Nothing is discarded. The underlying events stay exactly as they arrived, so a filtering rule that turns out to be wrong can be revised without having destroyed the evidence.
Filtered is shown
The headline figure is the one with machine traffic removed, because that is the figure you would want if somebody explained the choice to you.
The label is not optional
Every report says which number is which and what share was machine. A report that silently mixes them is worse than no report, because you act on it.
The practical consequence is sharpest in A/B splits. If variant A happens to be opened by more Apple Mail users than variant B, an unfiltered comparison picks A — and it picks it decisively, with a satisfying margin, on no information whatsoever. Filtering is not fastidiousness here; it is the difference between a test and a coin toss you believed.
The uncomfortable part: your open rate will look lower here than in a tool that counts everything. It was always this number. You were simply reading a different one.
Sent is not a result.
Sent is a beginning.
Reports
Counted from the events, every time.
Campaign and funnel counters are recomputed from the delivery events rather than incremented in place. A counter incremented on each webhook is a counter that a retried webhook inflates — and provider webhooks retry by design, which makes that a certainty rather than a risk.
Campaign performance
Delivered, opened, clicked, bounced, complained, unsubscribed — raw and filtered — plus the basis mix of the audience it went to. Exportable as a PDF for the client who wants one on paper.
Funnel drop-off
How many entered, how many reached each node, how many hit the goal and how many are still waiting. The shape of the answer to "where are we losing people", node by node.
Why did this person get that?
Every run writes a step for every node it entered: what it decided, on what values, and what it emitted. Three weeks later there is a real answer rather than a reconstruction.
Deliverability by domain
Bounce and complaint rates per sending identity and per sending domain, so a problem is attributable to a mailbox rather than to a vague sense that things got worse.
What we do not do.
There is no cross-site tracking, no third-party pixel, no advertising identifier and no data sold to anybody. Contact behaviour is recorded because a funnel needs it to decide what to do next, and it stays inside the project it belongs to.
This website you are reading sets no cookies of its own (the only one it reads is a name label the application leaves after you sign in — see the privacy page) and loads scripts only from this site and castomo's own servers — Observer, our own analytics, which stores nothing on your device, and the chat assistant — and its fonts are served from this domain. That is not a marketing position; it is simply what the site is.
Coming to dash-b: campaign performance, funnel drop-off, list growth, recent sends and deliverability by domain as tiles you drop onto any board, on the same account, with the viewer's own credentials and no key ever reaching the page.
Tell us what you send today,
and to how many.
We will show you what it looks like here — usually with the funnel drawn, because that is faster than a paragraph about it.
Open ClickSomo Talk to a person