Exploiter
Délivrabilité
Comprenez les événements du fournisseur, la réputation des identités d’envoi, la montée en charge quotidienne et la récupération automatique.
Événements
SendHQ's sending infrastructure reports delivery, bounce, complaint, reject, open, and click events. Provider message IDs connect notifications to the original SendHQ email without crossing workspace boundaries.
Isolation de la réputation
SendHQ calculates its circuit breaker for each normalized From email address inside a workspace. A failing bulk identity therefore does not locally pause a clean transactional identity. Suppressions remain workspace-local. The workspace's sending tenant is still shared by its identities, so a tenant-level enforcement action can affect more than one identity.
Seuils
Sur une fenêtre glissante de sept jours, au moins deux bounces sur cinq envois avec un taux de bounces de 10 % ou plus entraînent le ralentissement d’une identité ; au moins trois sur dix envois la mettent en pause. Une plainte à un taux de 0,05 % ou plus entraîne un ralentissement ; deux à 0,1 % ou plus, une mise en pause.
Envoi en pause et récupération
Une identité From en pause reçoit 423 avant la soumission au fournisseur. Arrêtez le flux en cause, inspectez les événements et les destinataires, et corrigez-en la source. Il n’existe pas d’interrupteur manuel pour lever la pause : l’identité redevient active automatiquement lors de sa tentative d’envoi suivante, une fois que les métriques glissantes sont repassées sous tous les seuils.
Every pause, for one From identity or the whole workspace, is refused with 423 and its reason, and the dashboard shows the same reason. If you think a pause is a mistake, write to support through the contact page. We reply within 12 hours and will unblock your account once you explain what happened.
Inspecter l’état
GET /deliverability/reputation liste le dernier état évalué pour chaque identité d’envoi exacte. Une identité apparaît après sa première tentative d’envoi, et sa ligne est mise à jour lorsqu’une tentative ultérieure réévalue la fenêtre glissante.
Métriques
GET /deliverability/stats aggregates delivery events across the workspace for 30 days; it is not the seven-day per-identity circuit-breaker calculation. GET /deliverability/account adds the daily ladder, today's usage, the 7-day bounce and complaint rates against the promote, hold, and pause thresholds, and a daily trend. GET /deliverability/domains and GET /deliverability/senders break the same outcomes down by sending domain and From address. GET /emails/:id/delivery is one message's timeline, including the bounce type and SMTP diagnostic. GET /threads/:id/delivery counts delivered, bounced, complained, and failed messages in a conversation. Diagnostic text is plain text, not HTML.
Score an email's content
POST /api/v1/deliverability/score takes a subject, the html and text as you wrote them, and from, and returns a score from 0 to 100 with what to change. Nothing is stored and nothing is sent. The email editor for automations shows the same number for the same content, a second after you stop typing.
A plain email with no links and no images scores 100. Each thing below takes points off, up to the limit shown, and the score never goes below 0.
- A designed layout: columns, or a coloured or pictured background
- 5
- Links
- 2 for the first, 1 for each more, up to 10
- Links to another site than the sender's
- 4 a site, up to 12
- A public link shortener
- 8 each, up to 16
- An
http://link or image - 2 each, up to 6
- A link to a number address
- 8
- Link text that shows a different site than the link opens
- 15
- Images
- 2 for the first, 1 for each more, up to 8
- Images beside fewer than 400, 1,200 or 2,400 bytes of words
- 12, 8 or 4
- Images and no words at all
- 20
- More than about 200 characters of hidden text
- 15
- Text the same colour as its background
- 8
- A plain-text part that says something else than the HTML
- 5
- A subject in capitals; a body more than half, or more than three quarters, in capitals
- 5; 5 or 10
- An empty subject; a subject that starts with
Re:orFwd: - 10 each
- More than 80 KB of HTML
- 3
- A form, a script, an embedded page, embedded media, a linked style sheet
- 5 a kind, up to 15
Words inside links, and text too faint to read, do not count as words beside an image. deductions lists what was found, most points first, each with a one-line suggestion. warnings carry no points: today there is one, for HTML over 102 KB, which Gmail cuts off.
Only what you can change is scored. The unsubscribe footer and postal address SendHQ adds are not, the opens and clicks it tracks are not, and neither is the frame around an email made in the block editor. Styles are read where they are written on an element, not from a style block in the head.
{
"from": "hello@example.com",
"subject": "Your cart is still here",
"html": "<p>Pick up where you left off.</p><p><a href=\"https://bit.ly/3xyz\">Finish checkout</a></p>"
}{
"score": 90,
"deductions": [
{
"id": "link_shortener",
"points": 8,
"count": 1,
"suggestion": "Remove the link shortener"
},
{
"id": "links",
"points": 2,
"count": 1,
"suggestion": "Remove the link"
}
],
"warnings": []
}The score looks only at the content of one email. Whether mail reaches the inbox depends mostly on things it cannot see: your sender reputation, authentication (SPF, DKIM, DMARC), complaint rate, and how recipients engage with your mail. A high score does not guarantee inbox placement and a low score does not mean spam. Use it as a checklist.