Real-time alert when a marketing email shows a hard-bounce rate above the safe threshold. HubSpot suspends sends when bounces sustain, so escalate to list cleanup before that happens.
At a glance
Hard Bounce Spike on Email Send is a real-time alert card. It watches each marketing email send in the recent window and fires when the hard-bounce rate for any single send crosses a configurable safe ceiling. Hard bounces are permanent delivery failures (address does not exist, domain dead, mailbox closed), as distinct from soft bounces, which are temporary. A sudden hard-bounce spike is the single fastest way to wreck sender reputation, and HubSpot will throttle or suspend sending from a portal if hard bounces stay elevated. The card exists to put a list-hygiene decision in front of the team within hours, not days.
| What it counts | For each marketing email sent in the trailing 6-hour window, the hard-bounce rate (hard bounces divided by attempted sends for that email). The card fires when any one send crosses the configured threshold. The underlying alert row carries the email name, send time, attempted recipients, hard-bounce count, and computed rate. |
| Hard vs soft bounce | Only hard bounces count toward the spike. Soft bounces (full mailbox, server temporarily unavailable, message too large) are tracked separately and do not trigger this alert, because they often self-resolve on retry. |
| HubSpot Hub scope | Marketing Hub sends only. One-to-one Sales emails and Service emails route through different infrastructure and are not included here. Transactional email add-on sends are excluded unless explicitly wired in. |
| Per-send, not portal-wide | The alert evaluates each send independently. A clean 50k newsletter and a dirty 2k re-import to a stale list are judged separately, so a single bad import does not get diluted by healthy volume. |
| Threshold | A configurable ceiling expressed as a percentage of attempted sends. The shipped default sits in the low single-digit-percent range, which is well above a healthy list’s normal hard-bounce floor. Tune it per portal in the Sensitivity tab. |
| Currency | n/a (count and rate). |
| Time window | RT (real time). The detector re-evaluates as new send-result data arrives, typically within hours of a send completing. |
| Alert trigger | Any email whose hard-bounce rate exceeds the configured ceiling within the trailing 6 hours. |
| Roles | owner, marketing, engineering. Marketing owns list hygiene and suppression; engineering owns any sync that may be writing bad addresses into HubSpot; the owner needs to know because portal suspension stops all marketing email. |
Calculation
Calculated automatically from your HubSpot data. See the At a glance summary above for what the metric tracks and the worked example below for a typical reading.Worked example
A DTC homeware brand on Marketing Hub Professional, roughly 180k contacts, sends a weekly newsletter plus periodic re-engagement blasts. Reading on 17 Mar 26. The team imported a 9,400-row list from an old trade-show CSV and sent a “we miss you” email to it the same afternoon. Two hours later the card fires.| Send | Attempted | Hard bounces | Hard-bounce rate | Alert? |
|---|---|---|---|---|
| Weekly newsletter (clean list) | 162,400 | 980 | 0.6% | No |
| Re-engagement blast (CSV import) | 9,400 | 1,212 | 12.9% | Yes |
| Product-launch teaser | 44,100 | 310 | 0.7% | No |
- The newsletter is healthy. A 0.6% hard-bounce rate on a regularly-mailed list is normal churn (people change jobs, close mailboxes). No action needed.
- The CSV re-engagement blast is the problem. A 12.9% hard-bounce rate is catastrophic. The CSV was two years stale and never validated. Most of those addresses no longer exist.
- Reputation damage compounds. Mailbox providers read a 12.9% spike as a sign the sender is mailing purchased or scraped lists. Even the clean newsletter’s inbox placement can suffer for days afterward because reputation is portal-level, not per-send.
- HubSpot’s own guardrail may already be engaging. Sustained high bounce rates cause HubSpot to flag the portal and, if it continues, restrict sending. The card’s job is to surface this before that restriction lands.
- The fix is suppression, not retry. Hard bounces should be added to the suppression list immediately and the remaining un-sent portion of the import held until the addresses are validated through an email-verification service. Re-sending to a hard-bounced address makes reputation worse.
Sibling cards merchants should reference together
This alert is the acute, per-send view of deliverability. Pair it with these to see the chronic picture and the downstream revenue effect:| Card | Why pair it with Hard Bounce Spike on Email Send |
|---|---|
| Bounce Rate | The trend-level companion. This alert catches a single bad send; Bounce Rate shows whether the portal’s baseline is drifting upward over time. |
| Delivery Rate | The inverse health signal. A bounce spike pulls delivery rate down; confirm the magnitude of the dent here. |
| Email Deliverability Drop | The reputation early-warning. Hard bounces are one of the main inputs to a deliverability drop; this card explains the why behind that one. |
| Spam Complaint Rate | The other reputation killer. Bounces and complaints together are how mailbox providers decide to throttle a sender. Watch both after a bad import. |
| Email Health KPIs | The roll-up panel. Use it to confirm whether one bad send has moved the aggregate health score. |
| List Health Summary | The root-cause view. A bounce spike almost always traces back to a specific unhealthy list; this card finds it. |
| Email Send Volume (trend) | Volume context. A spike on a tiny send is contained; a spike on a portal-wide send is an emergency. |
Reconciling against HubSpot
Where to look in HubSpot: HubSpot reports bounces per email and across the portal, but it does not surface a real-time per-send hard-bounce alert. The closest native views:HubSpot → Marketing → Email then open the specific send and read the Recipients tab, which breaks out Sent, Delivered, Bounced, and lets you filter to hard bounces. HubSpot → Reports → Analytics tools → Marketing email for portal-level bounce trend. HubSpot → Settings → Marketing → Email → Subscriptions / Health where HubSpot warns about sending health and any restrictions.The merchant traditionally finds out about a bounce spike either by opening the email report manually or by receiving a HubSpot health warning after the damage is done. This card moves that discovery to within hours of the send. Why our number may legitimately differ from HubSpot’s own view:
| Reason | Direction | Why |
|---|---|---|
| Bounce-classification timing | Our rate higher early, then settles | Some bounces are reported by mailbox providers minutes after send, others hours later. A reading taken right after send can show a different rate than HubSpot’s eventually-final number once all bounce notifications arrive. |
| Hard vs soft split | Our rate lower | HubSpot’s headline “bounced” figure can include both hard and soft bounces; this card isolates hard bounces only, so our spike rate is usually below HubSpot’s combined bounce percentage. |
| Time zone | Boundary effect | The portal’s reporting time zone may differ from UTC, shifting which send falls into the trailing-6-hour window on edge cases. |
| Per-send vs aggregate | Either | HubSpot’s portal-level report blends all sends; a bad small send can be invisible in the aggregate but loud in this per-send card. |
| Suppression already applied | Our rate lower on resend | If addresses were suppressed after the first bad send, a later send to the corrected list bounces far less; comparing the two reads without noting the suppression looks like an unexplained improvement. |
| Card | Expected relationship | What causes legitimate divergence |
|---|---|---|
| Klaviyo Bounce Rate | If the merchant runs both ESPs, a shared bad import shows a bounce spike on whichever platform sent to it. The platforms do not share suppression by default. | An address suppressed in HubSpot can still be live in Klaviyo (and vice versa), so the same stale list can bounce twice, once per platform, until both suppression lists are synced. |
| Shopify Total Revenue | No direct relationship, but a deliverability collapse caused by a bounce spike eventually shows as softer email-attributed revenue. | The lag between a reputation hit and a revenue dent is days to weeks, so do not expect same-day correlation. |
Known limitations / merchant FAQs
What is the difference between a hard bounce and a soft bounce, and why only alert on hard? A hard bounce is a permanent failure: the address does not exist or the domain is dead. A soft bounce is temporary: the mailbox is full or the server is briefly down. Soft bounces usually clear on the next retry, so alerting on them creates noise. Hard bounces never clear and directly damage sender reputation, so they are the ones worth waking someone up for. Why did the card fire on a send that only went to a few hundred people? The alert is rate-based, not count-based. A 12% hard-bounce rate on 400 recipients is the same reputation signal as 12% on 40,000. Mailbox providers judge senders on rate. The card is doing its job by flagging the small dirty send before that list gets reused at scale. HubSpot suspended our sending. Did this card cause that? No. The card is read-only. It detects the same bounce data HubSpot uses for its own sending-health guardrails. If HubSpot restricted the portal, that is HubSpot acting on sustained bounce rates. The card’s purpose is to give the team a head start so they can clean up before HubSpot intervenes. We re-sent to the same list and the alert did not fire the second time. What happened? Most likely the hard-bounced addresses were added to suppression after the first send, so the second send skipped them. That is the correct outcome. Never re-send to an address that hard-bounced; it should be suppressed permanently. Can we tune the threshold? Yes. The safe ceiling is configurable per portal in the Sensitivity tab. A young portal mailing recently-validated addresses can run a tight threshold; a portal that periodically re-engages older cohorts may set it slightly higher to avoid alerting on expected churn. Set it to your real baseline rather than the generic default. Why does the rate I see here differ from HubSpot’s email report? Two common reasons. First, timing: bounce notifications trickle in over hours, so an early read differs from the eventual settled figure. Second, classification: HubSpot’s headline bounce number can combine hard and soft bounces, while this card isolates hard bounces only. Match the hard-bounce filter in HubSpot’s Recipients tab to compare like for like. Action playbook when this fires:- Open the offending send and confirm the hard-bounce rate and the list it targeted.
- Add all hard-bounced addresses to the suppression list immediately; do not retry them.
- Identify the acquisition source of the bad list (CSV import, form, third-party sync) and pause any process that is writing unvalidated addresses.
- Run the remaining un-mailed portion of that list through an email-verification service before the next send.
- Watch Bounce Rate, Delivery Rate, and Spam Complaint Rate over the following week to confirm reputation recovers.