About this article: This article is reviewed against the cited sources. This is source research, not lab validation. Check the stated versions and your own requirements before making changes.
Direct answer: Publish planned maintenance with a consistent template: what will change, when in the customer’s local time zone, expected user impact, what they should do beforehand, and how you will confirm completion—delivered through agreed channels (email, in-app banner, status page) with lead times matched to severity, not only on the day via a frantic support blast.
Why it matters
ISV customers experience your hosted product like SaaS: they expect predictable Tuesday night patches, advance notice for risky database work, and honest updates if the window overruns. Silence trains users to blame “the cloud” for every glitch; over-mailing trains them to ignore you.
This is recurring maintenance communication for software vendors, not migration cutover day staff messaging (different audience and stakes). Your ops team coordinates with the hosting provider’s maintenance where applicable, but customer-facing words are the ISV’s voice.
How it works
Maintenance classes (notice lead times below are illustrative policy examples, not legal minima):
| Class | Example | Typical notice |
|---|---|---|
| Routine | Security patch, no login impact | 48–72 hours |
| Urgent security / emergency | Critical CVE or active exploitation fix | As soon as practical; post-incident summary if window could not wait for routine lead time (contractual approval rules apply) |
| Degraded | Read-only mode possible | 5+ business days |
| Major | Hard outage, schema change | 2+ weeks + reminder |
Align windows to Australian business hours preferences in contracts—e.g. avoid EOFY weekends unless customer signed high-risk window.
Message template (sections)
- Summary — one sentence impact.
- Schedule — start/end with time zone label (per-tenant default zone).
- User actions — save work, pause integrations, delay batch jobs.
- Scope — regions, modules, tenants affected (franchise parents may forward to sites).
- Updates — where to watch (status page RSS, email reply-to).
- Completion — all-clear notice including validation performed.
Channels
- Email to admin contacts—not every end user—for routine work.
- In-app banner for logged-in users during the window.
- Status page for timestamped history (subscribers include partners).
- Partner pack when resellers must cascade—give them forwardable text.
After action
Send completion or extension within promised SLA. If rollback occurred, say so plainly—customers detect silent rollbacks via mismatched versions.
Relationship to hosting
When a provider maintains hypervisor or network layers, merge notices so customers receive one coherent story. Do not forward raw infrastructure jargon without translation to application impact.
What to choose and why
Maintain a notice library: routine patch, certificate rotation, identity provider cert, major upgrade. Variables: date, duration, ticket id.
Automate scheduling from your change calendar; human approval before send for major class.
Measure open rates on admin emails and status page subscribers—if nobody subscribes, fix discovery in the portal.
Read next
Related articles you may want next.
- HostingWhat partners need to see that end customers should notGive partners the oversight they need while protecting each customer’s data through tenant boundaries, limited permissions and clear support access rules.Read article
- HostingHandling change requests after an ISV customer is liveRoute live customer requests to the right owners, distinguishing incidents, standard changes, defects and enhancements while recognising security and privacy concerns.Read article
- HostingTime zones and financial calendars when Australian customers share a hosted stackDesign time zones, daylight saving and financial calendars for Australian customers so scheduled work and reports use the intended business dates.Read article
Risks and limits
- Wrong time zone in notice erodes trust—generate from tenant metadata.
- Promising zero downtime when shared schema migrations run is dishonest.
- Legal or regulatory customers may need contractual minimum notice—contract beats template.
- Communications guidance not lab-tested.
Sources
- Site Reliability Engineering — Chapter 6: Monitoring distributed systems — Google SRE; accessed 2026-11-16 — observable signals that maintenance completed or overran (monitoring focus; not a customer-comms standard).
Related reading
- Handling change requests after an ISV customer is live
- Time zones and financial calendars when Australian customers share a hosted stack
- What partners need to see that end customers should not
- Software vendor hosting (related service context)
Learning notes
- Maintenance comms are a product ops discipline, not ad-hoc emails.
- Classify maintenance to set notice lead times.
- Always label time zones for Australian customers.
- Completion messages matter as much as advance notice.
- Partners need forwardable text, not forwarded panic.
This article was originally published on .