Ask three people at your company who owns a slow hosted app, and you may get three answers. Support says hosting, but the hosting partner’s dashboard says green. The customer has been transferred twice and still hasn’t heard “I own this.” That’s what unclear ISV hosting support responsibilities look like from the customer’s chair.

Agreeing the map before your first customer goes live takes an afternoon, while repairing trust after the first messy incident takes much longer.

Step 1: Name the First Contact and the Incident Owner

These are two different jobs. The first contact is who the customer reaches, while the incident owner is who drives the problem to a fix and keeps the customer informed, even when several parties are working on it.

In a white label software hosting setup, the ISV is the first contact, and hosting issues reach the host through you. In a setup that shows both brands, however, customers may go to the hosting partner for infrastructure problems. Our guide to white label and shared brand hosting covers how that choice changes your support surface.

Microsoft’s incident management guidance says a plan should define key roles, such as an incident manager, a technical lead, and a communications lead, and that clear ownership reduces confusion. A small ISV may put all three on one person, but write the names down anyway, with a backup for each. We’d rather agree who owns an incident than who’s to blame for it.

Step 2: Map Application, Hosting, Identity, and Device Issues

Most disputes over ISV hosting support responsibilities come from four territories, so decide who leads in each, who assists, and what a customer sees.

APPLICATION

Wrong results, error messages, broken features, data problems inside the software, license errors.

Lead: the ISV. The host assists with logs and access to the environment.

HOSTING

The environment is unreachable or slow for all users, sessions won’t start, backups or restores fail.

Lead: the hosting partner. The ISV relays the customer’s evidence and updates.

IDENTITY AND SIGN IN

Password resets, MFA problems, new accounts, removing a departed employee, access rules.

Decide in writing. The ISV may take the customer’s request while the host makes the change.

CUSTOMER DEVICE AND NETWORK

A customer’s PC, local printer, scanner, signature pad, or office internet line.

Lead: the customer’s own IT. The ISV helps triage, and the host confirms the environment is healthy.

Then name the gray zones. Slow screens, printing, and “I can’t sign in” can each land in more than one territory, so write a triage question for each. Is it one user or several? One workflow or everything? Did it start after a change? Those answers decide which territory owns the first move.

Step 3: Agree Escalation Evidence and Customer Updates

Support escalation fails when the second team starts from zero, so agree what travels with each handoff. Microsoft’s guidance also calls for a defined escalation structure covering how incidents are reported, who is notified, and through which channels, and for diagnostic data to be shared through secure channels.

An escalation note carries

  1. Customer name and the affected users
  2. When it started, with the time zone
  3. What the user was doing, and the exact error text
  4. Whether others are affected, and how many
  5. What has already been tried
  6. The incident owner’s name and the next update time

Customer updates need an owner too. Decide who sends them, on what schedule, and what each one says: current status and the time of the next update. Keep customer data out of shared channels, and give the host the access it needs and no more.

Step 4: Separate Response Targets From Resolution Promises

A response target is how fast someone acknowledges the ticket and starts work, while a resolution promise is how long the fix takes, which depends on the cause. Hosting faults and application defects need different kinds of work, so the two can’t share a timeline. Microsoft’s guidance also says a plan should document what each party covers, the exclusions, the escalation procedures, and the expected response times.

Promise what you control. A hosted edition support commitment can say how fast you respond and on what schedule you update, but it shouldn’t say when a code defect will be fixed before anyone has found it. Hosting availability is a different promise: an uptime commitment, like a 99.9% target, describes availability while saying nothing about how long one ticket takes to close.

Party Can commit to Shouldn’t promise
You (the ISV) Response time to the customer, update cadence, severity ratings A fix date for a defect you haven’t diagnosed
Hosting partner Response time to the ISV, infrastructure recovery targets agreed in writing Fixing application bugs
Customer Providing access and information fast Fixing the hosting or the software

Response targets are easy to promise, but resolution promises are where vendors get hurt. So ask any hosting partner for its numbers in writing, and write your own.

Step 5: Test the Handoff With a Sample Incident

Microsoft’s guidance says to test incident processes with realistic simulations, so communication and escalation plans are proven before they’re needed. Run one with a made up incident before your first customer goes live, and not with a real customer. Here’s a sample, shown in lanes.

Time Customer You (first contact and owner) Hosting partner
2:10 PM Calls: screens are slow office wide Names itself owner and asks the triage questions: how many users, which screens, since when
2:20 PM Several users, one workflow, since 1:45. App logs show nothing obvious, so it sends the escalation note Receives the note and checks environment health and session load
2:35 PM Gets an update: investigating, next update at 3:00 Sends the update Reports the environment healthy and shares the load data
2:50 PM Finds a report running a long query inside the app, and classes it as an application issue Confirms no environment changes and stands by
3:00 PM Gets the update: cause found, fix in progress Sends it on time
Next day Both teams hold a blameless review and update the map and the escalation note

The incident above is illustrative, not a real customer.

The test passes when the right person was reached on time, the escalation note arrived complete, updates went out when promised, and one owner was named throughout. Otherwise, fix the map and run it again.

Where CloudTop SaaS Fits

CloudTop SaaS is the engine room behind a hosted edition, not the storefront. In a white label setup or one that shows both brands, our engineers based in the U.S. run the hosting, and the people who provisioned your environment answer the tickets. This article doesn’t state our response targets or severity definitions, so ask us for them in writing before you sign, and ask each provider the same. Our guides to ISV recurring revenue hosting and white label cloud hosting pricing cover the business side of the same decision.

CloudTop SaaS is a CloudTop Office company, hosting Windows business software since 2000 through the parent brand. For the full incident planning framework behind Step 1 and Step 5, see Microsoft’s incident management guidance.

FAQ

Who should be the first contact for a hosted edition?

In a white label setup, the ISV, with hosting issues relayed to the host, while in a setup that shows both brands, customers may go to the host for infrastructure problems. Either way, write it down and name one incident owner for each ticket.

How do I tell an application defect from a hosting fault?

Check the scope. When one workflow fails for several users while the environment is healthy, suspect the application. Everything slow or unreachable for all users points to hosting, and trouble for one person points to their account or device.

What is the difference between a response target and a resolution promise?

A response target is how fast someone acknowledges the ticket and starts work, while a resolution promise is how long the fix takes, which depends on the cause. So commit to response targets and update cadence, and be careful about promising fix times you don’t control.

What should a support escalation note include?

The customer and affected users, when it started with the time zone, what the user was doing and the exact error text, whether others are affected, what has been tried, and the incident owner with the next update time.

How do I test ISV hosting support responsibilities before launch?

Run a made up incident through the whole chain: first contact, triage, escalation note, host check, customer updates, and a blameless review. Judge it on whether the right person was reached on time and one owner stayed named, then fix the gaps and run it again.

Planning a hosted edition and want the support handoff mapped?

A call covers who answers first, how issues escalate, and what you’d want in writing before launch.

Discuss hosted edition support