Appendix D — Suggested Support Workflow

What this appendix is

A suggested support workflow you can adopt or adapt. You're going to get tickets from your tenants — having a process means you handle them faster, look more professional, and burn out less.

Tier your support

Three tiers serve most partners well:

Tier 1 — Front-line

• Answers basic questions ("how do I add an extension?")
• Resets passwords and PINs
• Walks customers through self-service tasks
• Escalates anything they can't solve in 15 minutes

Skill required: comfortable in the dashboard. Doesn't need to know SIP internals.

Tier 2 — Technical

• Diagnoses call quality issues (codec mismatch, RTP packet loss)
• Investigates one-off failed calls (CDR analysis, SIP trace review)
• Configures complex call flows for tenants
• Escalates platform-level bugs to us

Skill required: understands SIP, can read a CDR, can use the call recording / transcription features for diagnosis.

Tier 3 — Engineering / vendor

• That's us. You file a ticket; we triage and respond.

Most small partners run with one or two people doing Tier 1+2. Don't worry about formal tiers — just understand which type of issue you're handling.

Your support intake channels

Pick one or two; don't spread yourself thin:

Channel	Pros	Cons
Email	Async, reliable, has a paper trail	Slow, hard to track at volume
Helpdesk software (Zendesk, Help Scout, Freshdesk)	Tracking, SLAs, reporting	Subscription cost, learning curve
In-platform tickets	Customers raise tickets from inside their dashboard	Requires you to build the UI/integration
Phone	High customer satisfaction	Doesn't scale, no record by default
Chat (Intercom, etc.)	Fast, modern feel	Always on, hard to staff after hours

Recommendation for new partners: start with email + a public knowledge base. Move to a helpdesk system once volume justifies it (~50+ tickets/month).

Response time targets

Industry-standard SLAs for B2B phone service:

Severity	Definition	Response	Resolution
Critical	Service down, multiple users affected	< 15 min	< 4 hours
High	Service degraded, single user affected	< 1 hour	< 24 hours
Normal	Bug or question, no service impact	< 4 hours	< 3 business days
Low	Feature request, minor cosmetic issue	< 1 business day	best effort

Set these in writing, communicate them to tenants, and measure yourself against them. Customers will forgive an outage; they won't forgive being ignored.

Self-service first

The cheapest support ticket is the one your customer never files. Invest in:

• A knowledge base — articles for the top 20 questions. Search-friendly. SEO-friendly so your customers find them on Google before they email you.
• In-product hints — tooltips, helper text on form fields, links from common error messages to relevant KB articles.
• Status page — when there's an outage, customers check the status page first (if you've set the expectation that you have one). Saves you 100 individual "is your platform down?" emails.
• This manual! — your tenants already get a white-label version. Link to it from your welcome email. Update it when features change.

Goal: 80% of customer questions answered without a ticket. The other 20% are real issues you should focus on.

Common issues and the playbooks

"I can't make outbound calls"

Diagnostic order:

1. Are they registered? Check the Extensions list in their tenant; status should be online.
2. Are they trying to dial a valid number? Some carriers block calls to certain country codes.
3. Are they over their usage cap? Check their limits.
4. Is their wallet exhausted? Check billing status.
5. If all the above are fine, check the CDR. If a call attempt was logged with hangup cause NO_ROUTE or similar, that's a routing issue — escalate to us.

"My voicemail isn't being delivered to email"

Check:

1. Voicemail box has emails configured (Voicemail page → Edit box)
2. "Email Attach" is on if they want the audio file
3. Their email isn't blocking platform email (check spam folder; check for SPF/DKIM issues if their IT is strict)

"Caller ID name is showing as the number, not our company name"

CNAM hasn't propagated yet. Caching in the national CNAM database takes 24-48 hours. Set the CNAM, wait, ask them to test again.

If still wrong after 48 hours, check the DID's CNAM Sync Status (DID Inventory). If "Failed", the CNAM may have been rejected (too long, special characters, profanity). Try a shorter version.

"Phone won't register"

Check:

1. SIP credentials are correct (extension number, password, server)
2. Phone is on a network that allows outbound SIP (port 5060 UDP is common; some firewalls block it)
3. Phone firmware is up to date
4. Try a softphone with the same credentials — if softphone works, phone-side issue
5. If softphone also fails, escalate to us with the SIP credentials and the calling IP

"I want to port a number to your service"

Send them the LOA template (see Appendix B). Once you have signed LOA + carrier bill + account info, file the port from your dashboard. Set expectations: 2-4 weeks total.

"I want to port my number away"

Don't make this hard. Help them. They'll come back later.

Tracking metrics

The bare minimum to track:

• Tickets per week
• Average first-response time
• Average resolution time
• Customer-satisfaction score (post-ticket survey: "On a scale of 1-5, how satisfied with this resolution?")

Most helpdesk software does this automatically. If you're using email, do it manually.

If you're seeing the same issue from multiple tenants, that's a documentation problem, a UX problem, or a bug. Address the root cause; don't just answer the same ticket 20 times.

After-hours support

Most B2B phone customers expect 24/7 support. You don't have to actually staff 24/7 — you have to appear to be reachable.

Options:

• Auto-responder + escalation — autoresponder explains business hours, gives an emergency-only number for critical outages
• On-call rotation — you and a partner alternate weeks
• Outsourced after-hours — services like Smith.ai handle off-hours triage

Whatever you pick, document it. Tell customers what to expect at 2am Sunday so they're not disappointed.