Support Desk & Calls

How it works

The mechanics of this kit, plainly — what runs and when, where the data lands, and what you look at.

• Queue sweep runs every hour, on the hour and keeps the The Queue dashboard fresh. Every hour it reads the support inbox and the call log, and keeps the queue honest.
• Support digest runs weekdays · 8am. Each weekday morning, one card on where the desk stands.
• Support replies runs weekdays · 9am, in 2 stages: Haven reads and picks (handing over needs reply), then Haven drafts the replies (handing over drafted).
• What is captured lands in the Ticket log dataset, which lives at Data/Ticket Log in your vault — yours to open, search, and edit like any other file.
• What is captured lands in the First responses dataset, which lives at Data/First Responses in your vault — yours to open, search, and edit like any other file.
• What is captured lands in the Ticket messages dataset, which lives at Data/Ticket Messages in your vault — yours to open, search, and edit like any other file.
• What is captured lands in the Saved replies dataset, which lives at Data/Saved Replies in your vault — yours to open, search, and edit like any other file.
• The The Queue dashboard reads from the dataset and shows, at a glance: status · this ticket · open by priority · the queue, oldest first · open vs resolved · past target · gone quiet · calls today · call back · on the line · recent calls · needs you now · asks by account · first reply · median hours, by month · top issues, by category · every ticket · how this board works.

The wiring, in one breath: Ticket capture writes to Ticket log · Queue sweep writes to Ticket log · Queue sweep writes to First responses · Ticket capture writes to First responses · Ticket log shown by The Queue · First responses shown by The Queue · The Queue kept fresh by Queue sweep · Ticket log summarised by Support digest · Ticket log feeds Support replies · Ticket messages feeds Support replies · Reply drafts writes to Ticket messages · Ticket capture given to Haven · Reply drafts given to Haven · Saved reply keeper keeps Saved replies · Saved replies feeds Reply drafts · Saved replies feeds Support replies · What keeps breaking reads Ticket log · Saved reply keeper given to Haven · What keeps breaking given to Haven.

Getting the most from it

A kit earns its place when it changes what you look at and what you say, not when it is installed. This is the rhythm that makes it pay.

• Look at The Queue at the start of the day and once after lunch, not at the inbox. The sweep has already filed what came in and aged it; Open by priority and Past target are the honest signal of where you are behind.
• Read the drafts before you read the tickets. Support replies runs at nine on weekdays and Haven writes into your drafts folder, never sends — so a five-minute read-edit-send is the whole morning pass. Run it by hand when the queue spikes.
• Say what happened in passing — "Alice rang about the login loop", "close the invoice one", "I've answered T-0142 by phone" — and Haven keeps the log in step. A call you take from the board's Call back button counts as the first reply on its own.
• Watch Gone quiet more closely than Past target. A ticket in progress with no word from you for a week is the one where the customer starts chasing; the digest leads with it, and a one-line "still on it" from Support replies is usually all it needs.
• Say where a ticket stands as it moves — "I'm on it", "it's fixed, they should check", "we won't do this, it's out of scope" — and Haven sets the status, and offers to draft the note that tells the customer. When several of you share the desk, "that's Jo's" puts a name on it.
• Read Top issues for what to fix upstream, not for whom to answer next: four login tickets in a month is a reset flow to mend, and the category bar is where that shows first. Ask "what keeps breaking?" for the tickets counted by category and by linked issue, this month against the last, with the ticket numbers beside each count so you can check it; Monday's digest ends with the same count in one line. It counts what people asked: why something broke, and what to do about it, are yours.
• Keep the replies you find yourself writing again. Say "keep what I told Mira", or paste a reply and give it a name, and the Saved reply keeper files it on Saved replies with the kind of ask it answers. From then on Reply drafts starts from it and says which saved reply it used, and one you mark out of date stays on the record and is never drawn on again.
• Once a ticket is resolved, ask Haven to follow up: "ask Farah how it went". It drafts two or three sentences on the same thread asking whether it worked, for you to send. Nothing goes out, and a ticket the customer has already thanked you on gets no follow-up. If you would rather be offered one every time you close a ticket, say so and Haven remembers.
• Correct the sweep's judgement and it sticks: "that address is not support", "billing questions are never urgent", "anything from a paying customer is". Each becomes a rule in Haven's preferences, and the next sweep files by it.
• If your volume is light, retime the sweep from hourly to once a day on the Routines screen — twenty-four short runs a day is the honest cost of an hourly queue, and the morning digest already gives you the picture.

Make it yours

Nothing here is fixed. The pieces are ordinary agents, routines, datasets and dashboards, so anything you know about shaping those applies. Retire whatever earns nothing.

Things people usually change first

Say any of these in chat, or edit them on the kit’s page in the app.

• Which address is the support address, and what should never become a ticket.
• The reply target per priority — what "urgent" means for your desk.
• How often the sweep runs: hourly by default, once a day if your volume is light.
• Whose tickets get a drafted reply, and the sign-off and register those replies use.
• How many days before a ticket in progress counts as gone quiet (a week by default), and how long a closed ticket can be reopened by a reply (30 days).
• The replies you keep, whether Haven offers a follow-up when you resolve a ticket, and how many tickets make a problem count as coming back (three by default).
• Whether your own product files tickets here too. An org admin can give it a key that reaches only the ticket records; each ticket then keeps its conversation in Ticket messages, and Haven's replies wait there as drafts until someone sends them from your product.

Your first run

Installing starts a short setup conversation — not a form. Answer roughly, correct later.

1. Adding the Kit offers the first sweep right away: connect your email on the same screen, start it with one tap, and watch it read back through thirty days of the support inbox — and your call log, if your phone line is connected — and build the queue: one ticket per ask, aged, with a priority and a reply target judged from the words. Prefer not to connect? Tell Haven what is waiting instead, and the log builds by hand.
2. When that first sweep finishes, Haven leaves a short note in your 1Presence inbox saying what it filed — how many tickets, from how many requesters — and invites a few minutes together: to fill what email and the call log could not show (a ticket nobody is working, an ask that came by text or chat, one that was finished long ago), which address is the support address, what counts as urgent for you (the reply targets per priority), what should never be a ticket, and how the drafted replies should read. Haven starts from what 1Presence already knows, in its memory and in files in your vault, and asks only for the gaps.
3. Whatever is still blank waits on the Kit page with a way back to Haven, and Haven offers it once the next time you talk. Tell it a part does not apply to you, or that you have finished setting up, and it stops asking.
4. From then on the sweep runs every hour and stays quiet unless something urgent lands, a ticket passes its target, or one in progress goes quiet; the digest arrives each weekday at eight; the reply pass follows at nine. Tell Haven who your support mail comes from (a contact form, a helpdesk or a group address) and the sweep can run within minutes of that mail and check once a day instead of every hour: up to 24 extra runs a day, Gmail only for now.
5. You correct what it filed, in chat — a priority, a category, a ticket that was never one. Those corrections make the next hour's sweep better than the last.

Watch the first pass, then tell it what you would have done differently — that correction teaches it more than any setup could.

The essentials

Learning, preferences and run costs work the same in every kit — so they are covered once, in the Kits guide.

Living with a Kit — the Kits guide →

← All kits