Unknown block type: SchoolHeader
Unknown block type: SchoolNav

Support & Tickets

Queues, SLAs and history against the client

What actually goes wrong

Seven ways support costs you the client

Rarely the answer. Almost always how long it took and how many times they had to explain.

A client contacts support at the worst moment they have had with your firm: a withdrawal that has not arrived, a position that closed, a document rejected for a reason they do not understand. Whatever the answer turns out to be, the experience of getting it is what they will remember and what they will tell other people about.

Most of what makes that experience bad is not the agent. It is a ticket that arrived in a shared inbox with no context attached, an agent who cannot see the account, and a promise made in March that nobody recorded. Each failure below is a version of the client having to do the firm’s remembering for it.

  1. 01

    Tickets arrived in a shared inbox nobody owned.

    Mail, chat, the portal, the app and WhatsApp all land in one queue with an owner and a clock on each item. Nothing depends on somebody noticing an unread message, and nothing is answered twice by two agents who both thought they had picked it up.

  2. 02

    The agent had no idea who they were talking to.

    A ticket opens against the client record, so the account, the last deposit, the KYC state, the open positions and the previous conversation are already there. The first thirty seconds go to the question rather than to establishing who is asking it.

  3. 03

    Nobody knew a ticket was old until the client complained.

    SLAs run per queue and per priority, with first-response and resolution clocks that pause when the ticket is genuinely waiting on the client. Breaches are visible before they happen, on the agent’s board and the team lead’s at the same moment.

  4. 04

    The same question got three different answers.

    Saved replies and an internal knowledge base keep the standard answers standard, in every language you support. An agent who is guessing at the withdrawal policy will eventually guess differently from a colleague, and the client will notice.

  5. 05

    Escalation meant walking over to somebody’s desk.

    Tickets route to finance, compliance or the dealing desk as a recorded handover with the context attached, and come back to the same agent to close. The client keeps one thread rather than being passed between departments that each restart the conversation.

  6. 06

    What support promised never reached sales.

    Tickets sit on the client record permanently, so the next person to open that client — an agent, a manager, the sales desk — sees what was asked, what was answered and what was promised, including by somebody who has since left.

  7. 07

    Nobody could say what support was actually costing.

    Volume by reason, by country, by platform and by time of day, with resolution times behind each. The reason that produces four hundred tickets a month is usually a process fault somewhere else, and it stays invisible until the tickets are categorised.

What is in it

In, queued, worked, and learned from

The last of the four is the one most helpdesks never get to.

  1. Stage one

    Intake

    However the client chooses to reach you, it becomes the same thing: a ticket against their record with an owner on it. Identification happens automatically where the client is signed in, which removes the most irritating two minutes of any support conversation.

    • Portal, app, email, live chat and WhatsApp
    • Matched to the client record on arrival
    • Category and priority on creation
    • Attachments and screenshots
    • Tickets raised by agents on a call
    • Duplicate thread detection
    • Auto-acknowledgement to the client
  2. Stage two

    Queues and SLAs

    Where a ticket goes and how long it may sit there. The clock is the control: an SLA that nobody can see before it breaches is a reporting category rather than a service level.

    • Routing by category, language and country
    • Priority rules and VIP handling
    • First-response and resolution clocks
    • Pause while waiting on the client
    • Business hours and holiday calendars
    • Breach warnings before the deadline
    • Assignment caps per agent
  3. Stage three

    Working it

    The agent has the client and their whole history on the same screen as the conversation. Escalation moves the work without moving the client, who keeps one thread from beginning to end.

    • Client record beside the conversation
    • Accounts, deposits and KYC state visible
    • Saved replies in every language
    • Internal notes and mentions
    • Escalation to finance, compliance or dealing
    • Linked tickets and merges
    • Full action trail per ticket
  4. Stage four

    Closing and learning

    A closed ticket is evidence. Categorised properly it tells you which part of the business is generating the work, which is the only route from answering questions faster to being asked fewer of them.

    • Structured resolution reasons
    • Ticket history kept on the client record
    • Reopen handling
    • Client satisfaction on close
    • Volume by reason, country and platform
    • Resolution and first-response times
    • Agent and queue performance

Who signs in

One thread, six people on it

The client should never be able to tell how many of them were involved.

  • Support agents. Their own queue, the clock on each item and the client record beside the conversation — so the answer is looked up rather than asked for.
  • Team leads. The queue as a whole: what is breaching, what is unassigned, where the backlog is building and which agent needs the next ticket to go elsewhere.
  • Finance. Escalations about deposits and withdrawals arriving with the payment record attached, rather than as a message asking somebody to check something.
  • Compliance. Escalations on documents, restrictions and complaints, with the trail of what the client was told and by whom.
  • Sales. What their own clients have been asking about and what was promised — which is usually the context that makes the next conversation land.
  • Management. Volume by reason rather than volume alone, because the categories with the largest counts are a map of what is broken elsewhere.

Clients forgive a slow answer. What they remember is having to explain it twice.

Get Your Demo Read the FAQ

Questions about the module

Before you retire the shared inbox

What a support lead wants settled before the queue moves somewhere new.

  • The client portal, the mobile app, email, live chat and WhatsApp, all arriving into the same queue rather than into separate tools that each have their own backlog. Where the client is signed in they are identified automatically; where they are not, matching runs on the address or number they contacted from. The point of consolidating is not tidiness — it is that a client who asks in chat and follows up by mail should not become two tickets answered by two agents who do not know about each other.

  • The conversation, and beside it the client: their trading accounts and balances, the last deposit and its status, their KYC state and what is outstanding, open positions, the introducing partner and every previous ticket. Most support questions in this industry are answerable from exactly that set of facts, which is why a helpdesk that sits outside the CRM produces so much escalation — the agent is not lacking authority, they are lacking the screen the answer is on.

  • Per queue and per priority, with separate clocks for first response and resolution, running against your business hours and holiday calendar rather than wall time. The clock pauses while a ticket is genuinely waiting on the client and restarts when they reply, so agents are not penalised for a client who took the weekend to send a document. Warnings appear before a breach rather than after one, on the agent’s board and the team lead’s at the same moment — which is usually enough for the breach not to happen.

  • It escalates as a recorded handover with the context attached, and it comes back to the original agent to close. Finance sees the payment record with the escalation, compliance sees the document or restriction in question, the dealing desk sees the positions — none of them have to ask what this is about. Crucially the client keeps one thread throughout and deals with one person, rather than being passed between departments who each open the conversation by asking them to explain the problem again.

  • Saved replies and an internal knowledge base, maintained per language, so the standard answer to the standard question is the same one every time it is given. This matters more in a brokerage than in most businesses: answers about withdrawal timing, document requirements or margin policy have regulatory weight, and an agent improvising one under time pressure will eventually improvise differently from a colleague. Where an answer needs to change, it changes in one place and every desk has the new version immediately.

  • More than most desks extract, provided resolutions are categorised rather than free text. Volume by reason, country, platform and time of day, with resolution times behind each, tends to show that a small number of causes produce most of the work — a payment method that fails silently, a document requirement worded ambiguously, a platform quirk on one instrument. Those are process faults elsewhere in the business, and they stay invisible for as long as the tickets they generate are only counted rather than classified.