The Intake Illusion: Where Demand Enters Before the Ticket Does
Most service audits begin at the queue. It is visible, countable and easy to report on. But the queue is a consequence, not a cause. By the time a request becomes a ticket, a series of decisions has already been made: which channel the customer used, whether the request was recognised as service demand, who took ownership, how it was classified and whether it was logged at all. Those decisions are the operating model in action.
If a team cannot describe its front door, it cannot defend its service model. A channel list is not a description. The front door includes the routes demand uses, the rules that govern acceptance, the people empowered to triage and the exceptions that bypass the formal process. An intake audit makes that front door visible.
Why intake is a control issue, not a courtesy
Service operating models depend on a credible relationship between demand, capacity and promise. Intake is where that relationship is first tested. When channels multiply and informal requests flourish, demand data becomes partial. Capacity is planned against a queue that does not represent the work. Service level clocks start late. Customers learn that the quickest route is a direct email to an engineer or a call to a senior manager. The formal model continues to report, but it no longer describes reality.
An intake audit is therefore a governance test. It asks whether the organisation can account for demand at the point of entry, whether triage rules are applied consistently and whether the service promise is made before capacity is considered.
Four entrances, not one
Most organisations have more than one front door. A useful audit distinguishes four types.
Declared channels are the official routes: the web form, the contact centre number, the service mailbox, the customer portal or the chat function. They are usually documented, measured and owned.
Adjacent channels are legitimate but less integrated: account managers, supplier portals, reception desks, social media inboxes and field engineers who receive requests on site. They often sit outside service reporting.
Informal routes are the corridor request, the direct email to a known specialist, the message in a team chat or the voicemail left on a personal number. They may be well intentioned, but they create work without a record.
Bypass routes are escalations that enter above the queue: a complaint to a director, a request through a regulator or MP, or a supplier asking a senior contact to expedite work. They are not automatically wrong, but they need a controlled path back into the model.
A two-week intake audit
The aim is not a perfect census. It is a defensible map of where demand enters and where it leaks.
- Assemble the paper trail. Collect the channel inventory, service catalogue, routing rules, telephony menus, mailbox auto-responses, web forms, chat scripts and any local triage guidance.
- Trace a sample. Select 30 to 50 requests across channels and follow each from first contact to logged ticket. Record the time lag, who touched it, how it was classified and whether the customer had to repeat information.
- Shadow the front line. Observe calls, walk-ins, inbox triage and the handling of exceptions. Ask staff what they do when the system does not fit the request.
- Interview adjacent teams. Speak to account management, complaints, field operations, suppliers and reception. They often see demand before the service desk does.
- Run a leak log. For five working days, ask staff to note requests that did not become tickets. Keep it simple: date, source, rough effort and reason for non-logging.
- Test the triage rules. Use realistic scenarios to check eligibility, priority, routing, rejection and appeal. Compare what the rules say with what experienced staff actually do.
The output should be a one-page intake map: entrances, owners, volumes, rules, exceptions and known leaks.
What to measure
Precision matters less than direction. Useful measures include volume by entrance, estimated unlogged work, time from first contact to logged request, first-time-right routing, misroute and rework rates, rejection and redirect volumes, abandoned contacts and the number of requests reopened because intake missed a material fact.
These measures should sit alongside the service catalogue. If a service is promised but has no clear entrance, that is a model gap. If an entrance exists but is not in the catalogue, that is a governance gap.
Reading the map
Channel sprawl is the most common finding. It is rarely solved by closing every route. Customers and staff use multiple channels for good reasons. The question is whether each route has a purpose, an owner and a rule for what happens next.
Look for shadow queues: shared inboxes that are not ticketed, spreadsheets used to track requests, and message threads that substitute for a record. Look for triage drift, where priority is set by whoever is available rather than by published criteria. Look for promise gaps, where a channel advertises a response time that the downstream model cannot meet.
Repairing the front door
Repair starts with clarity, not technology. Establish a channel register with an owner, purpose and service hours for each route. Define acceptance criteria and a small set of priority rules. Decide who can accept, divert, refuse or escalate demand. Publish the rules in plain language and review exceptions monthly.
Make logging easy enough that the informal route loses its advantage. That may mean a single demand record, a common reference number or a simple intake form that routes automatically. It does not necessarily require one tool, but it does require one version of the truth.
Finally, protect the human front door. Vulnerable customers, urgent safety issues and complex cases need routes that do not force people through a rigid form. The operating model should make those routes deliberate, not accidental.
The front door is the operating model in miniature. Audit it before debating capacity, because capacity planned against an incomplete intake is capacity planned against a fiction.
