MILabs
Build

Email categoriser for shared inboxes and team routing

An email categoriser reads incoming messages, assigns each to a category and routes it to the person who handles that work.

Updated 2026-07-16

What a email categoriser is

An email categoriser reads each incoming message, works out what kind of request it is, and places it where the right person will see it. Typical categories include a new enquiry, a client document, an invoice query, a complaint or a message that needs an answer today. The sorting can run on a language model that reads the subject line and body, with rules layered on top so known senders or keywords always land in the same place. For a team lead, the payoff is a shared inbox that stays readable. Messages stop sitting unread because they arrived under a general address, and staff spend less time working out who should handle what. The build is usually modest: a connection to the mailbox, a short list of categories agreed in advance, and a log of where each message went, so mistakes can be reviewed. It suits any team where email carries the same kinds of work over and over, such as intake for a firm or a brokerage. The owner stays in charge of anything unusual.

Email categoriser in plain language

Most businesses receive email faster than anyone can read it. A general address for a firm, agency or broker collects requests that look alike on the surface but need different handling: a document for a client file, a payment question for the accounts team, a complaint that should reach someone senior. Triage is the name for the step that tells those messages apart. Done by hand, it depends on whoever opens the inbox first, so items can sit for a long time under the wrong label.

Automated categorisation treats triage as a classification problem. Each message receives a label from a fixed set, and that label decides where the message goes. Older approaches relied on rules such as a known sender, a phrase in the subject or a domain name. Rules are predictable and easy to audit, but they struggle when wording shifts. Language models read the content and infer intent, so they handle varied phrasing well, though their errors are harder to predict.

Good categorisation depends less on the technology than on the category list. Overlapping categories produce inconsistent results, and categories that nobody owns produce messages nobody picks up. Because each label stands for a piece of work, it should use the terms the team already uses.

What you get first

  • One connection to a shared mailbox
  • An agreed list of categories
  • Classification of each new message
  • Rules for known senders and keywords
  • Routing to a folder or team member
  • A log of where each message was sent

What comes later

  • Draft replies for routine categories that staff review before sending
  • Reporting on message volumes by category over time
  • A review queue for messages the model is unsure about
  • Connection to a ticketing or case management system

Tools it uses

Where it is used

  • Sorting enquiries arriving at a real estate agency's general address
  • Separating invoice queries from client mail at a bookkeeping firm
  • Flagging complaints at an insurance broker for senior review

Who it suits

How MILabs runs it

A build runs through the same four steps as every MILabs engagement. Discover finds the stuck part with you and ends in a scoped plan. Craft builds it with you against an agreed scope. Measure reviews how it is going each month, and Repeat scopes the next step.

Email categoriser in practice

Inside a business that uses this, the work starts with an agreed map of request types. Someone who knows the inbox well writes down each category, the person or folder that owns it, and what a message in that category should look like. That document matters more than the software, since every later decision refers back to it. The build then sits between the mailbox and the team. A workflow tool such as Zapier, Make or n8n can pass each new message to the classifier, receive a label and move the message to the right place, while the mailbox stays the record staff check.

Day to day, the useful signal is in the exceptions. Messages with low confidence, or that match nothing on the list, belong in a holding folder where a named person reviews them rather than guessing a category. When staff correct a misfiled message, those corrections become the raw material for adjusting rules or the model's instructions. A periodic look at the log, comparing decisions with what people actually did, shows where categories have drifted.

For the team lead, the job is to treat categories as a living list. New kinds of request deserve a deliberate category rather than a spot under whichever label fits least badly.

good questions, straight answers

Frequently asked questions

What is an email categoriser?
An email categoriser looks at each arriving message in a shared mailbox, places it in one of a few agreed categories, and sends it to the person best placed to act on it. The label can come from a language model that reads the text, from plain rules about senders and keywords, or from a blend of both. Your staff choose the categories and can check a log of every decision made.
How do I know if an email categoriser is working?
Review a sample of sorted messages on a regular basis and tally how many sit in the wrong category. Set the log of decisions beside what staff actually did with each message. When people keep pulling messages out of one category, the rules or the model instructions need adjusting. A healthy build is one where the team trusts the sorting enough to stop checking every item by hand.
Which industries use an email categoriser?
Firms whose inboxes receive the same types of request over and over tend to benefit most. Typical examples include accounting and bookkeeping firms, legal services, real estate agencies, insurance brokers, IT managed services and recruitment agencies. Categories shift from sector to sector, so a recruitment agency might sort candidate applications while an insurance broker sorts claims questions.
Is an email categoriser right for a very small team?
Yes, as long as the inbox is busy enough to make the setup worthwhile. A team of two or three that gets only a few messages a day may find manual sorting quicker. Where one person handles everything, the main benefit is a record of what arrived and where it went, which is useful in its own right.
What does an email categoriser cost to maintain?
No single figure applies, because cost depends on how many messages pass through the mailbox, which tools are chosen and how often the categories change. Ongoing work usually covers checking the sorting, adjusting rules when new types of request appear, and any usage charges from the tools involved. The easiest first step is to say hello through the enquiry form, and we'll take it from there.
let's make something happen

Custom development: how we can help

Choose how you want to work on this with MILabs. Both options start with an enquiry.

Discovery is free, no obligation from the first chat.

Let us build it for you

We build it and look after it.

Build it yourself, we'll help

You build it with our direction, teaching and a weekly check-in.