Helpdesk software for a company that has outgrown the mailbox

Somewhere between fifty and five hundred people, support stops working by memory and starts needing a system — but an enterprise tool is still absurd. This is what to put in place at each stage, in what order, and roughly what it costs.

The stages, and what breaks at each one

Companies do not gradually need a service desk. They hit three fairly sharp walls, and the useful thing is to recognise which one you are standing in front of, because the right response is different at each.

• Under about fifty people: nothing is broken. One or two people handle everything and know every requester by name. A helpdesk here mostly adds ceremony. Write down the five things you explain most often and carry on.
• Fifty to about a hundred and fifty: the first wall. Requests outrun memory, things get dropped during leave, and nobody can answer "how much support load do we have". You need tickets, ownership and a knowledge base. You do not yet need process.
• A hundred and fifty to five hundred: the second wall. Multiple teams answer requests, the same problem recurs without anyone connecting the instances, and somebody asks for a report before a budget cycle. You need queues, SLAs, real reporting and the beginnings of asset tracking.
• Beyond five hundred, or earlier if you are audited: the third wall. Change control, problem management, licence compliance and a defensible answer when an auditor asks who approved a production change. This is where ITIL stops being overhead.

What to do first, and what to resist

The mistake at the first wall is buying for the third. An enterprise suite bought by a hundred-person company is typically 15% used, disliked by the team, and replaced within two years — not because it was bad, but because the configuration effort it demanded was never going to be repaid at that size.

The opposite mistake is cheaper but slower: staying on the mailbox for two years past the point where it stopped working, and accumulating a support reputation that takes longer to fix than the tooling did.

The order that works is boring. Intake, then ownership, then a knowledge base, then measurement, and only then process. Each step should be visibly paying for itself before you take the next.

A realistic budget

For a two-hundred-person company with three or four people answering requests, you are buying three or four agent seats. Everyone else raises requests for free. That is the entire software cost of a working service desk, and it is smaller than most teams expect — usually less than what the same company spends on video conferencing.

Budget more time than money. The month of deciding categories, agreeing targets and writing the first dozen knowledge base articles is the real investment, and it is the part that determines whether the tool is loved or tolerated. Our pricing page carries the current per-seat numbers and the GST treatment.

Growing without a migration

This edition is deliberately the first two walls: tickets, queues, SLAs, knowledge base, portal, chat, reporting and an API, with problem and change management from the Growth plan. When you hit the third wall — the asset estate, licence compliance, joiners and leavers — SupportCentral Enterprise at https://supportcentral.in is the same product with that machinery included, and your workspace moves across (tickets, knowledge base, people, settings, attachments) rather than being re-entered.

Either way nothing about the early setup has to be undone. Categories, articles, SLAs and history all carry forward, which is the honest reason to care which vendor you start with: not the feature list today, but whether the next step is an upgrade or a migration.

The three numbers worth tracking from day one

A growing company does not need a reporting suite. It needs three numbers that a founder or a head of operations can read in ten seconds and act on, and everything else can wait until somebody asks for it.

Volume by category tells you where to write knowledge base articles and where the company is quietly hurting. Median time to first reply is the number your colleagues actually feel, far more than resolution time — people tolerate a slow fix and resent silence. And the count of requests still open past their target tells you whether the desk is coping or drowning, which is the number you take into a hiring conversation.

Deliberately not on that list: agent-level productivity. At your size it measures who picks up the easy tickets, it corrodes a small team faster than almost anything else, and there is no version of it that survives contact with a colleague who spent a day on one genuinely hard problem.

The questions worth asking any vendor at this size

• What is the billable unit — agents, requesters, tickets or devices? Only one of those stays predictable as you hire.
• Can I export everything myself, today, without raising a support request?
• Where is the data stored, and what happens to it after I cancel?
• What does the price do at renewal, and is that in the contract or in a salesperson's email?
• How much configuration is required before the first ticket? If the answer is "a project", you are buying for the wrong wall.

Questions