Open the AMC software market in India and every product page says the same five things: track renewals, schedule preventive visits, assign engineers, log complaints, generate invoices. AntMyERP, ServiceKhata360, amcdesk, ServeWell — the feature lists are almost interchangeable. They are all competent at the part of an annual maintenance contract that is easy to model: a record with a start date, an end date, and a calendar of visits hanging off it.
That model is not wrong. It is just shallow. An AMC is not a date with reminders attached. It is a standing obligation — a year-long promise to keep someone else's equipment running, priced once, against a register of assets that changes, with parts economics that decide whether the contract makes or loses money, and an SLA that quietly defines what "breach" means. The packaged tools track the easy half and leave the half that costs you money in a spreadsheet. This is a field guide to the half they skip.
The contract is the wrapper; the asset register is the thing
A packaged AMC tool starts from the contract: customer, value, dates, type. The asset is an afterthought — a line of text, maybe a model number. But the obligation does not live at the contract level. It lives at the asset level.
A single customer might have one AMC covering eleven split ACs across three floors, two of which were replaced mid-contract, one of which is chronically faulting and eating visits. The contract says "₹X per year." The truth is eleven different maintenance histories, eleven warranty positions, eleven different risk profiles — and your margin is the sum of those, not the headline number.
When the asset register is a real first-class object, you can answer the questions that actually matter: which assets are consuming more service than they earn, which are out of warranty and should be priced differently next year, which customer's book is quietly going underwater. When the register is a text field, you find out at renewal, from a loss. Building that register-first model is the kind of internal operations tooling that does not exist as a clean SKU you can buy — it has to match how your book is actually shaped.
Comprehensive vs non-comprehensive is an economic fork, not a label
Every AMC tool lets you tag a contract as comprehensive or non-comprehensive. Almost none of them treat the two as the different businesses they are.
Under a non-comprehensive (labour-only) AMC, you supply the visits and the customer pays for spares. Your cost is bounded — it is engineer time. Under a comprehensive AMC, you carry parts too, which means every compressor, every PCB, every gas top-up comes straight out of that contract's margin. One contract type has a predictable cost. The other has a cost that can swing from healthy to negative on a single failure.
If the software bills, schedules, and reports both the same way, the comprehensive contracts that are bleeding look identical to the labour-only ones that are printing money — right up until renewal. A system that earns its place tracks parts consumption against each comprehensive contract in real time, so a contract going underwater raises its hand in month four, not at the renewal table. That is a margin-protection mechanism, and it is exactly the sort of thing a packaged tool flattens into a dropdown.
The SLA is the obligation, and most tools don't model it
The renewal date is the obligation everyone tracks because it is the one everyone can see. The obligation that decides whether your customer renews — and whether you are exposed to a penalty — is the SLA.
A real AMC carries response-time commitments: a breakdown call answered within four hours, a critical asset restored within twenty-four, four preventive visits delivered across the year on a defined cadence. Each of those is a clock. A missed preventive visit is not a scheduling inconvenience; depending on the contract, it is a breach you may have to make good on, and it is certainly the first thing the customer's procurement team raises when they are deciding whether to renew.
Packaged tools schedule visits. Fewer track whether the visit happened inside the window the contract promised. Almost none surface "you owe this customer two more preventive visits and have nine weeks left to deliver them" as an operational signal. That signal is the difference between a service business that renews 90% of its book and one that loses a quarter of it to "we never saw the engineer." We built exactly this kind of obligation-aware scheduling for a field-service firm in our Northridge Mechanical engagement — the system's job was not to look busy, it was to make sure the promise got kept.
Renewal is a re-pricing event, not a reminder
Here is the quiet assumption baked into "AMC renewal management": that a renewal is the same contract, one year later, at roughly the same price, and the software's job is to remind you before it lapses.
For a healthy service book, a renewal is a pricing decision. This asset is now out of warranty — the comprehensive cover should cost more. This customer consumed double the visits — re-price or move them to labour-only. This site added four assets mid-year — the contract value is stale. A renewal handled as a reminder produces flat re-prices that erode margin every cycle. A renewal handled as a re-pricing event, informed by the service history the system has been quietly accumulating all year, is where a service business actually makes money.
This is the difference between software that stores history and software that uses it. Pulling a year of visit logs, parts consumption, and SLA performance into a renewal recommendation is a small piece of process automation — but it only works if the underlying data was modelled as an obligation in the first place, not as a calendar.
The statutory layer sits on a different clock
An AMC is a supply of service and generally attracts GST at 18%. That part is simple. The part that is not: an AMC is almost always invoiced upfront for a year of service that has not yet been delivered.
So three clocks run at once. The invoice is raised on day one. The tax is due on that invoice. But the revenue is earned across twelve months as visits are delivered, and the obligation is discharged over the same period. A tool that treats an AMC invoice like a one-off product sale gets the GST right and silently corrupts your view of profitability — booking a full year's revenue on day one, so a contract looks profitable in April even though you will carry its cost until March.
A real AMC ledger separates invoiced from earned, keeps the GST layer correct against CBIC rules, and lets you read contract margin honestly through the year. This is unglamorous accounting plumbing. It is also the difference between knowing your AMC book is healthy and assuming it is.
What a real AMC system actually tracks
Strip away the feature lists and a system that fits an Indian service business tracks five things the packaged tools flatten:
- The asset register — the obligation lives per-asset, not per-contract, so margin and risk are visible where they actually accrue.
- The parts economics — comprehensive and labour-only are different businesses, and consumption is tracked against the contracts that carry it.
- The SLA clocks — visits owed and response windows, surfaced as operational signals, not buried in a calendar.
- The renewal as a decision — priced off real service history, not copied forward.
- The statutory layer — GST-correct, with invoiced and earned kept on separate clocks.
None of that requires ripping out the dispatch board or the accounting software you already run. The right shape is usually a thin layer that reads from what you have and owns the obligation model the packaged tools ignore — which is also why it pairs naturally with ongoing support: the contract book changes as the business does, and the software has to change with it.
To make that concrete: picture a Pune HVAC firm running roughly 400 live contracts, a mix of comprehensive and labour-only, across a few hundred customer sites. The build that matters is not a CRM — it is an asset-register-backed contract layer, visit scheduling with SLA clocks, parts consumption tracked against each comprehensive contract, and renewal recommendations generated off twelve months of service history. That focused engagement is typically INR 6–11 lakh over 6–9 weeks, sitting on top of the dispatch and accounting tools already in place. Add a customer self-service portal and an engineer mobile app that captures the visit on site — photos, checklist, customer signature — and you are at INR 12–18 lakh. The number tracks the obligation you are modelling, not a count of screens.
Where this leaves you
If your contracts are genuinely uniform — one asset type, one template, no penalties, parts billed simply — buy the packaged tool. It is built for that, and it will serve you. But most service firms that have grown past a few hundred contracts have discovered the hard way that their margin lives in the things the dropdown flattens: the multi-asset customer, the comprehensive contract going quietly underwater, the SLA nobody is watching, the renewal handled as a reminder.
That is the gap worth closing. Not with more features — with a model of your obligation that matches the one you actually signed. If that sounds like your book, start a conversation with us and we will map where your AMC book is leaking before we write a line of code.