Skip to content
EncryptTechBD
Home Industries Our Work Blog Pricing About Contact Discuss Your Project
বাংলা (BN)
Software Engineering 6 min read

Software maintenance contract scope: problems and fixes

Diagnose scope disputes in software maintenance contracts, then fix clauses for severity, change requests, security patches and handover.

E
EncryptTechBD · 0 comments
Programmer working on code in a dark room
Photo: Negative Space / StockSnap

Most maintenance disputes are not technical. They are arguments about definitions: what counts as a defect, when the clock starts, whether a security patch is included and who owns the documentation. By the time the invoice arrives, the contract has usually been signed for a year and nobody remembers what was meant.

This guide is for the person who inherited the contract and now has to make it work. It follows the symptoms that appear in monthly service reviews and gives a fix for each, plus the clauses to change at renewal.

Read the contract as four schedules

A maintainable agreement has four parts, even if they sit in one document. If any part is missing, that absence is the root cause of most disputes that follow.

  • Scope schedule: the applications, versions, environments, integrations and third-party components covered, with explicit exclusions.
  • Service schedule: severity definitions, response and resolution targets, support hours, channels and reporting.
  • Change schedule: what counts as a change, the threshold for pre-approval, estimation rules and who can authorise work.
  • Exit schedule: documentation, repository access, credentials, data return, deletion, transition assistance and intellectual property.

A diagnostic sequence for a disputed invoice

  1. Take the last quarter of tickets, not the last two arguments.
  2. Map every ticket to a clause in the scope or service schedule.
  3. Separate first response from actual resolution; a fast acknowledgement is not a fix.
  4. List the work that was done outside the contract and who approved it.
  5. Interview the person who signs off vendor work. Approval habits often create the dispute.
  6. Compare invoice categories with ticket categories. If the categories do not match, the reporting is not usable.

The output is a one-page evidence table. It is far more useful in a review than a general complaint that the vendor is slow.

Every fix turns into a change request

Symptom: a broken report, a validation rule or a small label change is quoted as new project work, and the backlog grows while the business waits for approval.

Likely cause: scope is written as “support” without defining a defect, an enhancement or a change. There may be no small-work allowance, and the vendor may treat any code edit as a project to protect its margin.

Fix: define the three categories with examples from your own system. Set a threshold below which work is pre-approved, and a monthly review of consumed hours. Require estimates in hours and a written acceptance test for anything above the threshold.

Prevention: run a scope workshop before renewal and attach the examples to the contract. Ambiguity is not neutral; it always favours the party with the invoice.

The SLA looks healthy while outages drag on

Symptom: the monthly report shows high compliance, but a real outage lasted most of a working day and nobody was accountable.

Likely cause: the clock starts when the vendor acknowledges rather than when the incident is reported or detected. It pauses while the vendor waits for the customer, excludes third-party providers and lets the vendor set severity after the event.

Fix: define severity by business impact with examples. Start the clock at detection or report, whichever is earlier. Write pause rules narrowly, require evidence for a third-party exclusion and publish response and resolution separately. Reported compliance should be reproducible from the ticket log.

Prevention: test one real incident against the clause before signing. If the contract cannot describe last month's outage accurately, it will not describe the next one either.

Security patching sits outside the retainer

Symptom: framework, plugin and server patches are months behind, or each patch becomes a separate quotation.

Likely cause: the contract treats maintenance as break-fix only. Patching is classified as project work, and there is no agreed cadence, maintenance window or emergency path.

Fix: add patching to the scope schedule with a routine cadence, an emergency path for critical vulnerabilities, a maintenance window and a rollback plan. Require evidence: patch version, date, test result and any deferred item with an owner.

Prevention: make patching a line item in the service report. If it is invisible in reporting, it will be invisible in delivery.

The vendor holds all the knowledge

Symptom: small questions take days, onboarding a new developer is impossible and the business is afraid to change vendors.

Likely cause: documentation was never a deliverable. There are no runbooks, no architecture notes, no repository access and no transition clause.

Fix: make documentation a paid milestone with named artefacts: architecture diagram, deployment guide, runbook, dependency list, credentials inventory and known-issues log. Review the runbook with the person who actually operates the system. Get repository and pipeline access before the renewal conversation, not after notice is served.

Prevention: put a quarterly knowledge-transfer session in the service schedule and an exit schedule that survives a change of account manager.

Feature work crowds out maintenance

Symptom: the same team handles emergencies and new features, and routine fixes wait behind the roadmap.

Likely cause: one capacity pool for everything, no published priority rule and no separation between keeping the system healthy and changing it.

Fix: reserve capacity for maintenance, publish the rule that security and availability work pre-empts feature work, and route larger changes through a change board with dates. Report both streams separately so the business can see what it is buying.

Prevention: include a monthly service review with a standing agenda: incidents, patching, backlog, changes, risks and capacity.

Rewrite the clauses at renewal

  • Name the systems, versions and environments in scope, including integrations and third-party components.
  • Define severity with business examples and state when the clock starts and stops.
  • Set response and resolution targets, and explain how both are measured.
  • Add patching cadence, emergency handling and a maintenance window.
  • Define defect, enhancement and change, with a pre-approved small-work threshold.
  • Require a monthly service report and a quarterly review with named attendees.
  • Include documentation, repository access, credential rotation and transition assistance in the exit schedule.
  • State how data is returned and deleted, and how subcontractors are disclosed.
  • Write a price review mechanism rather than an automatic increase.

Metrics to put in the service report

Ask for tickets by severity, first response and resolution times, reopen rate, change requests, patch compliance, unplanned downtime, backlog age and knowledge-transfer completion. Every metric needs a definition and a data source. A number that cannot be reproduced from the ticket system is a marketing claim.

If you need help rewriting the scope or taking over a system that was never documented, start with Custom Software & Integrations. The pricing page explains how assessment, maintenance and handover are scoped. For the budgeting side, read how to build an SME IT budget that survives the year.

Sources

Services that deliver this

Want this implemented properly?

Pricing →

Every engagement starts with a written scope. If a requirement is outside it, we quote it separately before starting.

Responses (0)

No comments yet. Add the first one below.

Leave a response

Your email address will not be published. Required fields are marked *.

Keep reading

More field notes