Contract Obligation Management: A Working Guide

Contract obligation management is the work of knowing every promise inside a signed agreement, on both sides, making sure each one is performed and evidenced, and deciding what happens on the day one of them is not.

That last part is the discipline. Everything before it is filing.

Tracking tells you what is happening. Managing means somebody acts when the answer is unwelcome, and that somebody is usually you, on a Thursday, with two other things on fire.

Most of what gets written about this subject stops at the halfway line. You will find a great deal of content explaining that obligations should be extracted, logged and assigned, all of which is true and none of which is the hard bit.

What contract obligation management actually is

It sits inside post-award contract management and it is the engine room of it. Performance regimes measure outcomes. Change control handles movement. Obligation management is the running check that both parties are doing the things they said they would do, in the way and at the time they said they would do them.

It is not contract administration. Administration is the record-keeping: the filing, the variations log, the invoice checks. Necessary, and about a third of the job.

It is also not the same as obligation tracking, though the two words get used interchangeably by people selling software. Tracking is the mechanism, the register and the cadence and the chasing. I have written the method out in full in a separate guide to tracking obligations, so this page will not repeat it.

Management is the layer above. It is the judgement about which obligations matter, what you do when one slips, and whether you are prepared to defend how you handled it eighteen months later.

The six kinds of obligation you are actually managing

Most articles give you a flat list of obligation types sorted by subject. Payment, confidentiality, data protection, delivery. Accurate, and no use at all, because it tells you nothing about how each one fails.

Sort them by failure mode instead.

Recurring obligations. The monthly report, the quarterly review, the annual insurance certificate. These fail quietly through attrition. Someone leaves, the report stops, and nobody notices for two quarters because nothing breaks when it is missing.

Dated deliverables. The exit plan within 90 days of signature. The implementation milestone. The transition-out plan. These fail loudly, but only much later, when somebody asks for the thing and discovers it was never produced.

Conditional obligations. The ones where their commitment depends on your performance. They supply the resource, you sign off the change within five working days, and their service level only applies if you did. Lose the map of these and you eventually pick a fight you cannot win.

Notification obligations. Notice of a claim, notice of delay, notice of a change in control, notice of intention not to renew. Short windows, prescribed formats, hard consequences. This is the clause people get wrong most often and it is almost always the one that decides who wins.

Value-creation obligations. Gain-share, benchmarking rights, continuous improvement targets, volume rebates. These fail through dormancy. Nobody wakes up thinking about the gain-share. If you are sell side these are your margin recovery, and if you are buy side they are your credits and rebates, equally unclaimed.

Standby obligations. Business continuity plan, disaster recovery testing, data return on termination. Invisible for years, then load-bearing for one very bad afternoon.

Six failure modes, six different management responses. A register that treats them as one undifferentiated list will manage none of them well.

Nobody else in the building is looking at the whole contract

Ask four people whether a contract is performing and you will get four answers.

Ask the engineers and they will tell you the technical specification is being met beautifully. Ask logistics and everything is arriving on time, in the right quantity, in the right place. Beautiful. Ask finance and the purchase orders are flowing, the invoices are clean, receipting is on time. Also beautiful.

Now change one variable. Goods arriving exactly on schedule, quality poor.

Logistics still reports the contract as performing well, because from where they stand it is. Finance still reports it as performing well, because the invoices are clean. Engineering reports failure. The business hears three good and one bad, concludes things are broadly fine, and quietly decides engineering is being difficult again.

Every one of those functions is looking at a real thing and reporting it accurately. None of them is looking at the contract.

On a large agreement, the kind that wraps engineering, construction, logistics and financial services into one document, nobody else in either organisation holds the whole picture. They cannot. They are accountable for their slice and they see their slice clearly.

That is the actual job, and it is why obligation management cannot be delegated to whoever has spreadsheet time this month.

It runs on a rhythm, not on a register

A register is an artefact. The rhythm is what makes it a system.

Obligations need to sit inside the governance you already have: the red ones in your monthly commercial review, the amber ones in the quarterly, the green ones at the annual. Not because that is elegant, but because a meeting that already exists is the only meeting that will still exist in month 14.

The colour, incidentally, describes how much attention the obligation needs rather than how healthy it is. An obligation can be red and entirely on track. Say that out loud every time you hand the register to somebody new, because people read it backwards and then panic.

The other half of the rhythm is escalation. When a review cannot fix something, everyone in the room needs to know where it goes next and whose name is on it. Escalation that skips the commercial layer and lands as a solicitor's letter is experienced by the other side as an ambush, and the relational damage outlasts the issue by years.

An escalation that disappears into an inbox has not been escalated. It has been forwarded.

What to do when an obligation is missed

Here is the part almost nobody writes about, probably because it does not demo well.

You have three honest options and only three.

Enforce it. Raise it formally, in the forum, in writing, with the clause reference and the consequence. Appropriate when the obligation carries real commercial weight and the relationship can take it.

Vary it. Sometimes the obligation was never achievable as written and everyone has quietly known this since month two. Changing the wording so it matches reality is not an admission of failure. Managing to a target that was never going to be met is the failure, and it is remarkably common.

Let it go, knowingly. Not everything is worth the fight. But do it deliberately, record that you decided it, and understand what you are spending, because rights you consistently do not enforce can become rights you have given away by conduct. Waiver is a slow leak rather than a single event.

What is not on the list is option four, which is the popular one: notice it, say nothing, and keep the evidence for a rainy day. That looks like leverage and behaves like a landmine. World Commerce & Contracting (WorldCC) research with Resolutiion finds roughly one in five contracts hits at least one dispute during performance, and the disputes are rarely about the thing that appears in the letter. They are about two years of unraised issues arriving at once.

Could you evidence a year of oversight?

Ask yourself a blunt question. If a regulator, an auditor or a counterparty's lawyer asked tomorrow for your last 12 months of contract oversight, could you produce it?

Agendas. Minutes. Action logs. Decisions and who made them. Performance data in a form somebody else could interpret without ringing you.

The reason this matters is asymmetric and worth naming. Evidence protects you when you are right, and it disciplines you when you are not. The bad version of that conversation starts when you go looking for the minutes of the quarterly reviews your contract required and find that nobody minuted any of them.

Buy side feels this hardest, because buyers audit supplier obligations forensically and almost never audit their own. WorldCC and Accenture put average value erosion at around 8.6%, and only 8% of organisations have invested in integrated contract management capability. Those two numbers are having a conversation with each other.

How TermHive helps you manage obligations rather than list them

Everything above is doable with a spreadsheet, a diary and considerable stubbornness. I did it that way for years.

TermHive exists because stubbornness does not scale. Signed contracts and schedules go into the Library once, so there is a single version rather than five inboxes and a shared drive nobody trusts. Track turns each agreement into individual obligations with a named owner, a due date and a rating for how often it needs attention, including the conditional ones your own side owes and the value-creation clauses that otherwise sit dormant for the term. Alerts does the chasing, particularly on notice windows and renewal dates, where being slightly late is the same as not doing it. ROI reports what the whole exercise is recovering, for the day somebody asks you to justify the licence.

What it will not do: draft your contract, negotiate it, sign it, or decide whether to enforce, vary or let something go. That judgement is yours and it should be. TermHive handles the mechanics so you have time left over to make it.

If your obligations currently live in a spreadsheet that only chases when you remember to open it, have a look at Alerts.

Frequently asked questions

What is contract obligation management? It is the practice of identifying every commitment in a signed contract, assigning each to a named owner with a deadline, monitoring performance on a defined rhythm, and taking a deliberate decision when an obligation is missed. It covers both parties' commitments, including your own organisation's, and it sits within post-award contract management.

What is the difference between obligation tracking and obligation management? Tracking is the mechanism: the register, the owners, the dates and the reminders. Management is the layer above it, deciding which obligations carry real consequence, how often each needs attention, and what to do when one slips. A register nobody acts on is tracking without management, which is the most common failure in the discipline.

What are the main types of contractual obligation? Sorted usefully, there are six: recurring obligations such as monthly reports, dated deliverables such as an exit plan, conditional obligations where the other side's duty depends on your performance, notification obligations with short windows, value-creation clauses like gain-share and rebates, and standby obligations such as business continuity plans. Each fails in a different way.

What happens if a contract obligation is missed? You have three honest choices: enforce it, vary the obligation so it reflects what is actually achievable, or let it go as a deliberate recorded decision. Ignoring it is not a fourth option, because rights that are consistently not enforced can be lost by conduct, and unraised issues have a habit of arriving together.

Who is responsible for contract obligation management? The contract manager owns the system, but individual obligations should be owned by the person who actually performs or evidences them, never by a department. That allocation should be made at handover and re-confirmed whenever somebody changes role, because a register full of owners who left in 2024 is not a register. It is a memorial.