How to Track Contract Obligations Without Losing Your Mind

Here is how to track contract obligations, in one paragraph, before we go anywhere near a framework.

Read the contract. Pull out every promise, on both sides. Give each one a named human and a date. Decide how often each one needs looking at, and then actually look at it on that rhythm.

That is the whole method. It fits on a beer mat.

It goes wrong because somebody produces a register with 300 lines in it, circulates it with a certain amount of pride, and then discovers that nobody in the building, including them, has the slightest intention of reviewing 300 lines every month. So it gets reviewed once. Then never. Then it becomes an item of furniture.

The hard part of obligation tracking is not extraction. It is designing something that survives contact with a normal working week.

Step one: extract, and accept that it is a reading job

Obligations are not helpfully gathered in a section called "obligations". They are scattered through the operational schedules, the reporting annex, the change control clause, and that one paragraph in the pricing schedule that everyone skimmed because it was three weeks before signature and morale was low.

So somebody has to read the thing. Properly. Every shall, must, will, shall not and is required to, plus the softer ones that look like context and are actually commitments, like "the Supplier will attend the quarterly review".

A 200 page facilities agreement will typically yield somewhere between 80 and 300 discrete obligations depending on how granular you get. If your extraction returns 12, you have found the headings, not the obligations.

If you want a structure rather than a blank spreadsheet, the UK government's NISTA contract management guidance publishes a free obligations matrix toolkit. A perfectly sensible skeleton to steal. No licence fee, no sales call.

Step two: one named human, one date, no departments

An obligation owned by "Procurement" is owned by nobody. An obligation owned by "the FM team" is owned by nobody with extra steps.

Names. Dates. Every row.

This sounds petty until you watch what happens when the row says "Facilities" and the report is late. Three people each assume one of the other two has it. Nobody is lying, nobody is lazy, and the report is still late.

Record the handover too. People move on, and an obligations register whose owners left in 2024 is not a register, it is a memorial. The CCM Institute's benchmarking puts clarity of responsibilities at 3.0, tied for the most severe capability gap in the discipline. That number is not measuring ignorance. It is measuring exactly this.

Step three: split your obligations from theirs

Two lists, or one list with a column, but the split has to exist.

External obligations are the ones the other party owes you. These are comparatively easy, because you have governance meetings, an account manager and a certain amount of leverage. You can chase them.

Internal obligations are the ones your own organisation owes. These are the ones that quietly sink you, because the counterparty is under no obligation to warn you that you are in breach. They will simply note it, keep the file, and mention it in eighteen months when you raise something they would rather not discuss.

Buy side and sell side get this wrong in mirror image. Buyers watch the supplier's obligations forensically and almost never audit their own. Sellers monitor their own delivery metrics obsessively, because they are being beaten with them internally, and pay far less attention to whether the customer is doing the things their delivery actually depends on.

Both half-blind, in opposite directions. It is almost sweet.

If you are sell side, look hardest at the conditional earning mechanisms. Submit the report by the date, hold the review, hit the milestone, get paid. Miss the condition and nobody sends you a letter. The money just does not arrive.

Step four: rate every obligation for attention, not for status

This is the step that makes the difference between a register you use and a register you apologise for.

Do not try to review everything on one cycle. Instead, decide for each obligation how often it needs touching, and record that as a RAG rating.

Red means monthly attention or more often. High impact, high risk, or currently going wrong.

Amber means quarterly. Real consequences if missed, but not moving fast enough to need constant watching.

Green means annually. It matters, or it would not be in the register at all, but a yearly check is proportionate.

The critical bit, and say this out loud whenever you hand the register to anyone: the colour describes the attention required, not the current status. An obligation can be red and perfectly healthy. Red is a diary decision, not an alarm. People get this backwards constantly and then panic about a register that is 20% red by design.

The worst register I ever inherited arrived in a handover pack from a consultancy who had been paid a genuinely impressive amount to produce it. 287 rows. Every shall in the agreement, extracted diligently, colour coded, sorted by clause number. It was a lovely artefact.

It was also 14 months old and nobody had opened it since the day it landed, because opening it was a two hour job with no obvious place to begin.

We rated it for attention and ended up with 31 red lines. Thirty-one is a Tuesday morning. The other 256 obligations did not vanish, they just stopped pretending to need monthly attention, and the register started getting used the following week.

An annual review of the right 40 obligations beats a quarterly review of all 300, every time. I will defend that in any room you like.

Step five: make the review produce a decision

"Touching" an obligation means monitoring it, checking it is on track, and where it is not, going and doing something about it. That last part is not admin. It is the job.

Put the red list in your monthly commercial review, the amber list in the quarterly, and the green list in the annual. Same meetings you already have, assuming you have them, which is its own article.

A review that produces minutes and no decisions is a status update wearing a suit. If nothing changed as a result of the meeting, the meeting was optional.

You cannot do this for every contract, so stop pretending

Here is the constraint nobody puts in the process document. You might have four contracts or four hundred, and nobody is resourcing 400 of these.

The instinct is to respond by doing a thin version everywhere. That is the worst available option, because it costs real time and produces the appearance of control without any of it.

Segment deliberately instead. Value alone is the wrong axis. A high-value commodity purchase might need almost nothing, while a mid-value contract with high uncertainty and a strategically important supplier might need the full machine.

Then be honest with the business about what the rest are getting, which is monitoring rather than management. That is a defensible position. Pretending you manage all 400 is not, and it will be tested at the worst possible moment.

How to track contract obligations, Where this falls over in practice

Three failure modes, in rough order of frequency.

The register goes stale after an amendment. Contracts get varied, extended, novated. If the tracker does not move when the contract moves, it becomes confidently wrong, which is considerably worse than being obviously empty.

The person maintaining the spreadsheet leaves, taking the working knowledge that never made it into the handover pack. This is the single most common cause of death for obligation tracking in organisations under 500 people, and it usually happens somewhere between 30 and 40 contracts.

Nothing chases. A register is a passive document. It sits there being accurate while a notice period closes in another tab.

None of this is a diligence problem. Diligence was never the missing ingredient. What is missing is a mechanism that works when everyone's attention is somewhere else, which is most of the time. The value at stake is not trivial either: WorldCC and Accenture put average value erosion after award at around 8.6%, and only 8% of organisations have invested in integrated contract management capability. Those two numbers are related.

How TermHive helps you keep an obligations register alive

Everything above is doable in a spreadsheet, right up until it is not.

TermHive is built for the point where it stops working. Contracts go into the Library once. Obligations come out as individual records in Track, each with a named owner, a due date and a RAG rating for attention rather than status, so the register you designed in step four is the register the software actually runs. Alerts does the chasing, including the notice periods and renewal dates that a passive spreadsheet will never shout about. ROI reports what the whole thing is recovering, which is what you will need when somebody asks you to justify the licence.

What it does not do: it will not negotiate your contract, it will not sign it, and it is not a drafting platform. There are excellent tools for the pre-award half and TermHive is deliberately not one of them. It also will not replace a competent contract manager, it will just stop that person spending their Tuesdays rebuilding a spreadsheet.

If you want to see what obligation tracking looks like when something else does the chasing, take a look at Track.

Frequently asked questions

How do you track contract obligations? Extract every obligation from the contract, record each one with a named owner and a due date, split internal obligations from external ones, then rate each obligation for how often it needs attention and review it on that cadence. The extraction is the tedious part. The cadence is the part that determines whether the register survives.

How many obligations does a typical contract contain? It depends on length and complexity, but a substantial services or facilities agreement will usually yield somewhere between 80 and 300 discrete obligations once you include both parties, the schedules and the reporting requirements. If your extraction produced a dozen, you have captured the headings rather than the obligations.

Can you track contract obligations in a spreadsheet? Yes, and at low volumes it works fine. It tends to fail somewhere around 30 to 40 contracts, and the usual cause of death is that the person maintaining it leaves. A spreadsheet is also passive: it will hold an accurate notice date and never tell anyone about it.

What is a RAG rating in contract obligation tracking? A simple way of recording how much attention an obligation needs. Red means monthly or more often, amber means quarterly, green means annually. The colour describes the required attention, not the current health of the obligation, so a red obligation can be entirely on track.

Who should own contract obligations? A named individual, never a department. The obligation owner should be the person who will actually perform or evidence the commitment, not the contract manager who recorded it, and the handover should be logged when that person changes role or leaves.