A COO called me in March because her board had asked a question nobody could answer: how much revenue is currently sitting in contracts that auto-renew in the next ninety days?
Forty-eight hours of digging produced a spreadsheet with 214 rows, 31 of which had a blank renewal date. Two of the missing ones had already renewed. One of those had renewed at a price the company had stopped offering eighteen months earlier, locked in for another three years. Nobody had done anything wrong. The contract lived in a Dropbox folder called "Signed 2023 Q4", the AE who closed it had left, and the renewal date existed only inside the PDF.
Her conclusion was the one most people reach: we need a CLM. Her budget owner's conclusion was that $40,000 a year for contract software at a 60-person company was insane. They were both partly right, and the argument they were having was the wrong one.
Here is what I told her, and what I tell most teams in that position. You probably do not have a contract management problem. You have a contract data problem. Those cost very different amounts to fix, and the second one is where almost all of the money is.
What CLM actually means, minus the vendor framing
Contract lifecycle management covers everything that happens to an agreement from the moment somebody asks for one to the moment it expires or renews. Request, draft, redline, approve, sign, store, track obligations, renew. A CLM platform is software that holds all of those stages in one place with an audit trail.
That is the honest definition. The market is worth roughly $3.6 billion in 2026 and growing at double digits, according to Precedence Research, which tells you how much money is chasing this problem. It also tells you how much marketing budget exists to convince you that you need the full platform.
Most B2B companies under 150 people run a partial version of this already, whether they call it CLM or not. A template in Google Docs. A Slack channel where legal gets pinged. DocuSign or PandaDoc for signature. A folder for the executed PDFs. That is a contract lifecycle. It is just not managed.
The gap is not the tooling for stages one through five. Drafting and signing are mostly solved. The gap is stage six onward, where the signed document becomes the only record of a dozen commercially important facts and then goes quiet.
Average share of annual revenue lost to poor contract management, per World Commerce and Contracting research. Top performers hold value erosion under 3%. The worst run past 20%. Almost none of it shows up as a line item anyone reviews.
I do not love that 9.2% figure, and I want to be upfront about why. It comes from a body of research heavily weighted toward large enterprise and complex procurement, and the vendors quoting it have an obvious interest in the number being big. For a 60-person SaaS company selling annual subscriptions on a standard MSA, the real leak is smaller. But it is not zero, and I have never audited a company where it was zero.
Where the money actually goes
When I audit contract leakage for a B2B company, the loss shows up in five places. They are not equally sized, and that matters when you decide what to buy.
Missed renewal windows. This is the big one. Research summarised by contract tooling vendors suggests around 15% of auto-renewing agreements miss their notice window, and roughly 48% of B2B contracts carry an auto-renewal clause. Do that arithmetic and about one in fourteen active contracts renews with nobody looking at it. Sometimes that renewal is in your favour. Often it renews a legacy customer at a legacy price for another twelve or thirty-six months, and you have just written off the price increase you spent a quarter planning.
Unbilled entitlements. The contract says 50 seats. Finance bills 50 seats. The customer has been using 78 for a year. Nobody reconciles the two because the seat count lives in a PDF and the billing record lives in Stripe.
Concessions nobody logged. An AE agreed to waive the first year's implementation fee "as a one-time thing" in an email, then it ended up in the signed order form, then it repeated at renewal because the CSM read the previous contract and assumed it was standard.
Price protection clauses. Somebody negotiated a cap on annual uplift into an MSA in 2023. Three years later your pricing team announces a 12% increase and legal discovers a third of the base is contractually capped at CPI.
Obligations you forgot you signed. Uptime SLAs with service credits. Data residency commitments. Notice requirements for changing terms. Each one is fine on its own. Forty of them with no register is a slow-moving liability.
Notice what all five have in common. None of them are drafting problems. None of them are signature problems. Every one is a problem of a fact being trapped inside a document instead of living in a system where somebody can act on it. This is the same failure mode I wrote about in the revenue leakage piece, just with contracts as the container.
A signed contract is a data source, and most companies treat it as a filing cabinet.
Six to ten fields per agreement drive nearly all the commercial risk. Getting those fields into your CRM is a weekend of work. Buying a platform to store PDFs you already have is a year of procurement.
The six fields that do most of the work
Before you evaluate a single vendor, run this exercise. Take your last 40 signed contracts. Open each one. Pull these fields into a spreadsheet:
- Effective date and initial term length
- Renewal type: auto-renew, evergreen, or expires
- Notice period required to cancel or renegotiate, and the resulting notice deadline
- Contracted quantity or usage limit, whatever your value metric is
- Price and any uplift cap or price protection language
- Non-standard terms: custom SLA, custom DPA, unusual liability cap, termination for convenience
That is it. Add a seventh if you sell into regulated buyers: any clause that triggers a compliance obligation on your side.
Two things happen when teams do this. First, it takes far longer than they estimated, usually four to six hours for 40 contracts, which is itself the finding. Second, they discover somewhere between three and nine agreements with terms nobody on the current team knew about.
The 60-person company I mentioned found eleven in 214 contracts. One of them was a most-favoured-nation clause promising a customer the best price offered to any comparable account. They had been running quarterly discount campaigns for two years.
The build that costs almost nothing
Here is the version I deploy for companies under roughly 200 contracts, and it has never taken more than two weeks.
Create a Contracts custom object in your CRM. In HubSpot this is a custom object associated to the company and to the deal. In Salesforce it is a custom object on Account. In Attio it is a new object with a reference to Company. The fields are the six above plus a link to the executed PDF wherever it currently lives.
Backfill it. Yes, manually, or with an LLM extraction pass over your PDF folder and a human checking every row. For 200 contracts this is two or three days of somebody's attention. It is the least glamorous work in RevOps and it produces more value per hour than almost anything else I do.
Then automate three things and stop.
I run these on n8n for clients who want the logic on their own infrastructure, or native CRM workflows where the data model supports it. The tooling barely matters. The notice clock alone, calculated correctly and pointed at a named human, recovers more money than most CLM implementations.
One detail people get wrong constantly: alert on the notice deadline, not the renewal date. If the contract renews on 1 March with 60 days notice, the date that matters is 31 December. A reminder in February is a reminder that you have already lost the option.
When you should actually buy a CLM
I am not against CLM platforms. I am against buying one to solve a problem a custom object solves. Here are the conditions where the software earns its cost.
Your contract volume is genuinely high. Over roughly 40 new agreements a month, the drafting and approval routing starts to consume real legal time, and that is what a CLM is good at. Under 10 a month, a template plus a Slack approval works fine.
Buyers redline you heavily. If most deals come back with markup, you need clause libraries, version comparison, and fallback position tracking. This is the strongest single argument for a platform, and it is mostly a function of who you sell to. Selling to procurement at Fortune 500s? Buy the tool. Selling to a VP of Marketing at a Series B? You will not use half of it.
You have obligations that cost money if missed. Service credits, milestone deliverables, regulated commitments. An obligation register you actually maintain is worth paying for.
Legal is a bottleneck with a headcount case. If the alternative to $30k of software is a $140k lawyer, the software is cheaper. If the alternative is nothing changing, the software will not change anything either.
That last row on the left is the one I want to underline. Buying a CLM on top of a broken CRM gives you a second system of record that disagrees with the first. If your account data is unreliable, fix CRM data quality first. Contract records inherit whatever mess your company records already have.
Why so many CLM projects disappoint
Gartner's finding, cited across the category, is that close to half of CLM implementations fall short of expectations, and roughly 40% of first-time buyers replace their system inside 36 months. That replacement rate is remarkable for a category this expensive.
The failure pattern is consistent and it is rarely about the software. Legal evaluates the tool, legal configures the tool, legal loves the tool. Sales encounters a contract request form with 22 required fields, opens a Google Doc instead, and sends it to the customer. Six months later you have a beautifully configured system containing 30% of your agreements, which is worse than having none because now you trust it.
The second pattern is scope. Teams try to turn on request intake, clause libraries, approval routing, e-signature, obligation tracking, and analytics simultaneously. Every one of those needs its own change management. Sequence them or you get none of them.
If you do buy, my ordering is: signature and repository first, then intake and approval routing, then clause library, then obligations. Six to eight weeks between phases. Nothing goes live until the previous phase has real usage numbers behind it.
What the tools cost, roughly
Public pricing in this category is thin, so treat these as negotiation-data ranges rather than list prices.
The number that surprises people is implementation. On the enterprise tier it routinely matches or exceeds the first year of licence. Vendr's marketplace data on Ironclad puts onboarding fees between $10,000 and $75,000 depending on integration scope. Budget for it or the project stalls at the exact point where it would have started paying off.
How contracts should connect to the rest of your revenue stack
A CLM in isolation is a document store with better search. The value comes from where the data goes afterward.
The contract record should be the source of truth for what a customer is entitled to, and three systems downstream depend on it. Billing needs quantity and price to invoice correctly. Customer success needs the renewal date and notice deadline to run a real renewal motion rather than a surprise. Finance needs term, price, and any variable components to handle revenue recognition without a quarterly reconstruction exercise.
If you already run a deal desk or a CPQ setup, the contract sits at the end of that chain, and the handoff from quote to signed terms to billing record is exactly where quote-to-cash breaks in most companies. The break is almost always a manual re-keying step. Somebody reads the order form and types the numbers into the billing system. That step has a defect rate, and the defect rate is your leakage.
My rule: any commercial fact that a human retypes from one system into another will eventually be wrong. Make the contract object write to billing, or make billing write to the contract object, but do not make a person the integration.
The 30-day version
If you want to act on one thing from this article, do this instead of scheduling vendor demos.
Week one, pull the six fields for every active contract into a spreadsheet. Assign it to one person and protect their calendar.
Week two, build the contract object in your CRM and load the data. Add a single required field: renewal owner, a named human.
Week three, build the notice clock. One automation, one alert, pointed at the owner 30 days before the notice deadline.
Week four, run the entitlement comparison once by hand. Contracted quantity against actual usage for every account. Whatever you find pays for the month.
At the end of that you will know whether you have a CLM-sized problem. Most companies discover they had a spreadsheet-sized problem and a discipline-sized problem. The ones who genuinely need the platform will know it by week two, because the exercise will be unbearable, and that is the real buying signal.
Not sure whether you need a CLM or a custom object?
Book a free 30-minute audit. We'll look at your contract data, your renewal process, and tell you honestly which one of the two you actually have.
Book an audit →FAQ
What is contract lifecycle management in simple terms?
It is the management of an agreement across its whole life: request, draft, negotiate, approve, sign, store, track obligations, renew or terminate. Every company with customers already does this. CLM as a software category means putting those stages in one system with an audit trail, so the commercial facts inside a contract stay visible after signature rather than disappearing into a PDF.
Does a 50-person B2B company need CLM software?
Usually not. At that size the binding constraint is almost never drafting speed, it is that nobody can tell you which contracts renew next quarter and on what terms. A contract custom object in your CRM plus one automation on the notice deadline solves that for the cost of a few days of work. Revisit the software question when you pass roughly 40 new agreements a month or when buyers start redlining most deals.
What is the difference between CLM and CPQ?
CPQ handles what you sell and what it costs: product configuration, pricing rules, approvals on discount, quote generation. CLM handles the legal agreement that results and everything after it. They meet at the order form. In a healthy stack CPQ produces the commercial terms, those terms flow into the contract record, and the contract record drives billing and renewal. The handoff between the two is where most quote-to-cash leakage happens.
How do I track contract renewal dates without buying software?
Store the renewal date, the notice period in days, and a calculated notice deadline as fields on a contract record in your CRM. Assign every record a named owner. Then run one automation that alerts the owner 30 days before the notice deadline, not before the renewal date. That distinction matters: if you learn about a renewal after the notice window has closed, you have no negotiating position left.
Why do so many CLM implementations fail?
Two reasons dominate. Legal buys and configures the system without sales in the room, so sales routes around it and the repository ends up incomplete, which is worse than no repository at all. And teams try to launch every capability at once instead of sequencing them. Gartner's number, quoted widely in the category, is that close to half of implementations miss expectations, and around 40% of first-time buyers replace the system within three years.
Abhishek Singla has spent over ten years in RevOps and is currently a Founding GTM Engineer at Peec AI. He builds CRM and automation systems for B2B teams through Ziel Lab.