CLM projects rarely fail because the software doesn't work. They fail because the organisation doesn't change around it.
Legal teams spend months evaluating contract lifecycle management platforms: demos, shortlists, security reviews, procurement. Then the contract is signed, and the real work begins. Data has to be cleaned, workflows built, people trained, and a whole business persuaded to work differently.
That gap between purchase and value was the focus of a recent ITGC session, We Bought a CLM. Now What? The Real Lessons from Legal Teams That Made It Work. ITGC founder Sarah Irwin, a former tech GC who has run her own CLM rollout, was joined by Hylton Lütge of Cera's legal team and Sabrina Pervez, Head of GTM at SpotDraft, the session's sponsor.
Sabrina put the core problem plainly: when a rollout fails, "normally it's because of implementation or adoption, not because the tech itself doesn't work." The seven lessons below are about closing that gap.
Almost every legal team already "has a repository": a SharePoint site, a shared drive, folders spread across individual laptops. Storage isn't the problem. The problem is that the knowledge locked inside those documents can only be reached through a lawyer.
Sabrina described the familiar pattern. Someone in the business asks a "quick question" on Teams or in the corridor. To answer it, counsel drops what they're doing, finds the contract, finds the clause and writes a reply. That can cost an hour, plus the time it takes to get back into the original task. Multiply that across a week and legal becomes, in Sabrina's words, the business's postbox.
Hylton's pre-CLM setup at Cera had the same flaw. There was a central store and named contract owners, but "nothing smart about it": no renewal prompts and no intelligence pulled from the agreements.
The lesson: Judge a CLM by how quickly someone outside legal can get a trustworthy answer, not by how neatly it files documents. In the demo, a plain-English question (which agreements expire in the next 60 days?) returned six contracts, ranked by urgency and linked to the source documents. That is the shift that frees legal's time.
In recent years, end-to-end CLMs have taken criticism, and AI point solutions for review or drafting have taken the spotlight. Sarah admitted she still sees a case for the full lifecycle tool. Hylton's experience explains why.
Cera grows through acquisition, buying and integrating companies much like a private equity firm. Its legal team works with around 1,570 local authorities, each with eight or nine agreements, some running to 160 pages. The team already had capable AI tools, including some that could review 150 documents at once. But each tool held its own slice of the data. Every acquisition meant manually loading new agreements into several systems, and legal ended up guarding information the wider business should have been able to reach.
One joined-up platform changed that. Even signatures moved inside it: Cera replaced DocuSign with signing in SpotDraft, so nobody has to hop between systems.
The lesson: Individual AI features are fast becoming table stakes. The lasting advantage comes from one connected source of truth that grows with the business rather than splintering as it scales.
The instinct after purchase is to migrate everything at once. Hylton resisted it. "How do you eat an elephant? One bite at a time."
Cera's first bite was its most important pain point: local authority contracts. The team brought every stakeholder in, set access rights, decided which metadata mattered, extracted it, migrated the contracts and then audited them. Only once that was solid did they move on. The next bite is more than 160,000 employment contracts.
Sarah ran her rollout the same way, with one deliberate addition. Alongside her two most urgent problems, she picked a quick win: automating NDAs. Within weeks she could show value while the heavier work continued in parallel. Cera did the same.
The lesson: A CLM rollout is a portfolio. Pair one high-visibility quick win, to earn credibility, with one high-impact problem, to earn the right to keep going. Getting either one wrong costs momentum that's hard to win back.
Nobody buys a CLM for its migration features. Yet the quality of what goes in sets the ceiling on everything that comes out. If a SharePoint holds hundreds of duplicate contracts, Sabrina pointed out, moving them into a new repository doesn't relieve the pain. It just relocates it.
At Cera, collating the data was one of the hardest parts of the project. It needed the contracts manager, the data protection officer, compliance and the tech team, all working through what should be shared and with whom. Hylton's payoff came afterwards: once the data was ready, "the system then takes the rest of that work away."
Vendors are taking more of this burden than they used to. SpotDraft's AI flags duplicates on upload even when files are named differently, and its migration specialists assess the current state before anything moves. Sarah noted that this level of support either didn't exist when she bought her CLM a few years ago or came at a steep premium.
The lesson: Treat data cleanup as phase zero, not an afterthought. Every hour spent there is repaid in reports, searches and automations people actually trust.
Legal teams often measure adoption by logins. The better measure is how little the business has to change to get what it needs from legal.
The SpotDraft demo made the point repeatedly. Sales can start a contract from Salesforce or HubSpot, with party details pulled in automatically, and often won't realise they're using a CLM at all. A third-party contract arriving by email goes to legal with one click, through a short intake form. Legal reviews it in Microsoft Word, where the playbook has already flagged the clauses that need attention. Nobody has to learn a new destination.
The guardrails are invisible too. Standard contracts go out without legal touching them. Non-standard ones trigger approvals automatically: an unusual jurisdiction routes to legal, and a deal over $25,000 a year waits for finance sign-off. Every step is logged, so months later anyone can see exactly who agreed to what.
The lesson: design the rollout around where people already work. Self-service with built-in guardrails beats training people to visit a new system, and it lets legal step back from routine work without losing control.
Sarah made what may be the session's most important point: contracts are "a business document, not a legal document." A CLM touches sales, finance, procurement, HR and operations, which makes the rollout a change programme for the whole company, not a legal IT project.
She treated it almost like a marketing campaign. Each function heard about its own pain point. For sales, that meant faster deals and faster commission, because they would no longer need to check routine questions with legal. She backed it up with self-service training in Notion, and recruited account executives as "guinea pigs" for a four-week pilot before going wider.
Sabrina's advice was to bring other teams in early, even during evaluation. That doesn't need a formal pilot. Sharing results at a weekly all-hands is enough to stop the business discovering the decision three months later. After that, trust spreads by word of mouth: if sales likes the tool, they'll tell HR, and HR will happily use it for employment contracts.
The lesson: legal can't mandate adoption, but it can market it. Lead with each team's pain, give them a stake early, and let early adopters do the persuading.
Lawyers instinctively frame legal tech around risk reduction. Hylton argued that this is the wrong opening. Risk, he said, isn't "a today problem." Finance will run a simple calculation: what does this cost, and what does it return?
He offered two ways to answer that.
Sarah built her own business case the same way, using data on the cost of slow manual work, and added internal allies in finance and sales ops who shared the pain. Her challenge to legal leaders was pointed: before you reach for headcount, ask whether the work is something technology could, and should, be doing instead.
The lesson: risk reduction is the floor, not the pitch. Lead with a number finance can verify, and keep reporting on it. Cera's automated monthly reports, fed by the CLM into other teams' systems, keep the value visible long after purchase.
Taken together, these lessons point to a change in what a legal team is for. The old model put legal at the centre of every contract question: the holder of the documents, the approver of every draft, the answer to every "quick question." The CLM rollouts described here move legal to a different role. It designs the playbooks, guardrails and data structures that let the rest of the business move safely on its own.
That is also why Sabrina's first piece of advice came before any demo: know your priorities before you speak to a vendor, because every tool looks impressive in a demo. A CLM amplifies whatever operating model you bring to it. If you bring clarity about your pain points, your data and the value you want to create, it compounds. If you bring chaos, it scales the chaos.
Hylton's closing advice is a good place to end. A rollout of this size is daunting. Break it into prioritised tasks, work through them methodically, and "you just go step by step by step."
This piece is based on the ITGC session We Bought a CLM. Now What? The Real Lessons from Legal Teams That Made It Work, hosted by Sarah Irwin with Hylton Lütge (Cera) and Sabrina Pervez (SpotDraft). Thanks to SpotDraft for sponsoring the event.
