Linda Evans is Senior Director, Commercial Legal & Operations at Hootsuite, and has run procurement, supported a high-volume sales organization, and sat through more vendor demos than she can count. She’s also spent the last year getting her legal team to actually use AI and has valuable insights about the experience.
Like many legal leaders at fast-moving tech companies, Linda faced an impossible math problem. The volume of work was rising, and headcounts weren’t increasing. Legal teams are being asked to absorb more work with the same amount of resources. The only answer is to innovate.
“It's part of our DNA to use software where we can. If you don't use the tools that are readily available, you won't succeed," Linda says.
Many legal leaders find themselves in a situation where they have a directive and a budget to find AI tools to streamline the workload, but they face a market full of vendors making similar promises. There’s no established roadmap of how to buy or implement an AI tool yet. So the reality is that you have to figure it out as you go, just like everyone else. And that was the experience at Hootsuite.
Buy for the problem you have
Linda came to vendor evaluations with a clear point of view: she knew her problem, she'd done her research, and she didn't have time for a fluffy pitch.
"You've got my time for a reason," she says. "I've done enough background research to get me in the room. Use your time wisely. Show me what you can do."
Her specific need was narrow and well-defined: faster contract review for a small legal team supporting a high-volume sales organization. She didn’t need a company-wide platform, and she didn’t want to roll out a repository that would require change management across the sales org. She wanted one thing done well for lawyers who needed to move faster.
That clarity made the evaluation process easier. She thinks having it is critical for anyone in the market for an AI tool; the best way to get what you want is to know what problem you're solving before you get to the demo. Many vendors will try to sell you more than you want. The right vendor will listen to what you actually need.
An important step—build the guardrails
Issues that Linda had to consider, as she went through the vendor selection process, were how to mitigate the risk of non-legal professionals using AI tools for legal purposes, as well as how to ensure that the AI tools are actually trustworthy.
General-purpose AI tools are accessible to everyone in a company, and not everyone using them to review a contract knows what they don't know. Anyone could run a contract through a general AI tool and miss something a lawyer would catch. This could create risk exposure that the legal team wouldn’t know about until it's too late.
The answer, Linda says, isn't to ban the tools. It's to build the structure around them.
"That's down to having governance policies, having frameworks, educating a team," she says. "And it has to be from the top down."
Linda’s point of view is that creating these guardrails is a responsibility for leadership, not a legal ops task. It requires clear internal communication about which tools are sanctioned for which purposes, what the boundaries are, and why they exist. The legal team can write the policy, but someone at the executive level has to support it.
Slow before fast is actually a good thing
When you roll out an AI tool, there’s a secret no one talks about. Things will often go slower before they go faster. And that’s expected. It’s even okay.
Playbooks have to be built, lawyers have to learn to prompt, and the tool has to be trained on how your organization works. In the meantime, work still comes in, and timelines don't shift. It’s a significant ask for a lawyer who is already under pressure to do something the harder way, on purpose, even just temporarily. That’s the situation Linda found herself in.
"In the beginning, the lawyers told me, it's taking us longer," Linda recalls. "And I said, I understand, because you're training it."
If she had the chance to do it again, she’d make a deliberate space in the implementation schedule for that period, and she’d plan it with intention. There might be a structured onboarding roadmap rather than an open-ended expectation that people would learn the tool themselves. There would be dedicated weekly training sessions with the team. There should be a realistic conversation upfront with her team about what the first few months would actually look like.
"Nothing is plug and play," she says. "Nothing is out of the box and immediately you've got an ROI. There is a real time involved in this. And you have to put the time in yourself upfront."
The teams that go through this bumpy introduction period are the ones who actually get value out of the tool. The ones who don't plan for it often wonder why adoption is low, and face renewal conversations they're not ready to have.
You can't delegate this
The single most important factor in whether a legal team actually adopts an AI tool is whether their leader is willing to use it.
"How can anyone buy into doing something differently if you're not doing it yourself?" Linda points out. "Change comes from the top. You have to walk the walk."
Linda notes that she doesn't ask her team to do things she won't do herself; she models the behavior she wants to see. Not only does she use AI, she shares results in the team Slack channel so everyone can see what she’s doing. To make AI adoption actually stick, the leaders have to be visibly, consistently in the tool alongside everyone else.
The roadmap for adopting AI is still nebulous. But Linda’s experience shows that the teams that are using AI well, with the adoption to show for it, are the ones that treated it as a way to think about changing behavior and ways of working, not as a pure technology problem. People use technology, and understanding how they work and what they need will make its adoption easier and the investment more valuable.
