The hackathon we planned with PointClickCare’s marketing team never happened.
We had the format, the workflows, and people ready to build. Then AI got more agentic, and the platform questions came up. Which tool would the company standardize on? Which connectors would security enable? What would meet legal guidelines? The hackathon slipped, then slipped again, and we let it go.
A hackathon, the way I run one, is a multi-week build window that ends in a showcase day, where the team sees what’s been built and works out what’s worth keeping. It’s the turning point, where a team stops building in pockets and starts building together.
I’m opening with the failure on purpose. You’ve read AI transformation stories that run in a straight line from kickoff to results. This one didn’t, and the zigs and zags are where the learnings are.
PointClickCare, a leading health tech company supporting over 30,000 provider organizations and every major U.S. health plan, is now running one of the most thoughtful marketing redesigns I’ve seen. They got there through stalled plans, platform changes, and governance debates, and they kept going anyway.
The context is what makes this one worth studying: healthcare data, three AI platforms in play, a skill that everyone in the company now uses with customers, and IT, security, legal, and finance involved from the start.
Renée Gapen, CMO of PointClickCare, agreed to share the good, the bad, and the messy version of the story. Her reason was simple: nobody is an expert at this yet. Her team got further because other companies were open about what worked and what didn’t, and she wants to do the same for the next team.
Quick Take
Just starting gets you further than waiting for the perfect plan. Every detour this team took taught them something they used in the next step.
- Asset development went from months to weeks, sometimes days. Specialists no longer create every asset. They started building the systems that produce them, encoding brand rules, templates, and product context into skills, which are instructions that tell AI how the company does a specific kind of work. The business owner gets a first draft in minutes, and nothing goes out until a specialist approves it.
- Marketing built a skill about 2,000 people across the company now use. A redesign that started in one function now changes what customers see, so PointClickCare shows up as one company rather than several.
- A governance question answered the platform question, and it wasn’t the question the team started with. Security went from the checkpoint that slowed things down to the group that helped test and deploy the first company-wide skill.
- The four problems driving the redesign have nothing to do with tools. They are about where knowledge lives, how fast work happens, who touches it along the way, and what the team learns from what it launches.
- Before FY27 planning, the team pulled five people out of their day jobs for 90 days to redesign the workflows first. Roles and org design follow the work, and I’d argue that’s the right order.
Watch or Listen Instead
To cater to different learning styles and time constraints, this newsletter is also available as a 4-min cinematic AI video explainer and a 16-min AI podcast with two AI hosts. Both were made with Google’s Gemini Notebook and personally reviewed for accuracy and responsible AI use.
Where They Started
When we began working together in late 2025, PointClickCare’s marketing organization of about 80 people was already using generative AI in real ways. Roughly 60% of marketing content was already being accelerated by AI-assisted first drafts. The energy was there. How work got from one person to the next was the problem.
Work ran on briefs and handoffs. A business owner who understood the product, the customer, and the market wrote a brief, handed it to a specialist, waited, reviewed, revised, and waited again. Each handoff added delay and interpretation risk. The people closest to the problem weren’t generating the output. And the business was speeding up, especially around new product launches, faster than the process could keep up.
This is the marketing model many of us grew up in, built for a world where production capacity was the constraint. That constraint changed.
Our engagement ran over several months: an in-person leadership and strategy workshop, virtual sessions for the full marketing organization, function-specific hands-on workshops, and a technical session for the ops and builder group with my automation business partner Justin Parnell. Weekly office hours and syncs with the internal program champion kept it going.
The Messy Middle
This is the part that doesn’t usually get covered in case studies.
The team’s early momentum ran on ChatGPT and custom GPTs. People liked them, built with them, and learned fast what good looked like. But that adoption stayed inside marketing.
PointClickCare runs on Microsoft. Teams, SharePoint, the tools everyone lives in all day. So Copilot was the natural enterprise path, and the company leaned in. The timing was the hard part. The agentic capabilities marketing wanted were early across the whole category in that window, and the security reviews for connectors and access were still being worked out. Those reviews were doing their job. This is healthcare, and the guardrails exist for good reasons.
That’s when the hackathon died. It wasn’t lack of interest. It was honest uncertainty about which platform to build on and what the rules would be.
The hypothesis going in was that everyone should be building, and what they found refined it. Plenty of work is self-contained, and the person closest to it should start. But workflows that need connectors, system access, and coordination with IT, legal, and finance require technical depth and alignment work that sits outside a marketer’s day job. More of their work turned out to be that second kind than the plan assumed.
“We tried to be overly efficient out of the gate. We tried to find the perfect platform first, with everyone building, and it stalled us. What moved us forward was being honest about where we were: experimenting in silos, with thin governance and rising expectations, especially as AI-enabled innovation halved product development cycle time. Once we understood that, we could design the operating model we actually needed.”
When Security Became a Partner
Here is my favorite part of this story. The same governance questions that contributed to stalling the hackathon ended up answering the platform question.
When the team looked at where their first company-wide skill should live, one they would build, test, and govern centrally, the deciding question became governance: which platform lets us push a skill to everyone, manage it centrally, and keep quality and compliance in one place. Working with their security team, they found their answer. Their IT organization then enabled Claude across the company, which gave marketing the technical orchestration that had been out of reach when the journey started.
The answer was a layered portfolio, with each tool doing the job it does best.
- Microsoft 365 is the foundation, running the collaboration and productivity backbone the whole company works in, with Copilot woven into those tools.
- Claude became the home for marketing’s centrally governed skills and agentic workflows, with more of their systems connected to it every couple of weeks.
- ChatGPT gets credit too, because it’s where the team learned what they wanted AI to do in the first place.
The brand skill itself tells you how the relationship with security grew. Marketing built it with brand guidelines, sample assets, and rounds of human testing with the teams that live in presentations every day. Security helped test and deploy it, and the skill now works as a brand governance layer, so work starts on brand instead of getting fixed after the fact.
None of this is theoretical in healthcare. Every skill launched to the entire company has to hold up under the same scrutiny the product does.
Read that again. Security went from one of the reasons plans stalled to the group that helped get the first skill into everyone’s hands. When you treat governance requirements as part of the design, the people who own those requirements build with you instead of reviewing your work at the end.
Willow Terry, VP of Market, Competitor and Customer Insights, led this work, coordinating across marketing leaders and partner teams.
“What felt like a stall was actually a turning point. The pause gave us the opportunity to bring together marketing leaders, each with a deep understanding of their function’s workflows and pain points, alongside experts from IT, Security, Legal, and Finance.
Together, we shifted from experimenting with AI tools to designing the capabilities, governance, and ways of working needed to scale AI across the company.”
About 2,000 people across PointClickCare now use a skill that marketing built. Whatever an employee puts in front of a provider or a health plan starts on brand, so PointClickCare shows up to the market as one company rather than several.
Reimagine the Workflows First
The people closest to the work kept bringing up the same underlying problems, whatever function they sat in. So instead of starting with tools, the team stepped back and redesigned the work itself. They pulled five people out of their day jobs for 90 days to rethink how marketing operated and find the root causes limiting speed, scale, and effectiveness.
Ninety days of five people is what it took at their scale. Yours might be different. What counts is that the time is protected, because transformation doesn’t happen on the hamster wheel. That sprint is where the four-stage workflow, the skill-building playbook, and the hub-and-spoke operating model came from.
Here are the four problems they were solving.
Notice that none of these are tool problems. They are workflow and knowledge problems, which is why a tool-first approach kept stalling.
Nobody at PointClickCare was collecting what worked, because an all-human team could only do so much and that work got skipped for years. Still hard, or just historically hard? For this one, the reason is gone. Check out 5 Reimagined Workflows: Work We Skipped That AI Now Does to learn more.
Dena Stewart, one of the marketing team’s 90-day orchestrators, put it this way: “The big move was about how do we move from experimenting as individuals to performing as a team.”
What They’re Building
The business owner still says what they need. What changed is who answers. Specialists encoded the brand rules, the templates, and the product context into skills, so AI produces the first draft in minutes and the specialist reviews it instead of building it from nothing. The brief didn’t disappear either. It gets produced automatically, downstream, at the point the work enters production.
The draft routes into production through Adobe Workfront’s AI briefing assistant, which fills in the project details, so the work is ready for execution. Humans still approve. Functional experts check the work instead of producing it, and quality is still their call.
The specialists’ work changed. They own the skills now, test them, and keep them current, and knowledge that used to live in their heads is written down where everyone can use it.
Cameron Ingrum, Director of Product Marketing at PointClickCare, is on the business owner side of that change.
“We used to measure asset development in months. Now it’s weeks, sometimes days. That didn’t come from the tool. It came from two things working together: the context we gave Claude about how we position and who we’re selling to, and the templates and brand rules our creative team encoded for each type of asset.
Together they get a product marketing manager to a v1 that’s 80% complete before a creative project even starts. Our designers spend their time at the top of their license now, refining instead of producing from scratch.”
What They Built to Keep It Going
Two other things came out of this work, and both are built to keep running after the 90 days end.
A playbook for building skills
Every skill goes through the same process. It takes longer than turning on a tool, and that’s the point.
- Define the job the skill is doing.
- Gather the source of truth: the actual brand guide, the real templates, the examples the team would point to as gold standard.
- Encode all of it into instructions the AI can follow.
- Run three structured QA rounds where the AI generates real artifacts and employees pick them apart, first for structure, then for edge cases, then for polish.
- Hand it to the teams who own the work to run on their real projects. Creative and Brand stress-tested the PowerPoint skill. Content put the copy guide through its paces.
- Only after those teams sign off, and legal and security have had their look, does IT validate the finished package and turn it on for the whole company.
Not every skill goes through all of that. One thing decides which ones do: how many people will use it.
An org design decision
The team looked at the three ways to organize AI in marketing: one central team that builds everything, every function building on its own, or a hub and spoke mix. They were already decentralized informally, which in healthcare is exactly the tool sprawl and compliance risk you’d worry about. They’re going to hub and spoke.
They landed on some version of democratized building with centralized enablement, which is where the lesson from the stall ended up. Their version puts the playbook, the quality gates, and the relationship with IT in the hub, while each function builds on that foundation and owns its outputs. How the pieces get divided will keep changing, and the part worth borrowing is the decision itself: the hub owns the system and the standard, and deliberately does not own every build.
Early Indications
The brand skill is in production across the organization, and it’s speeding up the new product introduction workflows. That skill was about brand consistency, and it was also about proving that a governed skill could be built, reviewed, and released across the business.
They haven’t re-measured the 60%. What changed in nine months is who produces the first draft and how many hands it passes through.
Most of what’s in production today is the same work happening faster, and faster is the floor. The work that couldn’t be done before, the learning loops and the insight repository, is next. That’s an honest place to be nine months in, and it’s the part I’d watch, because innovation is where the growth comes from.
Think big, start small, move fast. Proving one skill company-wide is what makes the next set faster to build.
As Jake Campbell, PMP, another of the marketing team’s 90-day orchestrators, says, “It’s not a build once and you’re done. We’re going to learn as we go.”
What to Take From This
Reimagine the work first, and the roles, the governance, and the org design follow. Three things from this team’s journey apply, whatever your industry.
- When a platform decision stalls you, the platform usually isn’t the problem. Look at the workflow and knowledge problems underneath it. Bring security, IT, legal, and finance in early and treat their requirements as part of the design. The team that slows you down at the gate can become the team that builds with you.
- Build the shared context before you scale the outputs. If the brand, audience, and product knowledge are in people’s heads, AI makes the handoff problem worse. If it sits in files your tools can reach, the handoff goes away.
- Start before you’re sure. This team brought in new platforms, cancelled a hackathon, and rebuilt their approach more than once. Every detour informed the next decision.
Each of those three is a reframe in practice. If you want the full set, I covered the eight reframes that change how GTM teams see AI in a recent issue.
Where to Start
Pick one workflow where a brief changes hands. For PointClickCare it was the new product introduction. Map where the knowledge for that work actually resides, whether that’s in someone’s head, in scattered documents, or somewhere your AI tools can reach. That map tells you what to build first.
The Practical AI in Go-to-Market newsletter shares learnings and insights in using AI responsibly. Subscribe today and let’s learn together on this AI journey.
For applied learning: Explore our applied AI workshops, offering both strategic sessions to shift mindsets through practical AI uses and hands-on sessions to build human+AI workflows. You’ll leave with either a clear plan or working solutions. Office hours provide follow-on support to keep momentum going.
For ongoing advisory: I work with leaders as a strategic advisor through their AI transformation, providing guidance, frameworks, and a sounding board as they build. If that’s what you need, let’s talk.
For team transformation: I partner with GTM leaders to move from scattered AI adoption to an operating model where human and AI teammates work together across workflows. See real examples: a lean GTM team’s step-by-step playbook and a global cybersecurity leader scaling to 150+ marketers with 57 AI teammates integrated into daily workflows.
For speaking: Here are virtual and in-person events where I’ve covered a variety of AI topics. I’ve also keynoted at many conferences and corporate events.
If any of this resonated, I’d love to hear from you. Reply to this email, find me on LinkedIn, or come say hi at an event.
