There are no shortcuts to change management. You still have to embrace the three Cs: communicate, communicate, and communicate. You still need enablement materials in multiple formats: video, email, Slack, in-CRM messages, Claude skills, and whatever else your team actually pays attention to. And you have to obtain and maintain executive buy-in the entire way through. If you need a refresher on the fundamentals, start with our nine best practices for better change management.
But this article is for people with challenges. Maybe it's a leadership team that refuses to agree on anything. Maybe it's a sales team that prefers shadow ops to sales ops, building everything in their AI of choice whether they need to or not. Or maybe it's a marketing team that uses a billion tools, except your marketing automation platform.
Below are the scenarios we've seen repeat across multiple orgs, along with how we've handled them ourselves. Because misery loves company.
Find your loudest naysayer and bring them into your requirements process. Tell them how insightful their feedback is (it usually is, once you get past the delivery), and ask them to test things before launch.
Salespeople talk to each other constantly. If your biggest skeptic loves what you built, they'll tell everyone. The squeaky wheel is always squeaky, so give them something good to squeak about and the rest of the team will follow.
That's the playbook. Here's what it looked like in real life.
I worked at a company with an over-engineered Salesforce instance. Nobody had been willing to challenge the CFO, whose motto was "data entry is why we pay salespeople a base salary."
That's not why salespeople get a base salary. And it's not an excuse for twenty validation rules on the opportunity object. (I'm not exaggerating.)
Creating an opportunity was nearly impossible. Error messages popped up like a RevOps person's Slack notifications after a long vacation. So reps didn't enter pipeline until a deal was almost ready to close, which left leadership with very little visibility into what was actually coming.
I didn't have executive air cover, so I had to build rapport slowly and painfully:
The turning point came in a meeting with the exec team. The CFO said he wouldn't allow us to change anything, so I asked him to create an opportunity in front of everyone. He hit five sequential errors and changed his mind.
I didn't win that argument by arguing. I won it by letting the system make the case.
Word of that meeting got around the sales floor fast. Once the loudest reps heard about it, they were on my side, and the rest of the team followed.
At a different company, the head of sales trolled the head of marketing on social media. His complaint was that the marketing leader didn't pay attention in meetings.
That disdain ran through the entire sales department. Marketing, in turn, wrote salespeople off as a bunch of complainers instead of looking for the valuable feedback buried in the complaints. When the two people at the top openly hate each other, their teams pick sides.
I couldn't fix the relationship between those two leaders. What I could do was take the arguments off the table on both sides.
On the marketing side, we held the team to a higher analytics standard:
On the sales side, reps didn't believe anyone was listening to them. So we showed them:
A sales team that sees its feedback show up in campaigns and results has a lot less to complain about, regardless of what their leader posts online.
And if your leadership team's favorite thing to fight about is account-based marketing, we've written a whole guide on how RevOps can make ABM work.
The first wave of this problem came from how easy CRMs made it for anyone to tinker and pull their own reports. Locking down dashboard creation and consistently communicating a single source of truth shut down a lot of those debates.
Then AI entered the chat.
People started connecting LLMs to their CRMs via MCP (so many initials). Without skills, and sometimes even with them, the LLM went rogue and started guessing at KPI definitions. We landed right back in the spot we'd fought so hard to get away from years ago.
There's no shortcut here. Sorry. The fix is a lot of hard work:
On that last point, you have two options. The first is building skills so people's LLMs at least attempt to use consistent definitions. The second is finding a tool with your key definitions baked in, where the metrics are calculated with consistent code first and an LLM interpretation layer comes in afterward.
The second option is my favorite. It's much more reliable than hoping your skills will hold and your LLM won't randomly barf out a number that starts a battle between marketing and sales.
This is happening everywhere. Every RevOps professional I've talked to is worried about the same three things I am:
Start with your own house. Document how RevOps views and adheres to data governance, then get IT, finance, and legal to review it to ensure the company is protected.
The order matters. In my experience, asking legal to document what they want first is a fool's errand. They don't want the liability, so they'll kick the can for.ev.er. Give them something to react to instead.
Once those teams review your documentation, they'll see why you've been asking for an AI steering committee for months, and "I don't have time for that" stops being a viable answer.
Every company should have an AI steering committee by now. Its job is to lock down who has access to what and when, in enough detail to avoid exposing data to malicious attackers.
Locking things down can't be the whole answer, though. If people can't see a path to the AI tools they want, they'll route around RevOps entirely, which is how shadow ops starts in the first place. Make the approved path visible, and communicate the progress you're making on it.
If you need more motivation to treat this as urgent, read the coverage of Microsoft's 2026 Digital Defense Report. Microsoft found that AI is compressing parts of the attack lifecycle from days to minutes, and phishing rose from 7% to 23% of incidents as the way attackers got in. That includes your website, your forms, and anything else marketing touches. RevOps should be worried about all of it.
There are a lot of great tools coming out in marketing land. (I'm not being sarcastic this time, I promise.) AI is legitimately fantastic at content generation, organization, and ideation. Tools like Quotient give marketers a chat-based agent for campaign planning and execution that also plugs into Claude or ChatGPT.
The catch is that many tools in this space haven't caught up to enterprise standards. Some don't integrate with your CRM or MAP in a way that tracks email opt-outs compliantly, which is mind-blowing for anything that sends email.
When there's no tech purchasing process in place, GTM leaders tend to buy first and ask questions later. So put one in place as soon as possible. Otherwise, you'll be the one discovering after the contract is signed that the new tool doesn't comply with your AI and data governance policies and doesn't integrate with your core systems.
Every one of these scenarios comes back to the same fundamentals: communicate constantly, show your work, and make the approved path easier than the workaround. What changes is which lever you pull when the usual playbook stalls.
If you take one thing away, make it this: the people resisting you the hardest, whether that's your loudest rep, your CFO, or legal, are usually your fastest route to adoption once you bring them in.
Have a change management horror story of your own? We'd love to hear how you handled it (or survived it) in the community.
