This piece is based on sponsored content published in Marketing Dive by Amy Crowther, CEO of Incubeta. The agency's promotional framing has been left out; what remains is the part practitioners can use.
Crowther describes walking into a meeting with a client's Chief Revenue Officer — only their third conversation. Instead of talking strategy, they ended up workshopping how she should write an email to her own boss.
That, she argues, is where a lot of client conversations sit right now. Not "what's our AI strategy," but "how do I even get the people above and around me to acknowledge this is happening."
Underneath the talk of technology, her conversations with senior stakeholders are deeply human. People ask to have things brought down to their level — not because they aren't smart, but because nobody has had time to build real fluency yet, and admitting that out loud still feels risky for a lot of executives.
Everyone is in the same boat. The only real difference is which shade of frustration or overwhelm someone happens to be experiencing on a given day.
Crowther thinks this has reduced a lot of ego in the industry, and that it had to. When nobody genuinely knows how to navigate something, the only way through is to be open about it and accept help. She has watched it level rooms that a few years ago would have been far more defensive and performative. People aren't posturing about AI maturity the way they once postured about their martech stack. They're mostly just asking someone to help them get unstuck.
This is the most operationally useful part.
The reason deals stall — the reason projects slow down even when everyone in the room nodded along and said "green light, let's go" — is rarely a strategy problem. There is almost always one person, somewhere in the workflow, with an unspoken fear that this will take their job, expose that they don't understand it, or shift the ground under a role they've held for years.
The moment you hit that person in the chain, everything stops. Not because they say no. Usually because they say nothing at all — and silence is a much harder thing to work with than a direct objection.
Her example is concrete. A client didn't need a yes or a no from a senior stakeholder; she just needed acknowledgment so the work could move forward. She pushed, she emailed, she tried every collaborative approach to support that person and bring them along. None of it landed, because the resistance was never really about the plan on the page. It was about something unspoken that no meeting agenda would surface on its own.
Eventually the only way forward was to build enough momentum around that person, with everyone else who was ready to move, rather than waiting for them to move first.
There is an old model of agency leadership: walk in and say, "We've figured it out. This is the way. Here's exactly how it's done and here's exactly how you'll do it with us." The appeal is understandable. In theory it reassures, especially a client who is exhausted and wants to hear that it's handled.
But right now, in AI and in almost every category moving this fast, if someone tells you they've fully figured it out, that should be your first warning sign, not your reassurance. Clients sniff that out almost immediately.
What holds is the foundation — the things that stay true regardless of which tool or platform is trending this quarter. The rest gets built with the client, in the room, in real time. And what gets built this week may look genuinely different in six weeks, because the ground is still moving. That is not a weakness in the process. It is the honest shape of the process right now, and pretending otherwise just delays the moment the client finds out anyway.
Trust used to be inherited. You worked with Coca-Cola, so a new client assumed you could handle their brand too, without much scrutiny — a kind of referred trust that came from the logo on your case study slide. That trust still exists in pockets, but it is not what wins new business anymore, and it is certainly not what holds relationships together through a period like this one.
Trust today is earned meeting by meeting, by being honest about what you don't yet know and showing up prepared anyway.
The advice compresses to one line: don't perform certainty you don't have. Name the uncertainty specifically and bring a real plan for working through it together, not a polished deck pretending the uncertainty isn't there.
"Specifically" is the operative word. "AI moves fast so nothing is certain" is evasion dressed as honesty. The usable form separates three things: what is definitively working (the foundation), what could look different in six weeks (the assumptions), and what happens when it does (review triggers and criteria). Written down separately, uncertainty stops being a risk and becomes a managed item.
The prescription for the silent stakeholder is equally practical. Instead of repeating the persuasion, build momentum with the people who are ready. If the resistance is not about the plan, explaining the plan better will not resolve it.
For the organizational version of asking without changing anything, see Why Company Assessments Change Nothing. For how AI shifts the ground under roles inside a company, see Why No New Giant Has Emerged in the AI Era.
Rarely because of strategy. There is usually one person in the workflow with an unspoken fear about their job or role who says nothing rather than objecting — and silence is harder to work with than a stated no.
Rather than repeating persuasion, build momentum around that person with everyone else who is ready to move. If the resistance is not about the plan, explaining the plan better will not resolve it.
In categories moving as fast as AI, a claim of total certainty should read as a warning sign rather than reassurance — and clients tend to detect it almost immediately.
Name it specifically and pair it with a real plan. Separate what is definitively working, what could change within weeks, and what the review triggers are — that turns uncertainty into a managed item.
To apply what you just read to your own site, start with a free audit of where things are now.
A strategist replies within 24 hours on business days.
Jensen Huang reads reports from every level, and founders everywhere took it as permission. But the variable is not frequency — it is whether your gate controls execution or collects information.
Assessments fail for two reasons: purposeless design and unexecuted results. Here is how to design an organizational diagnosis and actually use what it produces.
Cohort retention analysis exposes what growth rate hides. Here are the three curve shapes, benchmarks by category, and what investors actually read off the chart.