Two stakeholders, same meeting. You describe the same initiative twice: once as "agentic AI capabilities", once as "a self-service capability initiative". One framing creates resistance. The other creates engagement. Same work. Same scope. Same team. Different verb.
The first instinct is to feel this is dishonest. It isn't. Language choice is part of the PM's job, not separate from it. Translation is a stakeholder skill that most PMs underuse, partly because it feels like spin, partly because nobody teaches it.
1 · Why labels carry weight
In regulated and enterprise environments, certain words trigger loaded assumptions before the work is evaluated. AI triggers concerns about hype, cost, governance, and explainability. Transformation triggers memories of failed programmes. Autonomous triggers compliance questions. Agentic often triggers all of the above.
The assumptions aren't always wrong. They're just rarely about this particular initiative. When the label triggers a pattern, the evaluation happens against the pattern, not against the work.
A reframe isn't about hiding what the work is. It's about making sure the work gets evaluated on its own merits, by a stakeholder who isn't already running a pre-built objection in their head.
2 · Three reframes that earn their keep
Three patterns come up repeatedly. None of them changes the work. Each changes the conversation the work earns.
| Loaded label | Reframe | When it earns its keep |
|---|---|---|
| Agentic AI | Self-service capability initiative | When the audience reads "AI" as risk rather than opportunity. The operational framing anchors the conversation in utility before label. |
| Transformation programme | Operating model improvements | When leadership has been burned by previous transformation programmes. The smaller framing earns a hearing. |
| Data platform | Reporting reliability work | When the audience is operational rather than technical. Outcome-language reaches them faster than system-language. |
Each reframe trades abstraction for utility. The work is unchanged. The conversation it earns is different.
3 · When the label matters
Sometimes the loaded label is the right word.
- If you need infrastructure investment, AI may be the right word. Boards fund AI; they don't fund "self-service capability".
- If you need governance approval at scale, transformation might be the necessary frame.
- If you're talking to engineers, the technical term earns credibility.
The reframe is for the audience that won't fund the abstraction. The original label is for the audience that will. A skilled PM uses both: different audiences, same week, same work.
4 · Reframing vs spin
The line between reframing and spin matters. Get it wrong and you stop being trusted.
A mentee I work with reframed "agentic AI capabilities" as "self-service capability initiative". The scope didn't change. The architecture didn't change. The benefit story didn't change. What changed was that peers who'd been blocking on "AI" started discussing the use cases. That's translation, and it's the move worth practising.
5 · Building the reframe before the meeting
Reframing under pressure is hard. The audience asks a question, you reach for a label, you grab the closest one. Sometimes it's the wrong one.
The discipline is doing the work beforehand. Three quick questions to ask before a stakeholder conversation:
- Who's in the room? Operational, technical, financial, regulatory? Each audience hears different verbs.
- What's the loaded label here? The word you'd use casually with peers. The one that triggers a pre-built objection in this room.
- What's the cleanest reframe that's still true? Truthful framing first, anything else burns trust later.
Two minutes of preparation. The kind of preparation that distinguishes a senior PM whose initiatives ship from one whose initiatives stall.
- Reframing without delivering. The reframe buys you the conversation; the work has to back it up. A reframe that wins a meeting and then doesn't ship is worse than the original label.
- Using one reframe with all audiences. Different stakeholders need different entry points. A reframe optimised for resistant peers may flatten the work in front of leadership who would have funded the bigger label.
- Reframing instead of saying no. If the underlying work is wrong, a new label won't save it. Reframing is for good work that isn't getting a fair hearing, not for bad work that isn't getting buy-in.
- Hiding the reframe from yourself. If you can't say in plain English what the work actually is, you've crossed from translation into spin.