PowerUp Coaching — Living Knowledge Base

The Client Who Tried AI and Quit Isn't Being Difficult. They're Right.

5 min read

Topic

A resistance archetype coaches and consultants are about to meet constantly: the client who adopted AI, got unreliable results, and drew a defensible conclusion.

Target Reader

A coach, consultant, or educator selling AI-adjacent transformation. They’re competent, they get good results from their own stack, and they’re starting to encounter prospects who aren’t hesitant — they’re done. Prospects with a story about the time they tried it.

The Fear / Frustration / Want / Aspiration

The frustration of an objection that doesn’t respond to any of the usual moves. Behind it, a question the reader would rather not answer out loud: am I selling a result my clients can’t actually get?

Before State

Sorts resistance into fear, doubt, and overwhelm — all internal states, all coachable with reframing. Calibrates promises to results from their own environment, with its pinned models, its regression tests, and its owner’s tolerance for a bad week. Meets the refusenik with more enthusiasm, and watches it fail.

After State

Recognises a distinct archetype whose objection is factually correct and therefore immune to reframing. Sets the expectation before the disappointment arrives, designs first engagements to survive a model release, and treats their own failures as the credential rather than the thing to hide.

Narrative Arc

Ninety minutes into a mastermind session, a room of experts has traded stories about models that got worse after an update, agents that wouldn’t follow instructions, and a stack requiring 40% maintenance. Bally Binning asks the question nobody had: what does all this do to people who aren’t enthusiasts? The turn is realising that everything survivable-because-interesting to this room is simply broken to someone who bought a promise. The resolution is uncomfortable — the resulting objection can’t be dissolved, only pre-empted, and the coach who has never had a bad AI week has no credibility to pre-empt it with.

Core Argument

The volatility experts absorb as an interesting problem reads to everyone else as a broken product, producing a resistance archetype whose objection is true — which means it can only be managed by setting the expectation before the disappointment, never by reframing it after.

Key Evidence / Examples

  • Bally Binning’s question, verbatim: “Are we seeing the product as the promise? I don’t believe that’s the case yet. A lot of the stuff people are buying — the outputs are not going to be there.”
  • Scott Delinger’s two lines, usable as raw client language: “They’ll be swayed to the ‘avoid AI’ camp” and “it’s just too complicated and changes all the time.” The second is an accurate description of the present, not a rationalization.
  • The session’s own evidence base: a model reported as degraded post-update, a workflow built on a model facing retirement with no drop-in replacement, and a practitioner spending 40% of his time on maintenance — all from people who are good at this.
  • Lou’s connection to enterprise adoption: slow adoption as a rational response to a dependency that changes underneath you, not a courage deficit.
  • Bally’s line on the governance conversation she convened — not minding the gap, “filling the gap with more stuff” — as the shape of the wrong fix.

Proposed Structure (5–7 beats)

  1. Open in the room: experts comparing notes on volatility, comfortable, engaged — then Bally’s question, which changes the temperature.
  2. Draw the line between the two experiences. Same events, opposite conclusions, and the difference is whether you find it interesting.
  3. Introduce the archetype and separate it clearly from fear/doubt/overwhelm. This person did start. That’s what makes them hard.
  4. Concede the objection is true. Spend a full beat here — the piece has no credibility if it flinches, and its whole value is in not flinching.
  5. What this costs the reader specifically: a promise calibrated to expert conditions is a number the client can’t reach, and the gap gets blamed on the technology or on themselves.
  6. The counter — expectation-setting delivered before, a first win engineered to survive a model release, and handling a visible regression in front of the client rather than behind them.
  7. Close on the credential: a coach who has never had a bad AI week can’t do any of this. The failures are the qualification.

Editorial Notes

Lowest actionability score of this session’s briefs (3) and that’s honest — the counter-moves are real but softer than a checklist. Consider whether a diagnostic (how to tell a refusenik from a hesitant prospect in the first ten minutes of a sales conversation) would lift it; that would also make it far more useful to the reader’s next call. The hard editorial discipline here is refusing to make the client wrong — the entire argument collapses the moment the piece implies their conclusion was irrational. Bally and Scott’s language should carry the piece; it’s already the right register. Avoid AI-adoption statistics — the argument is about one person’s experience generalized, and borrowed data would blunt it. Note this is written for the coach, not the refusenik; the refusenik will not read it.

Next Step

  • Approved for drafting
  • Needs revision
  • Deprioritised

Graph View