Pip: Good Strategy Rebellion — where the sacred cows of Silicon Valley come to be skewered, preferably with footnotes.

Mara: Today we’re in the territory Martyn Jones has been carving out: a full-scale critique of data architecture hype, the case against fashionable frameworks, and a defence of the unglamorous systems that actually work.

Pip: Coherence over chaos, basically. Let’s start with the bonfire.

The Data Mesh Reckoning

Mara: The question this segment answers is whether Data Mesh is a genuine architectural advance or a repackaging of old problems in progressive-sounding language — and the book under review, F*CK DATA MESH, does not hedge its answer.

Pip: The review by Paula Prada Price in the New College Tech Review sets the tone early. She quotes Jones calling Data Mesh the “Brexit of data”: grand bluster about sovereignty and freedom from centralized oppression, followed by the inevitable discovery that decentralisation just means more silos, latency nightmares, duplicated effort, and a fragmented mess nobody can govern.

Mara: So the upshot is that the promised liberation — domains owning their data products, self-serve analytics, no more central gatekeepers — lands in practice as more fragmentation, not less. Jones’s argument is that the architecture didn’t solve the organisational problems; it just renamed them.

Pip: And he’s not gentle about who’s been selling the rename. The Kensington review, in The Wired Wire under Lady Amanda Percival’s byline, picks up the “Bullshit Detection Kit” thread — Jones’s term for the tools executives need to spot what he calls “bandwagon-chasers” and “buzzword dribblers” thriving on stack complexity.

Mara: That review lays out the specific targets methodically. The Data Lakehouse gets dismissed as a “naive, dishonest, and disruptive fraud.” SAFe Agile is flagged as disastrous in complex enterprise environments. And the rush toward what Jones calls “Degenerative AI” is framed as organisations preparing themselves for a technology they don’t understand.

Pip: Three distinct failure modes, one underlying pathology.

Mara: The third review — from Bella Carmela at WiredWiredWired — pulls out what’s arguably the sharpest line in the book: “Data doesn’t lie. But a bad analyst with a Power BI license and a KPI target will.” That’s the call to arms Jones closes on.

Pip: All three reviews converge on the same rehabilitation project: the rational data warehouse, subject-oriented, scalable, built for historical truth and actual BI decisions. Not fashionable. Apparently works.

Mara: Jones’s counter-proposal, across all three pieces, is less about a new architecture and more about reclaiming accountability — demanding that the business own its data rather than outsourcing the question to whichever framework has the best conference presence this year.

Pip: Which brings us to the question of what good strategy actually looks like when the hype has cleared.

—

---

Mara: The throughline today is that the most radical position in data right now might simply be insisting on coherence.

Pip: Accountability, consistency, systems that serve people rather than PowerPoint decks. Apparently that’s the rebellion. We’ll be back with more of it.