One hundred and seven posts of Frictionless Decisions, read end to end, and what they change about the review path set out in EgD-STR-001.
He is not the reach. He is the review. Those are different assets and only one of them
was in the plan.
I. What is held
The full public archive of Frictionless Decisions is mirrored into the EVEglyphDesign Substack archive: 107 posts, 4 September 2024 to 24 July 2026, 83,219 words, every one of them public, each one hashed and registered in registry/PROVENANCE.md.
Nothing is read on Substack. The reading surface is ours, the copy is ours, and the copyright stays his.
II. What he actually writes
Zane Hall's subject is decisions, not dashboards, and his authority is a specific one: he built the decision-support layer at Broadcom in the early 2000s, when it went from an edgy startup to eight billion dollars. His recurring claim is that every manager could see the P&L effect of their own decision and update the plan themselves, giving the CFO instant visibility. Eighteen posts return to that period.
The spine of the publication is a numbered series, The 00001010 Commandments of Data — ten rules in binary, each an imperative. Always define the system of record. Always warehouse the data first. Always store data as journals. Never mask bad data with logic. Never use reports to manage business logic. Never break the link between the business activity and the metric. Never give anyone access to raw data.
Around that spine sit four repeating veins: exception reports as the origin of his method, who owns the data and the security conversation as a partnership rather than a compliance checklist, the shape of a data team — first four hires, smaller teams, centralise or not — and the financial layer: profit centres, product P&Ls, the matching principle, the financial analyst's dream dataset.
That is an FP&A practitioner writing about semiconductors and the ledger. It is, almost exactly, the reader EgD-STR-001 went looking for.
III. What is working
The cadence. Ninety-six consecutive weeks without a gap. Very few publications of any size manage that, and it is the single hardest thing to fake.
The title grammar. Always X. Never X. Try X. An imperative and a noun phrase, no colon, no subtitle doing the work. A reader knows in four words whether the post is for them. The 00001010 device gives an ordinary listicle a memorable spine.
The length. Median 768 words. He does not pad, and he opens with a scene — Custer, South Dakota; a whiteboard formula from Ken Venner; Ray noticing bad numbers on a daily report — rather than a definition.
The newest form is his best form. Connecting Data to Decisions, the roundup he started in March 2026, runs 680 to 1,378 words and is the only format where the word count is rising and the engagement is holding. It is also the only place he reacts to other people's work rather than restating his own.
IV. What is not working, in numbers
The mirror makes this measurable, and the number is the reason to change the plan.
— 107 posts. Mean 2.4 reactions each, median 2.
— 85 comments in total, across all 107 posts and two years.
— 98 restacks in total. The best-performing post he has ever published drew 8 reactions; fourteen posts drew none at all.
Across two years the trend is flat: 3.0 mean reactions in late 2024, 1.7 in the first half of 2025, 2.5 in the second half, 2.7 in the first half of 2026. The rebrand from Frictionless Data to Frictionless Decisions in June 2026 did not move it. Nothing has.
The archive is being recycled rather than compounded. At least six posts are re-runs of an earlier post under the same title, roughly twelve months apart — Always Measure the Whole Before the Parts (January 2025, June 2026), Avoid "Best Practices" at All Costs (August 2025, May 2026), The First Four Hires for Your Data Team (May 2025, May 2026), Let's Game The Casino (June 2025, June 2026), The Most Important Data You Don't Have (September 2024, September 2025), and Connecting Business Intelligence to Artificial Intelligence three times over. The re-runs do not out-perform the originals. A weekly slot is being fed rather than an argument being built.
The plain reading. His following is not on Substack. It is on LinkedIn, where he posts under a headline that reads Author, Consultant, Board Member, and where the Frictionless Data article links actually get traffic. Substack is functioning as his warehouse, not his channel — which is precisely the mistake EgD-STR-001 §II.1 warns about in the other direction.
V. What this changes in EgD-STR-001
Three corrections. The rest of the note stands.
1. Do not ask him for reach. Ask number two in EgD-STR-001 §I.3 — the named mention — is worth very little here; a citation into a list with a median of two reactions moves nothing. Ask number one, the correction, and ask number three, the co-authored piece, are worth more than the original note assumed. He has judgement, a Broadcom P&L pedigree and two years of disciplined argument. That is the asset.
2. He is the reviewer we thought Bansal was. The Business Owns the Data, 14 July 2026, is the EVEglyphDesign sovereignty argument written in his own words: hand the business control of its own data, and the security review stops being a compliance checklist and becomes a partnership. Quote that line back to him and the position paper needs no introduction.
3. Expect him to push on the handover, and let him. His second commandment is never give anyone access to raw data — because raw data is a table copy with no lineage and no context, and people make decisions on it anyway. EgD-POS-001 hands over a whole database. On the surface those collide. They do not: what we hand over is structured, with the structuring published as the curriculum, which is his commandment satisfied rather than broken. But he will find that seam within ten minutes of reading, and it is the most valuable thing he can tell us. Do not pre-empt it in the message.
VI. He asked how. Answer him.
He has already said yes — "yeah I'd love to review it. How should I do that?" — and the answer is still owed. Send this:
Zane — everything is public, nothing to install. The paper is
section III is the part I want you to break — the case for entering an SAP or Salesforce
estate additively, under a non-competition agreement, instead of displacing the vendor.
Comments straight back to me, or open an issue on
the repository if you would
rather they were on the record. Fair warning on one point: your second commandment says never give anyone access to raw
data, and the paper hands the customer the whole database. I think we agree — what gets
handed over is structured, and the structuring is the curriculum — but if that seam does
not hold, I would rather hear it from you than from a client.
Then one sentence, separately, offering him the mirror: his own archive, 107 posts, hashed and readable, at a URL he owns nothing to keep alive. He has two years of work sitting inside somebody else's product. That is not a favour we are asking for; it is one we are doing.
VII. What we will not do
— Not tell him his numbers. They are in §IV because the operator needs them to size the ask. Sending a man his own reaction counts is not a review, it is a discourtesy.
— Not republish him. The mirror is read, cited and hashed. It is not a channel.
— Not treat one strong reviewer as a distribution plan. EgD-STR-001 §IV still holds: three pieces, then count forks.
© 2026 EVEglyphDesign. All rights reserved. Controlled copy. Pour le bien-être du peuple.