ZNU LABS

Counting is Hard

Tue, 01 Jan 2019 00:00:00 +0000

Counting is Hard#

On day one of a statistics class we are taught that counting is hard. Depending on what question is asked, the numbers will vary.

Reading the following will take a few minutes. Not reading it will cause hours of misunderstanding in quarterly meetings.

Three kinds of report#

Time domain reports show results from a time period. You can see results by progress category. You cannot see yield, or what came from what.

incept → tn-2020-000

Mon, 01 Jan 0001 00:00:00 +0000

The Science of Counting#

ZnuLabs // Origin note, written in lockdown, 2020

I. The stillness#

COVID had changed everything, and the first thing it changed was the pace.

The world stopped moving. Airports emptied. The consulting calendar — twenty years of it, the thing that had structured every week of my working life — went blank in a fortnight. And in that stillness I found myself staring at an old problem through a new lens, because for the first time in two decades there was nothing else to stare at.

The Orchestration Is the File

Sat, 14 Mar 2026 00:00:00 +0000

The Orchestration Is the File#

Buenos Aires, March 2026

Core thesis#

The GUI is a dead end. Every generation of marketing and RevOps orchestration tooling — Eloqua, Marketo, HubSpot, Pardot, LeanData, Clay — repeats the same pattern, and solutions built in that pattern break down at an early phase. They make the orchestration visual instead of data. The canvas looks great at Hello World and becomes unmaintainable long before it becomes large. No version control. No diff. No portability. No import, no export. Screenshots and prayer.

counting → tn-2021-001

Mon, 01 Jan 0001 00:00:00 +0000

Counting is hard: a formal treatment#

Three reports, one surface

Abstract. Business reporting produces numbers that do not reconcile, and the non-reconciliation is treated as a defect to be argued away in meetings. It is not a defect. Period, cohort, and operational reports are three projections of a single two-dimensional object — the Lexis surface of entity trajectories indexed by entry time — and the information each discards is exactly the information the others retain. We formalise the object, classify the three report families as the three ways to degenerate a two-parameter selection, and derive four consequences: that the naive yield ratio is valid only when transit time vanishes; that cohort estimates are downward-biased by a computable factor admitting a cure-model correction; that the stock–flow reconciliation residual measures unlogged destructive mutation; and that period conversion rates fall when sales cycles lengthen even under constant propensity to buy.

Sovereignty Is a Supply-Chain Problem

Thu, 20 Aug 2026 00:00:00 +0000

Sovereignty Is a Supply-Chain Problem#

Core thesis#

EU data sovereignty is achievable today, with today’s vendors, at commercial prices. We built a data services company on that premise and it holds.

But it is not a checkbox, and it is not a procurement filter you can apply in an afternoon. It is a supply-chain audit that runs the full depth of the stack, and the two failure modes that matter most are invisible on any pricing page.

trees → tn-2022-002

Mon, 01 Jan 0001 00:00:00 +0000

The tangled trees#

What CRM inherited from the subscription list, and what it left behind

Abstract. The customer relationship data model is not designed. It is inherited, and its ancestor is the magazine subscription list. We trace the descent — private post riders, the Postal Acts of 1794 and 1879, the rented mailing file, the business response card — and show that the modern schema retains the acquisition half of that model while having silently dropped the deletion half, which is what made the original work. We then state the structural complaint formally enough to be actionable: the stored schema is a forest of type hierarchies, the world it represents is a connected graph, and the edges that carry the most commercial information are precisely the ones with no column to live in. The formal treatment of the resulting projection is deferred; this note establishes what is being projected and why the shape is what it is.

attribution → tn-2023-003

Mon, 29 May 2023 00:00:00 +0000

Success Has Many Fathers: Attribution Without Identification#

Why credit models rank channels by diagnosticity, not by cause#

1. The problem#

Success has many fathers. Failure dies an orphan.

The proverb is normally read as a comment on human vanity. In marketing analytics it is a description of a data-generating process, and it is doing more damage than the vanity it describes.

A multi-touch attribution report shows that converting deals touched, on average, eleven channels. Credit is distributed across those eleven. Every channel appears to work. The budget conversation that follows is about reallocating fractions, and the fractions are defended with a model.

structure → tn-2022-04

Mon, 01 Jan 0001 00:00:00 +0000

Abstract. Customer relationship platforms are commonly described as fracturing their subject matter into silos. We show the fracture is not a schema defect but the composition of two distinct errors: the conflation of a referential layer with an evidential one, and a lossy projection applied to the second. The Lead/Contact division reduces to the definedness of a single partial function and carries no further information. Group interaction, by contrast, is genuinely destroyed: the stored activity set is not a faithful image of the incidence structure that produced it, and no query recovers what the projection discarded. We decompose the observable fraction of an account’s interaction history into three independent factors, give an identifiability condition under which one of them is recoverable by clustering, and bound account coverage from above by capture–recapture. Field correctness is out of scope and treated in the successor note.

entropy → tn-2025-005

Mon, 01 Jan 0001 00:00:00 +0000

The thermodynamics of business data: decay, injection, and the economics of standing remediation#

And, like this insubstantial pageant faded, / Leave not a rack behind. — The Tempest, IV.i

Abstract. The data held in customer and resource platforms is not a static asset but a decaying one. We model a record’s correctness as a survival process with a finite half-life, and the platform’s aggregate quality as a reservoir subject to two opposing flows: continuous decay of standing records, and continuous injection of new records whose quality is bounded by the discipline of their capture. We show that such a system equilibrates at a quality strictly below that of its own inflow, that the fraction of jointly-usable records collapses geometrically in the number of fields a decision requires, and that one-time remediation relaxes back to equilibrium on the decay timescale. A standing remediation term is therefore not an operational preference but a structural necessity; and because the same exponent that governs the collapse governs the recovery, the return on that term is leveraged rather than linear. We close with the recoverability floor that bounds what any method can achieve.

grail → tn-2026-006

Mon, 01 Jan 0001 00:00:00 +0000

The confluence: single stream as a reachable condition#

We are such stuff / As dreams are made on, and our little life / Is rounded with a sleep. — The Tempest, IV.i

ZNU LABS

Counting is Hard

Counting is Hard#

Three kinds of report#

incept → tn-2020-000

The Science of Counting#

I. The stillness#

The Orchestration Is the File

The Orchestration Is the File#

Core thesis#

counting → tn-2021-001

Counting is hard: a formal treatment#

Sovereignty Is a Supply-Chain Problem

Sovereignty Is a Supply-Chain Problem#

Core thesis#

trees → tn-2022-002

The tangled trees#

attribution → tn-2023-003

Success Has Many Fathers: Attribution Without Identification#

Why credit models rank channels by diagnosticity, not by cause#

1. The problem#

structure → tn-2022-04

entropy → tn-2025-005

The thermodynamics of business data: decay, injection, and the economics of standing remediation#

G. Righter/ZnuLabs · Copyright © 2021-2026, Grover Righter#

grail → tn-2026-006

The confluence: single stream as a reachable condition#