// SCORECARD · FOUNDATION · GOVERNANCE · AI

Most institutions do not have a data problem.

They have a trust problem wearing a data costume. There is plenty of data — what is missing is confidence that the number is right, that everyone means the same thing by it, and that a decision made on it holds up to a regulator, a board member, or an angry customer.

free guide·10 chapters·free guide file
free download

Get the Field Guide — free.

Workflows covered in the guide:

  • Score all eight dimensions with the business in the room — and find out where you disagree
  • Name one system of record for every entity that matters, and treat the rest as copies
  • Run a quality loop that profiles, tests, and fixes at the source instead of patching the report

instant download + emailed copy · no spam · unsubscribe anytime

// the problem

The order matters, and almost everyone gets it backwards.

Score where you really are, build a foundation people trust, fix quality, agree on definitions, govern it so you can move fast, then put AI to work and measure what it returns. Skip to the AI part before the foundation is solid and you will ship confident, expensive mistakes — faster, and with a better-looking dashboard on top of them.

// what you’ve tried

You already know which parts are shaky.

Bought the platform first, on the theory that the platform would settle the definitions

Watched two teams bring different numbers for the same metric and spend the meeting arguing about whose report was wrong

Shipped a dashboard nobody opens, because nobody believes the number on it

None of that is a tooling failure. A report is not wrong when two teams disagree — the definitions are. And a foundation is not a shopping list of tools; it is four decisions made on purpose, in an order that most roadmaps invert.

// what changes

Here’s what the guide covers.

p. 4

Score all eight dimensions with the business in the room — and find out where you disagree

from “Score your data reality”
p. 5

Name one system of record for every entity that matters, and treat the rest as copies

from “Build a foundation people can trust”
p. 6

Run a quality loop that profiles, tests, and fixes at the source instead of patching the report

from “Fix data quality”
p. 7

Certify the handful of metrics that cause the most fights, and end the argument for good

from “Agree on definitions”
p. 8

Know which model-risk regime you are actually under, now that SR 11-7 is gone

from “Govern it so you can move fast”

Choose a workflow that fits your task, then check the requirements before trying it.

Choose a workflow you can put to use.

Start with one task and the matching chapter. Gather the inputs, try the example, and check the result against your requirements. The guide is a reference for that process. Your tools, access and experience determine what is practical for you.

1 of 24 free guidesbuilt by an AI consultancyno fluff, no paywall
// inside the guide

A checklist, not a lecture.

Eight dimensions you score with the business in the room, so the disagreement itself becomes the finding. The six data-quality dimensions and the loop that keeps them healthy. Lake, warehouse or lakehouse, chosen for the right reason. What model governance actually requires now that SR 11-7 has been replaced. A one-page checklist and a first year, sequenced.

01

An eight-dimension scorecard you fill in with the business in the room

02

Lake, warehouse or lakehouse — and how to pick for the right reason

03

The six data-quality dimensions, and the loop that keeps them healthy

04

What model governance requires now that SR 11-7 has been replaced

05

A one-page checklist and a sequenced first year

10 chapters
Score your data realityBuild a foundation people can trustFix data quality (the unglamorous engine)Agree on definitions before you argue about numbersGovern it so you can move fastPut AI to work (on a trusted foundation)Ship insight people actually useMeasure what mattersYour first year, sequencedThe one-page checklist
// how it works

Three steps, and you’re reading in under a minute.

1

Tell us where to send it

Just your first name and email — that’s the whole form. No hoops, no credit card, nothing to install.

2

It’s yours in a second

The guide downloads right away, and we’ll pop a copy in your inbox too, so you’ve always got it when you need it.

3

Try one thing this week

Open it, pick one idea that fits your week, and actually use it. One small win is all it takes to see what this thing can do.

// before you ask

A few fair questions.

What should I check before using the guide?

Read the chapter list and examples to see whether the guide fits your task. The documents describe a June 2026 snapshot. Product features, access and pricing can change, so confirm any setup instructions against current vendor documentation.

Do I need paid software?

The download is free. Some workflows may require a product account, subscription or hardware. Check the requirements for the workflow you choose before buying or installing anything.

Will a prompt always produce the right answer?

No. Check factual claims, calculations, citations and any output you plan to publish or act on. Use information you are permitted to share and keep consequential decisions under human review.

How do I download it?

Enter your first name and email in the download form. The guide is offered as a download and an emailed copy. No payment card is required. You can unsubscribe from subsequent emails.

Find out where you actually stand.

Grab the Field Guide, run the eight-dimension scorecard with the business in the room, and let the two lowest scores write your next two quarters. Free, takes a second, yours to keep.

free download

Get the Field Guide — free.

instant download + emailed copy · no spam · unsubscribe anytime

An independent guide from Prometheus Consulting, for data & analytics leaders, and the executive funding the work. Downloadable guide snapshot: June 2026. Landing page revised 2026-10-14. Check product instructions against current vendor documentation. Unofficial: product names are trademarks of their respective owners, none of whom reviewed or endorsed this guide.