# Score your own organisation's readiness
> **What you will learn**
> Not a chapter to read but one to **do**. Put **your own organisation** against
> the three failure modes from the last chapter, score it, and find the items you
> cannot produce evidence for.
## What you need
```
□ 40 minutes
□ A sheet of paper or a spreadsheet
□ Ideally two or more people from different departments
· alone, this ends at "we're fine"
□ One AI initiative under way or under consideration (if none, use "evaluating")
```
**Keep that third line.** The goal is not a good score — it is **finding the
points where departments answer differently.** A split answer is the result.
---
## Step 1 — Score the three (10 min)
Take the three straight from [how transformations
fail](/guide/ax-failure-modes). Score **separately** and compare afterwards.
| Failure mode | For us | Score |
|---|---|---|
| **1. Adopted without a purpose** | Can you say in one sentence what problem this solves? | |
| **2. No measure of effect** | Is it already decided what success is measured by? | |
| **3. Workflow left untouched** | Whose order of work changes after adoption? | |
```
Score: 2 = you can answer right now
1 = roughly exists, but not written down
0 = does not exist / people answer differently
```
**A total of 6 will mostly be cut down in step 2.** That is normal.
---
## Step 2 — Produce the evidence (15 min)
This is the point of the exercise. For each item, write down **where you would
look to find it.** Not a description — **a document name and a location.**
| Item | Where the evidence is | Did you open it? |
|---|---|---|
| The problem being solved | (e.g. March proposal, p.2) | □ |
| The success criterion | | □ |
| The workflow that changes | | □ |
```
Evidence rule:
✓ you can name a document and a page -> keep the score
△ "it's somewhere" / "it was in emails" -> drop to 1
✗ nothing -> 0
```
**"We covered it in a meeting" is not evidence.** Ask the same question of
someone who was not there and you get a different answer. Ask the person next to
you right now.
### What usually shows up here
**The success criterion reads "improve efficiency."** That is a wish, not a
target. [The failure with no measure](/guide/ax-failure-modes) starts exactly
like this.
**"The workflow that changes" is blank.** The most common one. If the tools grow
and the procedure does not, people experience it as **one more job on top of the
job they already had.**
---
## Step 3 — Take the lowest two (10 min)
Re-total the scores and keep only **the two lowest items.**
```
□ 1st ( ) — current score
□ 2nd ( ) — current score
```
Do not try to fix all three. **Two is the limit of what actually moves in thirty
days.**
### If departments scored differently
**That item is your first priority.** The split matters more than the score —
it means the delivery team and the people doing the work are holding different
pictures, and adopting in that state produces "why are we doing this again?"
three or four months later.
---
## Step 4 — Write the next thirty days (5 min)
For each of the top two, write **one line containing a person and a date.**
```
✗ "Define the success criterion" <- no subject
✓ "By 14 March, K pulls the current case volume
and writes the baseline as one line in the doc" <- person, date, artefact
```
| Rank | Who | By when | What is left behind |
|---|---|---|---|
| 1 | | | |
| 2 | | | |
**If "what is left behind" is empty, delete the line.** A plan that leaves no
document makes you hold the same meeting again next month.
---
## Self-check
```
□ You each scored the three items separately
□ At least one score dropped once evidence was demanded <- if none, you were generous
□ You identified an item where departments disagreed
□ The top two have a person and a date attached
```
If the second is unchecked, go back to step 2 and **actually open the
documents.** Answering "it exists" without opening it is the most common wrong
answer in this exercise.
---
## Check yourself
**1. Why is "we covered it in a meeting" not evidence?**
Answer
**Because someone who was not there answers differently.** What the organisation
shares is what was written down; the rest is each attendee's memory.
**2. Why does a split score become the first priority?**
Answer
**It means people hold different pictures.** That is worse than a low score — a
low score is known to everyone, while a split has each side proceeding in the
belief that it exists, until they collide months later.
**3. Why not fix all three at once?**
Answer
**Two is the limit of what actually moves in thirty days.** Take all three and
all three produce documents and nothing else. Finishing one, with evidence, makes
the next discussion far easier.
---
That is **why transform**. In the next part we look at **what to do first**, and
how to choose when there are several candidates →
[Three questions for choosing work](/guide/ax-choosing-tasks)