- 2-3
- Numbers worth watching
- 6 wks
- Typical engagement
- 1 in 4
- Questions we say cannot be answered
- 0
- Dashboards we will build
Most of what a dashboard shows you cannot be acted on
We work on the two or three numbers in a business that actually move a decision, and we will tell you when the honest answer is that your data cannot support the question you are asking.
Engagements
See everythingHow an engagement runs
-
1
The question
Written down in one sentence before anything is queried. If it cannot be written in one sentence it is not yet a question.
-
2
What the data can bear
Two weeks reading what you actually collect. About a quarter of engagements stop here because the data cannot answer the question, and we say so rather than producing something that looks like an answer.
-
3
The answer, with its error bars
How confident we are and why, in plain words. A number without an interval is an opinion with a decimal point.
-
4
Two numbers to watch
At the end you get two or three numbers worth watching and the query behind each, not a dashboard with forty tiles nobody opens.
We do not build dashboards
A dashboard is a way of deferring a decision about what matters to everybody who looks at it.
Forty tiles is not a measurement system, it is an inventory of everything that happened to be easy to plot. We will help you pick two or three numbers, tell you how they can be gamed, write down what should happen when each one moves, and hand over the queries. If you then want those on a screen, any tool will do it in an afternoon and you will not need us for it.
Who does the work
What clients say
They stopped in week two and charged for two weeks rather than six.
- Two numbers and the queries. Our last consultants left a dashboard nobody opened.
- Every figure came with a confidence interval and how it could be gamed.
- They genuinely will not build you a dashboard. Do not ask twice.
- Taught our analysts to do it and then left. That was the whole point.
- They published an analysis they got wrong and how it was caught.
- Expensive for six weeks. Cheaper than the year we spent on the wrong metric.
Before you engage us
Scope, tooling, data quality and what happens when the answer is no.
Will you build us a dashboard?
What if our data is a mess?
Which tools do you work in?
What do we get at the end?
Do you stay on afterwards?
Have you been wrong?
Write your question in one sentence and send it
If it takes a paragraph it is several questions and we will help you split it. The first conversation is free and about a quarter of them end with us saying your data cannot support it.
Method notes
EverythingTalk to us
Your question in one sentence and roughly what you collect. Two working days to an answer.
14 Trafalgar RoadKingston, Saint Andrew