Issue 01, The Valuable Part, is on its way. Get it first, free.

We used to direct the tools. Now we direct the analysis.

For analysts working through the AI shift. Building the chart got easy. The analysis is where the value went. Tool-agnostic notes on how to think through a project, direct the work, and spend your time on the part that actually matters.

Issue 01 · the inaugural essay

The Valuable Part

The essay the whole publication grows out of. It opens on a confession, reframes the old job as assembly mistaken for analysis, names the automation fear honestly, and then turns it. Tool-agnostic, plain-spoken, no pitch.

Get Issue 01 when it lands
"I am not a coder."

The machine came for the assembly. It did not come for the thinking. It raised the stakes on the thinking and removed the hiding place.

From The Valuable Part

Viz is the easy part. The analysis is the valuable part.

For twenty years an analyst's hours went to the mechanics. AI collapsed the assembly. The value did not vanish. It moved back to the part we always rushed past.

The publication

Free weekly essays on how to think through a project before touching anything, decide what is worth building, and read what the data actually says. Written for the working analyst, not the beginner.

Subscribe free

The method

How I turn a vague question into a solved problem, packaged. Finding the real question, directing the AI, catching it when it is confidently wrong, handing back something the business can trust.

See the positioning

The guide

Fifteen-plus years of analysis, thousands of dashboards, and an origin in Excel line graphs that still make me cringe. I made the expensive mistakes so you can skip them.

About Eric

What's coming next

Then · where the hours went
Cleaning and joining the datamost of the week
Building, then rebuilding, the charttwo days
Framing the real questionwhatever was left
Now · where the value went
Cleaning and joining the dataDIRECTED
Building, then rebuilding, the chartMINUTES
Framing the real questionTHE JOB
Essay 02

Where the time should go now that assembly is cheap

The thing you spend your time on is the thing that is actually valuable. A worked look at an ordinary analyst week, before and after the shift, and what to do with the hours you got back.

Get it in your inbox
Essay 03

How I think through a project before touching anything

Follow along
Essay 04

What to pay for and what to refuse

Follow along
Worked example

A real project with the reasoning left in

Follow along

Building the chart got easy. Here is where the work went.

Four things the tools cannot do for you, in whatever tool you happen to use. Every essay here serves one of them.

Frame the real question

The question you were handed is rarely the one worth answering. Finding the real one is most of the job now.

How I find it

Decide what is worth building

Cheap assembly means everything can be built. That is the trap. Choosing what to throw out is where judgment shows.

How I choose

Direct the build

Tell the AI what to make, then catch it when it is confidently wrong. You cannot direct what you do not understand.

How I direct

Hand back something trusted

Say what the data knows and what it does not. The honesty around the chart is the craft the business pays for.

Why it matters
The moat
You can't direct what you don't understand.

AI did not remove the need to know analysis. It raised the stakes and removed the hiding place. The analyst who can think is the one AI makes more valuable, not less.

The method · in development

This is how I turn a vague question into a solved problem.

It is not a course about a tool. Tools change. This is how I think. How I find the real question under the one I was handed. How I decide what is worth building and what to throw out. How I direct the AI to do the work, and how I catch it when it is confidently wrong.

It is for the working analyst in the AI era. Not the beginner trying to break in. You are already in, and you have decided to move up instead of out.

I am building it in the open with a small founding cohort. Subscribers hear first.

Hear when the cohort opens

Who is writing this

Eric Summers. Analyst first, builder second. Directed Analysis is a Directed Building property, tied to no product, and it never will be.

See the data stories at ericsummers.io
15+

Years analyzing data for people who had to act on it

LinkedIn
1,000s

Dashboards built or improved, across industries and tools

Selected work
Excel

Where it started, with homely line graphs that make me cringe now

The confession
Daily

Building in public at TableauOps, so the method gets tested every day

TableauOps

Move up the value chain instead of out of the job

One essay a week on the thinking the tools cannot do for you. No tool tutorials, no pitch, unsubscribe any time.

Subscribe, free

AI isn't coming for the analyst who can think.