Agencies are groups of people, people are organisms, and organisms do what they’re rewarded for doing.

Most observers would suggest that agencies are rewarded for winning awards for their creativity.

An agency is a business, a business requires revenue, and awards bring clients — which provide revenue.

But agencies aren’t rewarded for winning awards for their creativity. Agencies are rewarded for making the best decisions about being creative.

If that process of making the best decisions about being creative leads to awards? Great.

But awards aren’t the point.

The point is the way an agency organizes itself to reliably and consistently produce great creative.

The same goes for big client wins.

The same goes for headline-generating hires.

The same goes for creating viral videos that land your agency in AdAge, or Campaign, or Ain’t Advertising Swell I Feel Pretty dot.com.

These are all wonderful results and should be applauded. But they are only necessary conditions for long-term success. They are not by themselves sufficient.

Process, not outcome, is what great agencies are rewarded for doing.

And if you don’t believe that’s true, then that’s why your agency doesn’t do postmortems after you fuck something up.

Actually, let’s not call them postmortems

A postmortem is the examination of a body to determine the cause of death.

The very name is quite grim, isn’t it?

It refers to a certain cadaverous conclusion.

It assumes your project has experienced a Mortal Kombat-like fatality, and that an investigation will determine whodunit.

How about something more forgiving? More encompassing. More appreciative.

Something that doesn’t assume an error has been made. Something less CSI, more Joy of Painting.

How about we use the term “retrospective”?

And let’s not just do them after a dramatic whoopsie

Here’s the thing about retrospectives: they’re easiest to perform after a dramatic whoopsie.

In the moments after a dramatic whoopsie everybody tends to look around, brow furrowed, mouthing the words “what the f!*k just happened?”

In those moments, everybody is in the mood to reflect.

But absent one of those impressively terrible clangers, we say, “We’re too busy.”

We say, “We have more problems to solve.”

And isn’t that a convenient response!

Reflecting on one’s idiocy is painful, and there are very few people, besides the Marquis de Sade and devotees of acid jazz, who enjoy pain.

But it’s precisely in those moments — when something has not, in fact, gone dramatically and irrevocably wrong — that retrospectives are even more useful.

Why? Because retrospectives are a form of “double loop learning”

The first time you make a decision you are engaging in single-loop learning.

You do a thing, or you don’t, and you’re rewarded for it, or you’re not. You learn from the results of that decision. You try again.

But if you question why or how you did a thing — if you honestly self-assess your assumptions and methods of action — you shift into double-loop learning.

That’s what a retrospective is. It’s not just acknowledging what went wrong or right. It’s questioning the assumptions that led you to that result. And that is a healthy exercise, regardless of whether a project has ended.

By reframing postmortems as retrospectives, you open the possibility of using them at any opportune moment — the conclusion of a project phase, the beginning of a new quarter, the end of the year, whenever.

Or as my people, the Catholics, say, there’s never a bad time to criticize yourself. Peace be with you.

Of course, retrospectives are difficult because they hurt

Reflecting on your errors has very visible short-term costs.

Reflection takes time. And because the future is not visible today, slowing down today to go faster at an indeterminate future point seems like a bad gamble (a cognitive error known as time discounting).

Reflection is also potentially embarrassing. People resist learning from fear of being seen as incompetent.

Happily enough, here’s the thing: we’re all incompetent.

It’s only realizing the reality of our own incompetence that’s tough.

That’s why successful retrospectives establish a no-blame atmosphere of learning.

They eschew defensiveness.

They do away with justification.

Instead, they collect and analyze data. They ask what conclusions we can draw from our experience. They ask how can we test those conclusions, and they ask what evidence we need to prove they’re correct.

And they don’t blame individuals.

Instead: they improve the processes that individuals use, which in turn allows the individuals to be more successful.

And that’s the real benefit of a retrospective.

It creates the opportunity for success.

It creates the possibility of making better creative decisions.

Which, if you’re playing the long game, is what agencies are rewarded for doing.

Go ahead and steal this retrospective

We use the following documents to organize our retrospectives. I hope they’ll be as useful to you as they have been for us.

Suggestions for a better process? Have a framework you prefer to use, or that you’ve seen work? Leave them in the comments; we’d love to learn from you.

The Survey

Step 1. Sent to all employees who participated in the project. (Clicking this link will prompt you to make a copy of the survey)

Findings Report

Step 2. Assign a researcher or strategist (who was not part of the project) to read and synthesize the survey results.

Findings Deck

Step 3. Collate the report findings into a deck. Use this deck to present to the team and guide discussion about the project.

Remedies Spreadsheet

Step 4. After presenting the retrospective findings to the team, use a copy of this spreadsheet to suggest remedies for problematic situations. These remedies should be approved and socialized with the appropriate teams.

Fin