Another Filter Early Opportunity: the Unfiltered Email


To:   all@yourcompany.com
From: bigprocess@yourcompany.com

ATTENTION ALL EMPLOYEES:
YOUR X-9 REPORT IS DUE BY CLOSE OF BUSINESS THIS FRIDAY. 
NO EXCEPTIONS! 
IT IS YOUR RESPONSIBILITY TO ENSURE THAT YOUR X-9 REPORT IS FILED ACCURATELY AND ON TIME.
Sincerely, etc., etc.
(Please disregard this email if you have already submitted your X-9 report.)

Have you ever gotten this email?

My favorite is when I have no idea how to even find out whether I’ve submitted my X‑9 report or not. I might or might not even remember what an X‑9 is.

An email like this has a bigger impact than you might have expected. Let’s imagine that the “bigprocess” department sends it to 1,000 people, and that it takes an employee 15 minutes to become enough of a temporary X‑9 expert to even answer the question, “Have I submitted my X‑9 report?” I’ve seen actual numbers that were way worse than these.

The cost of such an email would then be 1,000 person × 0.25 hour/person = 250 hours of the company’s time. That’s 31.25 days. And that’s just to calculate the recipient list.

The problem with a letter like this—with the fine-print if statement embedded into it—is that it exhibits the filter late antipattern: it computes recipient list later than it should. The “bigprocess” department that sent this email should have done their job a little differently:

What they did (BAD)What they should have done (BETTER)
1. Send the email to everyone in the company.1. Calculate the list of people for whom the email is relevant.
2. Let each employee figure out whether the email is relevant or not.2. Send the email only to those people.
The “bigprocess” department should not have sent the email to people for whom the email is irrelevant.

Chances are, the people in “bigprocess” know a lot more about X‑9 reports than you do, so they might not need long to figure out who has submitted the report and who has not. But even if it took them 30 FTE-days to compute the recipient list, it would still cost the company less than distributing the calculation to each of 1,000 employees.

Behind my math here is the implicit assumption that one FTE hour of “bigprocess” time has the same value as one FTE hour of time from your managers, leaders, and domain experts. An hour of your recipients’ time may be more valuable than an hour of your “bigprocess” team’s time.

It gets worse.

When a knowledge worker is interrupted, it doesn’t just cost the duration it takes to do the new thing, there’s a multi-minute penalty that you have to pay to get back into the flow of the task you were thinking before you were interrupted. So, a “15-minute interruption” may well cost you closer to half an hour.

There’s a reason that “bigprocess” departments all over the world sends email messages like this. The people in those departments probably don’t have the technology resources to (1) write a program to calculate the right sender list or (2) send an email to a filtered email list. They may well be doing the best with what they have.

But if this kind of an email comes at people every day, week, or month, it’s almost certainly worth the effort to figure out a solution to the problem. Frankly, if people are getting these more than about twice a year, they’re probably deleting them as soon as they see who they’re from.


If you’re interested in optimization—filtering late and interruptions and political problems and such—you can read more about them in my new book, How to Make Things Faster.


Posted

in

,

by

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *