11 Jan 2020

Five Seconds Could Change Everything

This post also doubles as my 2019 lessons post. To check out my 2018 lessons post (this practice is now two years old), see EV is Everything. I seem to like the word everything.

I have a buddy who always gets on my back when I order the same beer as him. I’ve now heard his rationale probably dozens of times - ordering the same thing as someone else is always a failure to take the five seconds to think about what you actually want. Have you looked at the menu? Have you evaluated the possibilities?

I know my own rationale for copying beer orders. I use the orderer as a proxy for good taste, and in doing so eliminate having to think about what good taste is. In the past I used to believe that relying on proxies like this was a great way to save time. I now believe decision quality is jeopardized by taking these shortcuts.

The beer example is inconsequential, so I want to relate this idea to every decision. Many of these decisions are how to respond in a conversation with one or multiple people. The how can be not responding; it can be asking a question; it can be disagreeing, agreeing, proposing an alternative, whatever. These responses can change the direction of the conversation and any resulting action from it.

In my day-to-day, most conversations of consequence involve engineering decisions (abstract into a general service this sprint or do inline and kick off into the future?), stock market or DFS actions (buy or sell stock or player?) or personal planning (go rock climbing or run?). Because I hold a stake in the final decision for each of these, my responses influence the future. Knowing this is extremely motivating, but that’s the subject for another post.

Conversations involve a lot of words, so there’s often dozens of micro-decisions to be made every few seconds. Especially in the case of information-heavy conversations, there are lots of decisions about which questions to ask. The internet is a thing, so many questions are a waste of time when they could be googled (ex. “How has the stock traded the last few months?”). Similarly, information-heavy conversations may blend fact and opinion, so I believe strongly opinions must be questioned when appropriate:

• Is the stock actually trading at a real discount to book? Is there a case to be made that book value is overstated?
• Why does this local service need to be changed to a remote service? You say that there is a generalized use case for it based on what the business wants to do, but how sure are you of that? Give some examples.
• If we go rock climbing at 4 instead of noon, how crowded do you estimate the bouldering walls will be? How much value do we get if we go at a slightly earlier time?

These questions I think can be difficult to ask because they force people to analyze their own assumptions, which suggests you as the questioner do not trust the questionee. That suggestion is real. Many people who ask these types of questions are considered by most people to be hardasses and potentially difficult to work with. The credit officer I worked with in my first banking job was famous or notorious depending on how you frame it for these types of asks. There is a price to be paid for asking these questions.

There is a much larger price to be paid for not asking these questions. Unlike a slightly uncomfortable interaction’s impact on a human relationship, code in production, a buy or sell in the market and waiting in line are impossible to reverse. Yet, it’s easier to not ask them because of the social niceties of the world we live in. Some management norms even argue asking too many questions is a sign of a bad manager that doesn’t trust their people.

Tradeoffs like this one must be weighed in conversational responses and really any decision, which is why I’m advocating for adding five seconds to what you think would be a normal response time given a complex problem. A seemingly easy problem may be complex given the right analysis, and vice versa.

The five-second rule feels even more relevant as weighty decisions move to Slack and text message from e-mail (where in my opinion it’s easier to take your time replying). Slack especially is a problem here as you can see other people typing, which no doubt impacts how conversations go down and your own response times in conversation. Social expectations (ex. in Slack in a real-time conversation, I should respond in <30 seconds) should not impact how you get to an optimal decision.

The five seconds here is somewhat arbitrary. I’m finding it’s just the right amount of time to either confirm a decision or stir a longer period of considered thought. If nothing else, five seconds reminds me of the consequences of not asking an important question or consenting to do something that may have huge ramifications on the future (that future could be one hour, one day or a lifetime).

The five second rule in 2020 is a request to myself to not let the instant response culture dictate how I respond. In the words of Stephen Covey:

“In the space between stimulus (what happens) and how we respond, lies our freedom to choose.”