Emotions aren't terrible

We're all human. Emotions are often useful indications that something is off, and you should try to treat them that way.

👋 Hello! I’m Robert, CPO of Hyperquery and former data scientist + analyst. Welcome to Win With Data, where we talk weekly about maximizing the impact of data. As always, find me on LinkedIn or Twitter — I’m always happy to chat. And if you enjoyed this post, I’d appreciate a follow/like/share. 🙂

---

Picture this: you work as a DS on a product team, and your product manager asks you to turn off default email subscriptions for new sign-ups. “It’s just a bad experience,” they say. You push back, citing the original experiment that led to launch of the feature. “20% lift in top-line conversions, 15% lift in revenue.” Your PM objects: “maybe these aren’t great users”, “those numbers seem inflated”. The PM’s arguments seem emotionally fraught, but of course, with cold, hard data on your side, you bludgeon them with reason until they relent. You walk away victorious, having bested another stakeholder with your rigor.

This has been me, and I imagine many of you as well. As technical folk, we tend to conflate emotional / intuitive with incorrect, but these are not always synonymous. “Emotions have no place in the face of truth,” you might say. And it makes sense, to some degree: I’ve written quite a fair bit about how we are the most appropriately positioned to bring objectivity to strategic discussions, for instance.

That said, while I still contend logical reasoning is critical to any reasonable discussion, emotional does not mean wrong — and this is a line I must admit I’ve been perilously close to crossing. Regardless of how objective I want decision-making criteria to be, this is often neither possible nor productive. We get angry/sad/frustrated because we care, and, however inarticulate, emotions and otherwise intuition-driven responses are often data — and it’s in our best interest (and the interest of our objectives) to consider them accordingly. So let’s talk about emotions, and discuss how to optimally navigate this biochemical world.

“Something like 90 to 95% of our decisions and behaviors are … shaped non-consciously by the emotional brain system.”
- Baba Shiv, in Think Fast Talk Smart

Emotions are not necessarily incorrect — they’re barometers.

Of course, emotions are manipulable and very painfully fallible, which we should be cognizant of, but emotional subjectivity is not inherently incorrect. Emotions are a mechanism of speedily filtering and synthesizing what’s important. An exceptional decision-maker can navigate this space well, avoiding the pitfalls of cognitive bias and synthesizing qualitative data into a well-formed, subjective map. Emotions are a barometer, and intuition is the deliverable.

… but sometimes they’re hiding intellectual dishonesty.

That said, the world of emotions is rife with hidden bias, and a universally helpful first step is to assess the nature of the bias, particularly with the aim of rooting out anything subversive. A useful way to bisect emotional arguments is as follows:

1. Those that shroud intellectual dishonesty.

2. Those that reflect something genuinely wrong.

If you’re dealing with the former, your best course of action is to try to help your counterparty recognize what they’re doing (or just move on with your life — sometimes arguing with folks who are playing their own covert game is futile). Fortunately, most of the stakeholders I’ve had to deal with are shooting for the same objective I am — maximizing impact to the business, and then, while untangling an emotional minefield might be difficult, there’s usually something important that someone cares about at the root of it.

And this might not be so simple as an emotional response to a product feature that seems sub-par. It could be that your stakeholders feel voiceless, or, for that matter, your direct reports or coworkers feel overworked, etc. Recognizing that genuine emotions are indicative of something to address can help you address the root cause of the emotional response, whether it's a product or a process failure.

How to deal with earnest emotions.

So how should we proceed? Let’s revisit the example I gave at the start of this post. Imagine if, instead of berating your PM for their senselessness, you explored the root of their anger. “I went through the onboarding experience, and the number of automatic emails I got was painful.” You push past the mixed bag of frustration and shame they felt about the product, and you discover that the concern is legitimate: your PM is acting under the supposition that brand is your strategic advantage, and the onslaught of onboarding emails they receive are degrading that — an effect you wouldn’t be able to see through any short-term experiment. So together, you figure out a plan to overhaul the email digests without destroying your team’s metrics. Your mutual open-mindedness brings you to a solution might not have otherwise discovered.

If you’re dealing with honest emotions, get in there and bring as much intellectual honesty as you can muster. Don't assume emotions are tainted bias and berate (even silently) others for having them. And from a stakeholder management perspective, don’t just present your work and expect your stakeholders to reorient all of their qualitative signals against yours. Recognize that the nature of their work is quite different from yours, and while yours may be more technically sound, theirs is likely more proximate to the decisions that matter, emotions and all. They might (read: surely will) know something you don’t.

And this is a posture that what we frequently fail to take — we should never force stakeholders to live in a world of where only quantifiable effects are tenable. Our analyses, our insights, our interpretations, and our data are all just that — data points in their tumultuous decision space. It’s not our position to sweep clean that chaos, but to help them understand how our insights can rotate that space.