This week we are talking about performance reviews with Alen Peacock, co-founder of Stretch Metrics.

Following the interview last week with Radina Nedyalkova about how to best approach talent acquisition, performance management is a great next topic to cover.

Alen is both a serial startup founder and an Engineering Lead, exposed to the ups and downs of setting up and managing high-performing teams at scale. With his latest venture Stretch Metrics, Alen is tackling the hard problem of creating a better framework to evaluate employees fairly by making the full process more natural.

In this interview we covered the challenges with the existing performance review frameworks, and how this process can be improved, the role of data as part of this process and how to better leverage it, how to best engage your team and the role of peer reviews in educating the overall evaluation.

Alen shared a ton of insights that will definitely provide you with an interesting perspective and hopefully help re-think your existing approach. Thanks, Alen!

Alen, could you tell us a bit about your background?

I started writing software as a kid and followed that passion through two degrees in Computer Science. I spent a few years after college working for MIT’s Lincoln Laboratory, and also spent some time working on tech in the financial sector, for a Fortune 500, and for a startup that was acquired before co-founding an ambitious tech company myself and raising funds from Google Ventures. The company we started was acquired in 2014, after which I spent a few years leading large engineering teams. Through it all, I learned a lot about working with really great people. The crucible of starting a new company in particular forged a much deeper understanding of how to build and maintain high-performing teams.

From your perspective, what’s the purpose of a performance review? And what are the benefits for both the employees and the company?

Learning to give and receive effective, high-quality feedback might be the greatest superpower there exists for driving teams towards excellence. There is no way to improve at anything without some sort of feedback loop. And getting better at things is immensely rewarding – it’s the reason people golf or practice the guitar. As Dan Pink famously points out, mastery is one of the primary drivers in the science of motivation. Performance reviews provide an opportunity for people to do just that, but often fall so far short of that goal that they veer into demotivation. While companies tout performance review process as a way to help develop individuals, in practice they serve mainly to support yearly compensation adjustments, promotions, demotions, justify firings, and to create inputs into workforce expansions and reductions.

Your goal with Stretch Metrics is to create a better framework for companies to conduct performance reviews. What’s wrong with the existing approaches and what are you trying to do differently?

The first problem with traditional performance reviews is that they are a time-consuming distraction from other high-priority work. The second problem stems from that first one: because of the distraction they create, individuals and companies can’t afford to do them very often. Combined, these problems spin out into many others: performance reviews end up tainted by recency bias, are soon forgotten, and don’t provide the continuous feedback necessary to truly build individual mastery. Taken together, these deficits create a credibility problem for traditional approaches.

Many performance processes are manager-driven, with the only feedback delivered via self-assessment and manager-assessment. 360-degree review processes gained popularity to address this shortcoming, but they also multiplied the amount of time and work, requiring everyone to complete even more assessments.

Due to these problems, traditional performance reviews are almost universally loathed and many companies (including 1/3rd of the Fortune 500) have abandoned the annual performance review process completely.

Alen and his co-founder Clint Gordon-Carroll

We live in an era where data is over hyped. Many companies claim to be data driven but track vanity metrics or fail to act on the data they collect. In regards to performance management, how much data is enough data, and what’s the best way to leverage it to drive value?

Data is King, but collected data that is never consulted is a powerless despot. Naval Ravikant observed:

Shorter feedback loops means more iterations, and it’s the number of iterations, not the number of hours, that drives learning.

So collecting and delivering performance management data in a timely manner – daily, all year long – really unlocks data’s potential for growth and learning. This observation seems to have been missed by much of the performance management industry, and by many standard performance management processes.

We’ve discovered that the data collected about perceived performance on a team, when collected often enough as is done by Stretch, provides insights beyond just performance. Currently, we surface insights to managers about individual engagement, flight risk, conflict hotspots, and communication in addition to core indicators of excellence or declining performance. These insights give managers new super powers to get on top of issues before they fester, reward behaviors in the moment, and keep their teams happy and motivated. Individuals gain continuous insight into how their own performance is viewed which creates a new kind of self-awareness. They also gain the ability to collectively make their own manager’s performance visible to the larger organization, both to hold them accountable and reward them.

The “decay rate” of insights gained from performance data is high. An indicator that surfaces a risk for a high performer to begin searching for a new job elsewhere must give a manager a timely nudge. If the underlying data used to create these indicators were only gathered yearly, quarterly, or even monthly, it would surface too late to take meaningful action.

In my experience, the level of engagement with ‘survey’ tools can be very inconsistent. For example, I noticed that Engineering teams tend to engage more if compared to say Sales or Marketing teams. Do you think this is a technology problem? And, if engagement is key to quality results, how do we facilitate adoption?

Your experience is typical. Getting people to answer surveys can feel like bathing a cat, and getting people to do traditional performance reviews often begins to feel truly sisyphean after the third or fourth email reminder goes out.

Our initial thesis at Stretch revolved around the idea that annual performance reviews could be displaced by a quick experience, done daily or even more frequently, right as interactions with colleagues happen. We knew this was a radical approach and we’d need to prove, first, that people would engage with the system at a high rate. So that was the first thing we focused on. It’s an insanely hard problem, as anyone who tries to increase engagement in an app can tell you. So we started at a place that we knew would provide the most fertile soil: inside of workflows and tools people already use rather than with yet another app or web page. And we focused on making it extremely quick and simple. And then we sat down with our users from the very beginning to ask them questions like “how distracting is this?”, “were you ever bothered by a request from Stretch?”, “what did you like best/least about this?”, etc., and iterated with small tweaks until we achieved greater than 70% weekly engagement, across our entire user base. HR tools in general are notoriously bad at this, typically focusing on the experience of the HR professional to the detriment of the individual contributor. Stretch makes HR the hero instead of the goat, by producing an experience that individual contributors will embrace and use, without prodding from their managers or HR leaders.

Technology alone cannot increase engagement – you have to start with keen observations about human behavior and then study that behavior in tandem with the development of technology.

An interesting supplement to the classic performance reviews is the feedback collected from peer co-workers. Having those who work with you provide insights around your performance can add valuable context. Considering that people build personal connections at work and, especially in smaller teams, you don’t have many peers with a similar skill set able to provide an unbiased evaluation of your competences, is this process really fair and beneficial?

The truth about performance management – no matter what system or process is used – is that everyone ultimately measures the perception of performance rather than performance itself. This is inescapable.

Getting a reading on that perception from everyone you work with increases the accuracy of those perceptions, rather than relying only on your manager, or a small handful of people chosen to help give a 360-degree review. We’ve almost all encountered individuals who are rewarded by an organization for managing-up well and managing-down terribly, or managing-sideways well but neglecting their subordinates, or by being adept at understanding the political power structure of a team or organization and playing nicely only with those in power. Many of us have seen truly great managers, beloved by their teams, punished or looked over because they didn’t manage-up well. All of these issues can be mitigated by collecting feedback from all sides, to help organizations recognize greatness and incentivize good work.

What we’ll continue to see in the workplace is a democratization of structures and processes, and that extends to performance management. A manager-led review process is tainted by the narrowness of experience and interaction between one subordinate and one manager. If we wanted the most accurate picture of how a person impacts an organization, we would gather input from every person and team they interact with. But such an encompassing perspective is nearly impossible with traditional performance review and evaluation processes due to the inordinate time and distraction that would require. In order to get a truly comprehensive understanding of how your performance is viewed, you’d need a tool that collects that view from all the people you work with, over regular intervals, throughout the year, without distracting those people – and that’s the system we think we’ve hit on at Stretch.

Advocating for transparency seems to be a growing trend. When talking about personal performance, what’s the impact of a radically candid approach?

Radical candor can be transformational, but it takes some getting used to. We’ve all experienced receiving really good constructive feedback after not receiving it for a long time. This could be the first time you had a really good writing instructor in school who red-lined more than you were used to, the first time you had a coach with truly high expectations in sports, or the first time you had a really effective mentor in the workplace. And that candid feedback usually stings and is uncomfortable at first. But after a short while, you not only acclimate to it but learn to crave it, because you learn that it makes you better. The same is true for having constant insight into how your own performance is perceived by others. It can be uncomfortable at first; we all have times when we perform more poorly than we are capable of. But you soon realize that understanding how your colleagues view your performance is liberating, because it does unlock higher achievement.

Unfortunately what we’ve found is that not only do most organizations not practice radical candor, they don’t even have good habits of regular feedback. Unleashing the power of radical candor requires people to take that first step of giving and receiving feedback regularly. Until your people are doing that, you have no hope of those people becoming experts at feedback.

We tend nowadays to rely on software and products more and more. Sometimes even at risk of dehumanising basic interactions. In cases where we for example measure performance via insights or score cards, without offering full context around it, what are the most negative effects you've witnessed? Or do you see these as a positive way forward if properly administered?

Context always matters. In the very earliest alpha versions of Stretch, we were focused on proving core hypotheses around engagement and accuracy of the system and hadn’t yet built mechanisms for people to provide richer context, and we found our early users begging for that capability. People wanted to recognize true excellence with a few words of description around that work. That’s also true when people notice lagging performance. And on the receiving end, performance data is meaningless without some concrete directional information to act on. I have a close friend who once received the feedback, “He’s rogue” in response to a 360-degree evaluation question. That was not enough feedback for him to take action to fix whatever problem the evaluator was hinting at – was he “rogue” because he ignored input from others? Because he didn’t show up to meetings? Because he challenged other perspectives too often? Is being “rogue” even inherently bad or good? Giving good context is the solution, and that usually involves describing behaviors instead of traits, giving specific examples instead of generalizations, and making sure that suggestions are actionable. Notice that none of those things can be done automatically by software or AI – they all require a sentient and empathetic human being.

One of the tricks we use to keep engagement high within Stretch is to make the quick daily reviews anonymous, so that people don’t have to worry about political consequences of telling someone they did exceptionally well recently vs. telling them their efforts seem to have fallen recently. And all reviews for a person are averaged together, so that the overall impact of a single rating (negative or positive) is absorbed and anonymized when delivered. But when people give additional context via a comment that will be delivered to that person, we make it clear that these comments are attributed and begin a conversation between the two people. This negates toxic or cancerous behaviors that can come from someone trying to anonymously lob passive-aggressive bombs toward someone they had a bad interaction with or simply don’t take the time to form feedback carefully. Partly because detailed feedback is attributed instead of anonymous, positive comments tend to dominate and critical comments tend to be delivered with great care -- both behaviors that the science on feedback say are essential for personal growth.

Are there any resources on the subject that would you recommend our readers to check out?

There is so much great research out there on these subjects. Some of my favorites:

Where can people connect or find out more about you?

You can find me on Twitter @alenp, and you can find Stretch at https://stretchmetrics.com or in the Slack App Directory


If you have experience revolving around startups, people and culture, ping us on Twitter @HumaansHQ or drop me an email at giovanni@humaans.io. We'd love to learn from your journey and share your learnings.


Humaans is a smart employee database created to simplify your people operations. Collect, store and leverage your people data to make your business better. Get started today – create an account in minutes.