Climbing off the CTO ladder (Before We Fall Off)
Early thinking about career development (2014 repost)
Almost 10 years ago, faced with the passing of my father and my journey into more senior leadership roles, I started my thinking about career ladders and career development frameworks. The thinking in this post from 2014 helped influence the definition of the career ladder at Spotify. While the specific details are different some of the core elements of non-linear growth and multiple vectors of growth persisted.
I am reposting it here in its original form, to offer a glimpse into my prior thinking, and encourage me to capture some updated thinking.
I have recently been thinking about my career development path and comparing that to my father who spent 40 years working as a machine builder in the automotive sector. Through this lens I have started to rethink how we view career development within the tech sector.
We’ve spent time and resources coming up with novel ways of organizing our companies - from lattice networks and holocracies to Spotify’s own tribe/chapter model. While specific organizational structures are important, it neglects to take into account career development. If we are serious about building great teams, we need to spend time on this area as well; instead, most organizations have adopted a variation of the “white collar model” and assume it works well. I argue that it does not (nor do I believe it works well anywhere).
A ladder
The most common model is the classic career ladder which typically has engineers becoming more senior, eventually becoming managers of other engineers and then ultimately directors and VPs. The attraction is the sheer simplicity. It sets up clear expectations and promotion criteria and is easy to describe to people. Because of this it is easier for the manager to mentor, get resources and make decisions about promotions and compensation. However, it is important to remember that this is optimized for simplicity of execution and not the career development of people. This should be reversed, we need systems optimized for helping people develop because there is a cost to this simplicity.
Typically it sets up patterns to promote people out of the jobs that they are performing well.
Predefined rungs on the ladder limit growth for those uninterested in that explicit narrow path.
Because of limited management positions, it sets up a funnel that will eject people who have no room to grow.
Experimentation is difficult since trying a new role and failing results in a demotion.
Explicitly sets up a “promotion” model as opposed to viewing it as a role change.
Comparing this career progression to my father’s path, who spent his entire career without ever being forced into management (although he did have opportunities), caused me to distrust this model. Although he never climbed a ladder, he was clearly getting better at his job, adding value to his company, and making the people around him more productive. If a single ladder system had been in place he likely would have been pressured into being a less valuable employee since he would have been pushed out of the role he enjoyed and excelled at, into one which he did not enjoy. So, why do we continue to use an antiquated idea, that may never have fit in the first place, to define career growth? Simple, because it’s easy.
A second ladder
To help mitigate the challenges of a ladder, many companies have added a second ladder that focuses on technology roles as opposed to management roles. This second track is an iteration on the junior/senior/staff/principal modifier used for a long time within tech, but instead of it building to a manager role as above, it allows for a parallel track that stays focused on technology. No longer does the engineer need to make a role change to continue their career development. A good example of this working is Google’s 9 steps (in tech) that promote technical achievements. This helps by removing the funnel of people exiting the company as well as removes promotions to positions where skills and interests are not matched. It does this while maintaining much of the simplicity of the ladder system. However, by itself there are still some holes:
Continues to assume the only way to add value to the company is through the combined effect of expanded responsibility, influence and technical depth as opposed to rewarding growth in some of those areas.
Experimenting with different roles is still challenging because moving “down” the ladder is seen as a negative thing.
In some cases this can also reinforce negative impulses and allow people to deflect responsibility because of their “level” and sets up hierarchies where those may not be beneficial.
Usually not as clear as originally developed and you may end up with systems with unclarity.
Spotify has recently started thinking about the non-management career track and we are working to see if there is something we would like to adopt. We’ve had feedback from engineers and managers alike that we need to add a tech ladder. I am not convinced that adding a ladder fixes everything, the real issue is that we are not communicating how to grow without a ladder. The second ladder is a blunt instrument that can easily add some clarity to this, but it is not a complete solution. Part of management is optimizing for the productivity of the team not our own work streams. This should not be taken lightly as it will have an effect on the organization’s culture. Understanding this may help invest in a technology ladder that works for the culture of your organization.
What do people look for?
I believe that in reality people are looking for a few core things when trying to get promoted.
Compensation: The ability to be rewarded for our efforts
Autonomy: The ability to exert control over our work
Mastery: Developing a deeper understanding of our craft and continuously improve
Recognition: The recognition that we are adding value to the company
Ladder based models are optimized for #1 and #4. Compensation is actually one of the easier ones to solve, because an organization is willing to pay a certain amount for their staff. Breaking this into bands may make it easier for managers, but the real issue is setting up transparent pay ranges so people know how they are being rewarded. Recognition can be more difficult to solve but can also be done through official announcements, one on ones, as well as more ad-hoc methods. More difficult for sure, but not unsolvable.
Instead I think concentrating on mastery and autonomy will yield better results. Promotions become a reward for mastery but usually are not well supported by the company and they also mean people may get promoted out of their area of mastery/expertise. It also can remove some autonomy since it sets up a narrow path of advancement (the fact that we call it “advancement” is part of the issue) and thus does not allow people to decide how to grow, it may actually dictate it. For the individual this limits their growth and for the company it creates a homogenous group of senior talent that excel in the specific framework.
Lest I think my father and his industry had it figured all out, these are some of the same challenges he faced. While he continued to get better every day and was rewarded for it, much of this was self led with little support from his organization. I am astounded how similar my experiences are to my father’s, different jobs, expertise and geography but some things remain common.
There must be more!
At Spotify we have traditionally not had a structured ladder. We opted for a system where roles/titles are defined by institutional need not career advancement. An engineer and a manager both add value, and both are important to the mission. Changing roles should not be see as a promotion, but a new challenge. One that will add different value to the company but in of itself does not indicate more impact. Removing this “upward” movement as the only career development path is imperative to real growth. It’s the long game where ladders look to the short game. This seemed to work quite well when the company was small and served us well for many years. However, we now have over 600 engineers and have begun looking for some structure to this without losing our culture of innovation and autonomy.
We began by looking at a second career ladder as mentioned above, but found that the structure that imposed was too limiting. Our first attempt at a flexible and non-linear structure was add-ons. Add-ons are semi-structured ways of adding value to individual and the business. Specific add-ons are defined such as “Agile”, “Speaker”, etc. that define the scope, steps and help needed to achieve it. They set up clear acceptance criteria and skills needed to earn the add-on. They do not change the fundamentals of your job, rather they aim at exploring other ways for people to make an impact. This was a compelling idea that removed the linearity but had some key challenges. First, it focuses on new skills/techniques that may not always be related to what you are using. This sets up a system where the reward is the add-on as opposed to rewarding the use of the skills/techniques learned. Also, it may have been too free-form because it was entirely driven by the individual as opposed to a collaboration between the individual and the mission. Finally, we may not have put enough institutional backing to allow for future growth of the model, and the rollout suffered.
But it did get me thinking....
Originally when I began thinking about my father and his journey I thought add-ons were the way. It was one of those things that attracted me to work at Spotify. I even ventured out to get others excited about it. But as we set out this summer to explore ways of providing guidance to career development I began to think of another idea. Not to say add-ons do not have merit, I still believe in them, I just think we need something more.
My current thinking is that we need something with multiple paths and possible areas of growth. With that in mind we brainstormed what types of things do we look for in those that are making more impact in their roles. We decided on 8 different goals that tried to get at the heart of what the individual is looking for was well as what things make the most impact on our organization (and reflect our culture). This covers mastery, collaboration and reaching to solve the teams mission. It values what we value here at Spotify, aligned autonomy to achieve our goals, build great products and move quickly. In other words the categories reinforce our what we push for in our culture. To achieve the recognition that people are looking for as well as tools for managers to help people grow we then visualised three “steps” that define the impact someone is having in each category. This includes individual, team, tribe and finally company wide impact. Behind each one of these cross sections we are building concrete examples of ways people can exhibit this. Achieving a step does not indicate things you are allowed to do, rather it is a reflection of what things you are doing and the expectation that you continue to do them.
This idea is still in its infancy, it’s still growing, changing and moving forward. For instance, we still have not figured out how this may impact compensation. One thought is to add in aggregate steps to show overall impact, and also align with more standard HR policies around “advancement”. I am not convinced this will be needed but remain open minded to where this goes. If we find things missing we will add them, if we find things that are not working we will replace them and things that do work we will make them better. If my first year at Spotify has taught me anything, it’s that the best solutions are ones that morph, adapt and change to meet the new challenges.
I have outlined one idea we are trying at Spotify, it’s an experiment, and it may not work out. We’ve all built innovative products, now is the time to innovate our career development before we all fall off the ladder. I think I owe it to my father to find a way to be as happy with my career development as he was with his own and I am thankful to have an employer that is supporting me in this discovery. It’s no small task, but I think we’re up to the challenge.