At OLX I send out a weekly update email every Friday. It is an attempt to be transparent about what I’ve been working on that week and helps me reflect. Sometimes I also muse a bit on more broader topics I’ve been thinking about, or books I’ve read. This is an excerpt of this Friday’s email, I thought it may be interesting for people outside of the company as well.
Last weekend I read Carol Sanford’s “No More Feedback”. It’s an intriguing book, and resulted in a “falling of my faith” type of situation. She challenges many “core beliefs” that I hold (or have been taught to hold). As a means for me to process, and to perhaps prepare you a bit for what you’d find in this book if you were to decide to read it, here’s my summary of its ideas.
Disclaimer: It is likely that if you actually understand the summary below it will leave you in ambiguous and confused state, and this may not be personally beneficial to you. If you’re not up for that: stop reading, have a good weekend!
Disclaimer 2: These are the ideas described in the book, the fact I repeat them here do not mean this my or OLX’s official standpoint, nor that they’re particularly compatible with our current practices. Take them as ideas, part of my “challenge everything” philosophy. It’s good to sometimes re-evaluate your core beliefs.
Here we go.
The basic premise of the book is: feedback is toxic. Feedback mechanisms come from systems theory, and are perfectly valid there, the idea that this concept has any place in systems that involve humans is flawed, unless we feel that humans are yet another part of the “system” to control, whip into shape and optimize in terms of efficiency.
Feedback is based on the premise that we need other people to tell us if we are doing well (and what “well” means), because we cannot assess this successfully ourselves. And, as it turns out, other people are really bad at it: cognitive biases, self interest, judgement with limited context, violent communication. It’s a mess. Think of your last performance review, did you feel it was fully fair and helpful? Chances are you at least have mixed feelings. Yes, we can work very hard to improve feedback, but it takes more and more of people’s time, lots of training and in the end people still feel bad and the results are questionable. The book describes a particular case of one company introducing performance reviews. Over a decade worth of iterations, more and more time was invested in it, up to 7% of total time of managers until they gave up, dropped the process, and saw no change in terms of performance. Let me repeat my disclaimer: this is what the book is saying, not a recommendation.
“So,” Sanford says, “feedback: bad.” No performance reviews, no peer feedback, no “good job”, no acknowledgements, no “look at this person, he/she’s an example for all.” All these things do more harm than good.
So, you will ask: what is the alternative? Self reflection. People are, or can learn to be, much better judges of their own performance than other people.
“But people can’t do this at all!” you’ll say. “Look at Trump!” No politics please, thank you.
Yes, Carol Sanford’s argument goes, because over the course of their lives people have learned to largely rely on external feedback to feel valuable. They’re used to be whipped into (predefined) shapes with feedback: good boy, bad boy, more of this, less of that. As a result, self direction drops, and we become dependent on other people’s opinions completely. If you’ve ever read a parenting book by Jesper Juul, this consistent with his message: children don’t look for a “good job” when they show you a drawing, they are looking for “I see what you did, tell me about it.” It’s only when you start to “good job” and sticker the crap out of them they lose intrinsic motivation, and just look for validation from their parents all the time.
But I digress. Let me not push my own parenting philosophy into this, except for to say: everything is related! :-)
Instead, if we learn (ourselves, supported by our organization) to get better at self reflection, we develop ourselves much more effectively, resulting in more impact on ourselves, teams, companies and the world in general. And Carol mentions a few companies that work this way, including such companies like Proctor and Gamble where many of these ideas originate and have been applied for decades.
There are three “core human capacities” that need to be developed to make this happen. And you’ll just love their accessible, academic names:
- Locus of control
- Scope of considering
- Source of agency
Locus of control determines how much control we feel over our own lives. The low end of the spectrum (external locus of control) is “I have no control over my life, I’m a victim of my circumstances, everything that’s wrong in my life is because of external factors.” Although you may expect that the high end (internal locus of control) is feeling in control of every aspect of your life, this is of course not realistic. Instead, it’s about taking responsibility for how you react to the things that happen and your ability to use them as opportunities for self development. This is the part you can (learn to) control.
Scope of considering is about the scope we look at when we reflect on our own performance. The low end of the spectrum is considering just yourself (internal considering): me, me, me. The high end is the world or even universe (external considering). So, when you do things and try to reflect on your success or what you’re trying to impact do you consider mostly yourself, do you also consider what’s best for your neighbor, your team, your company, your country, the world?
I was reminded of this while reading an internal discussion on the topic of OLX’s TechRadar. Somebody there commented:
“Although as a developer, I may actually benefit from not having any TechRadar at all, as I can then put on my CV all nice different tech from A to Z that I’ll use on every new project.”
I’m taking this 100% out of context just to make a point, because it’s the ultimate example of internal considering: why pick technology X? It would look great on my CV. This begs the question: if everybody in the company would have the level of considering of at least what’s good for the company, would we need a TechRadar to guide decisions? Maybe not.
Example: What language should I choose for this project? Let’s see, well for me: I know PHP, Perl and Kotlin. Personally Perl is my favorite, but none of my colleagues seem very excited, so that’s out (neighbor, team level considering). Kotlin is more widely used in the company already (this would still be a good use of the TechRadar), so I can get support easily, also the talent market for hiring Kotlin engineers is good, so that even if I need to hand over this project at some point, the company should not have a problem finding replacements (company level considering).
Are these all the factors to take into account? No, so still you need resources (documentation or people) to provide sufficient business context to make such decision on your own, but the purpose of those resources is to provide the context, not to give you feedback or a decision.
Source of agency is about who we expect to take action on things. The low end of the spectrum here is the authoritarian view: we expect others to tell us exactly what to do, and we don’t do anything until told explicitly. The high end of the spectrum is personal agency where we don’t wait around for somebody to act, we view the world as ours to change and step up and take action.
“Cool stuff bro’,” you’ll say. “What do I do with this?” And I’ll be honest, I don’t have a ready answer for you. My challenge with the book is that for me it’s quite hard to put it into action immediately. It requires some processing, and I’m not sure if some sort of revolution is required to use any of its ideas, or there’s an incremental approach.
So what I’m doing with it now is I talk to people about these ideas, and I’m writing about it here. I’m also reading some of Sanford’s other books to get more sense of this. Some ideas are forming, and I’ll share them when I feel they start to make sense, for those interested.