At the time of this talk (2016), Helen Toner was a senior research analyst at the Open Philanthropy Project. Here, she shares management lessons — such as holding regular one-on-ones and soliciting feedback — that she learned in her first year at GiveWell and the Open Philanthropy Project.
Below is a transcript of the talk, which we have edited lightly for clarity. You can also watch Helen’s talk on YouTube or read its transcript on effectivealtruism.org.
As I’m sure you all know by now, effective altruism [EA] is about doing the most good you can with the resources available to you. And one strategy that gets talked about a lot within EA is earning to give — increasing the amount of money you have access to. That is one resource you can do more good with. But working within and helping to lead an organization is a way to grow other resources: the amount of time, talent, attention, and energy to which you have access.
So, my hope is that this talk is relevant for anyone who might want to found a nonprofit, found a company, lead a team, grow a team, or work on any kind of team. That’s my hope.
That said, I’m not focusing specifically on collaborating in small groups, so for the Berkeley students among you, I’m sorry — I wish I could tell you everything you need to know about how to do group projects painlessly, but I can’t. The most I can wish you is professors who let you choose your groups or at least give you a peer feedback form.
One of the many things that drew me to GiveWell and Open Philanthropy Project when I started working with them a year ago was that it looked to me like those organizations were growing fast and growing thoughtfully. I wanted to be able to observe that. I wanted to learn from it. I wanted to be a part of it. And that’s what I’m hoping to talk about today. But with that said, I only joined a year ago, so I’m not a guru. I can’t tell you everything that you need to know about management or growing an organization.
I’m going to share some observations and things that I’ve learned over the past year that I hope will be helpful. And I’m planning to leave time for you to share your own observations or ask me about mine. And I should also clarify that none of this talk is GiveWell’s or Open Philanthropy’s official take on any of this. It’s all just my own personal observations and thoughts.
About a week after I started at GiveWell, Elie [Hassenfeld], one of the co-founders, said to me, “Do you want to go for a walk, have a chat, and get to know each other?” So we went on a walk. And one of the things he asked me on this walk was: “What’s your least favorite thing about working at GiveWell so far?”
That is a pretty tough question for someone who’s only been there for a week and is talking to the co-founder for one of the first extended periods of time you’ve had together. I thought for awhile and the best thing I could come up with was that there weren’t enough SodaStream bottles. We had a SodaStream machine in our kitchen, which converts normal water into sparkling water, and some staff had a SodaStream bottle on their desk. I couldn’t find a bottle to use. Grave problems! [Laughs.]
Elie and I continued to chat. We went back to the office and two days later Elie came to me with SodaStream bottles that he had immediately ordered after our conversation. From then on we’ve always had enough SodaStream bottles, even as our team has grown.
This is obviously a pretty trivial example of what I would claim is good management, but I think this particular story demonstrates three things about communicating openly that are important. The first thing it demonstrates is the importance of giving people lots of opportunities to communicate. At GiveWell, for example, every week we have one-on-ones, which means that each staff member goes for a half-hour walk or sits down for half an hour with their direct manager. And this is just a time to talk. It’s a pretty standard piece of management advice, but I want to repeat it because I think it’s so great. It’s just time to talk with the people you work with about what’s on their mind, what they’re worried about, and maybe for them to give you feedback, or for you to give them feedback, and so on.
As well as these weekly one-on-ones with your direct manager, GiveWell’s other senior staff also make it really easy to set up a time to talk to them. And they try to make people feel welcome to send them random emails and chat messages to keep the information flow going.
The second thing I think this SodaStream story illustrates is the value of actively soliciting negative feedback. This question — “What’s your least favorite thing about working here?” — really gave me the opening to bring up something negative, which could be something small like a SodaStream bottle, or something bigger.
I think it can be scary to bring up things like this, especially as a new employee, or if you don’t know if the thing is actually a problem, or if you’re not sure how you would solve it. Questions like “What’s your least favorite thing?”, or just making it clear in general that you are willing to hear what people have to say, even if it’s negative, can be really valuable.
And the third thing the SodaStream story shows is the value of following through or acting on what you hear. The fact that Elie immediately went straight to Amazon and bought more bottles shows that, even though what I said was a pretty small issue, it was taken seriously. That encouraged me to feel comfortable engaging more and spending more time thinking about this kind of thing — and then bringing it up when I did. I think that SodaStream story, simple though it is, is pretty illustrative.
Communication can also be badly calibrated. A good story that probably most people here are familiar with — and that shows how this can be done badly — is the jokes you see on the internet about what it means when your girlfriend says, “Oh no, it’s fine.” I don’t want to go into whether these jokes are accurate or blatantly sexist or some combination of the two, but I think it’s just a fantastic example of badly calibrated communication. What happens when someone says something they don’t mean is you need to adjust — and you don’t know quite how far to adjust. And when you’re working with people, there can be a real temptation to be nice. I do think it is really valuable to make an effort to keep the things you’re saying calibrated. And what that means is when someone tells me something I did was fine, I know that it was fine. If they tell me that it was very good, I can feel good about it having been very good. And if they say it was just below expectations or way below expectations, I can take that at face value as well.
This obviously comes up in the context of giving feedback. I think that is a good use case. But I think it can be useful in other places as well. For example, when talking about your own work, if you’re doing a research project and you’ve come up with some conclusions, it can be really useful being clear about which conclusions you feel confident in and which conclusions you feel less confident in — which conclusions you might change your mind on if you put in a little bit more work.
Anyone who’s read much of the stuff that GiveWell or Open Philanthropy writes online will know that we’ve used this self-calibrated communication in our external communication as well.
It turns out that if there is something you need done, and you’ve done it before, then having a process that someone else can follow is really pretty great. Just to take a simple example, if we want to recruit some summer interns, we’ve done that for a couple of years. If we have a process written down, then we can just hand that task to someone, they can follow the process, the job gets done, and that is great.
But to show how this can go wrong, I’m going to use a hypothetical that might not be as probable as I wish it were. You can imagine, for example, that five years from now GiveWell has really taken off as an internship destination. Berkeley or Stanford students are thinking about their summer, and they’re really thinking about Google, SpaceX, and GiveWell. You can imagine what might happen in this case is that people would put a very different amount of time into preparing for the interviews than they do right now. If we were to use the process we use right now to try and identify people who will really contribute to GiveWell, and just find the people who are willing to pay for training they can get on the internet or who knew people who had interned at Givewell before, for example, [that process might not work].
Obviously, this is a whimsical example, but my point is that even though processes can save time, make things more consistent, and codify well-established practices, processes can also get out of sync with what they’re trying to do. If you have a bunch of processes that start to pile up and ossify, then you don’t end up with an effectively functioning organization. Instead, you end up with the DMV.
I don’t have a full solution for this, but I think that a good start is to actively keep this in mind and actively remind the people whom you work with that processes are designed to serve your principles and serve your goals. It should always be an option to just throw aside a process completely or to redesign it from scratch if it gets out of sync with what’s going on.
If senior staff remind other people that the processes they’re following aren’t the word of God, you’ve gone one step towards not letting things ossify in a bad way. And when you pass tasks between people and give people new work, if you can make it clear why things are done the way they’re done, that makes it possible for the people who have just taken on the work to keep an eye out if the underlying circumstances happen to change. Then they can change the processes to keep them in sync.
In terms of growing capacity and increasing your ability to do things with more people, I’m thinking of three facts that apply to every organization. One is that you need tasks done. Another is that you need tasks done well. And the third is that you have incomplete information about the people on your staff — what they can do well, what they can do badly.
There are two extremes of how you could handle this, which seem to be pretty clearly failure modes. One of them would be micromanaging or overmanaging. The example I think of is when I interned at a management consulting firm and I had to make a bunch of slides with organizational charts showing who was working at the company, what their roles were, and how they were connected. And my manager really wanted to check not just that I’d gotten the names right and the relationships correct. It was also very important to him that I had chosen the right font, the right font size, the right shape and size of the boxes. I think that was both a waste of his time and quite frustrating for me.
And on the other extreme, you have macromanaging or undermanaging, which is where you just throw someone right in the deep end. I think this is something we have all experienced a version of. Maybe one example from college might be if a lecturer gives a long lecture on the details of a theory and an abstract, and hands you the problem set, and you look at the problem set and realize you have no idea how you’re supposed to do it. So, you go back home and figure it out all by yourself.
Both of those extremes are bad. In the first, if you micromanage, it’s a bad use of the manager’s time and very frustrating for the employee. And if you undermanage, it can lead to mistakes. It is also frustrating for the employee to not know how they’re supposed to go about their work.
There’s something I’ve seen at GiveWell that seems to me a pretty good way to handle this. When you give someone a new project and you are not sure how they’re going to do, start out following them really quite closely, checking on them frequently, keeping an eye on what they do well and where they need guidance, whether they’re improving, and then gradually hand people more ownership over things that they can do well as you become confident in them.
This also is much easier — this style of following closely at first and gradually handing out ownership — if you’re in an environment where there’s a lot of open and calibrated communication. The simplest way to demonstrate this is to share one of our dark secrets, which is that about 95% of our internal communication is done by emojis.
As my parting gift, I want to show you two of my favorite emojis, which I think are really useful for open and calibrated communication: the thumbs up and one we call “TMI.” We use it a little differently from the normal meaning.
The thumbs up stands for approval. You might send your manager what you’ve recently been working on and then you’ll plan next steps. They could use the thumbs up to indicate that your next steps sound good. It gives them a chance to comment if they want to, but if they just do a thumbs up emoji, then you’re really maintaining ownership over the project and just giving them a chance to weigh in quite cheaply.
TMI, likewise, can be used in the same situation. We use it to say, “I’m basically deferring to your judgment. I haven’t thought about this hard. You go, you do it, you decide if it’s good.”
I think this is really useful to have these two options. I don’t think that every organization necessarily needs these exact two emojis, but they’re an example of making it simple to check in and drive a project forward by yourself, and also making it easy to go to your manager or others when you need help.
Following people closely and then gradually handing them more ownership, as facilitated by open and calibrated communication, means that you can gradually give people the hardest work that they can do. This both keeps the people you work with engaged and it means that your organization can do more in the long run.
Thanks for making this available!
My impression from having worked at 2 EA organizations in the last few years, and the conversations I've had in the community more generally, is that paying more attention to and sharing more thoughts about how to build orgs and other organizational issues could be very valuable. E.g. on: management, hiring, how to scale, communication, feedback, pros and cons of different decision-making structures, how to productively confront interpersonal conflict, accountability/oversight mechanisms, ...
I think one of the strengths of the EA community is that improving our reasoning as individuals has really become part of our 'cultural DNA'. I'm thinking of all the discussions on cognitive biases, forecasting, how to weigh different types of evidence, etc. If the cultural status and available knowledge on 'organizational improvement' could become as strong as for 'self-improvement', that would make me really optimistic about what we can achieve.
It seems to me that some of those issues have been discussed quite a bit; e.g. how to recruit, how to communicate, and how to give feedback.
Thanks, that's good to know.
I guess I'm mostly comparing the two EA orgs I've worked with, and my memory of informal conversations in the community, with my experience elsewhere, e.g. at a student-run global poverty nonprofit I volunteered for while at university. It's possible that my sample of EA conversations was unrepresentative, or that I'm comparing to an unusually high baseline.