with Spencer Greenberg
the podcast about ideas that matter

Episode 047: Productive Conversations and Feedback Loops (with Julia Carvalho)

Enjoying the episode? Want to listen later? Subscribe on any of these apps or stores to be notified when we release new episodes:

June 4, 2021

How can we become better leaders? How can we give better feedback to others? How can we be better listeners? How can we give good advice? How do startups (or even existing companies) build great products? What sorts of things do experts actually know? When is it useful to poll customers for feedback?

Julia Carvalho serves as an advisor to startups doing impactful things in the world. She thinks a lot about product strategy, teams, and prioritization. She likes emailing with strangers and can be found at

JOSH: Hello and welcome to Clearer Thinking with Spencer Greenberg, the podcast about ideas that matter. I'm Josh Castle, the producer of the podcast, and I'm so glad you've joined us today. In this episode, Spencer speaks with Julia Carvalho about social and emotional dynamics, models for productive conversation, how startups build good products, and strategies for developing constructive feedback loops.

SPENCER: Julia, welcome. It's really great to have you on.

JULIA: Thanks, it's fun to be here.

SPENCER: The first topic I want to discuss with you is the question of how to be a better manager. I'd love to hear some of your thoughts on that.

JULIA: Well, I can give you a really specific example. I'm actually just gonna give you the example rather than the model. I've been recently working on this project with a colleague of mine. She had just made this incredible report and she asked her manager for feedback, and he said it was great. We started talking about it — because I wanted to actually talk with her about the report — and she was frustrated because she's been there for four months, and everytime she asks for feedback she gets a "Great," or a "Good," or "Everyone loves this," or just some other generic feedback. And I remember reading through her report. It was good, it was great, but there's so many ways that something can be good. And this particular report was great because she really nailed the tradeoff between a broad enough report that you got the highlights without getting bogged down. She had just the right level of information and follow-ups that I actually believed the conclusions that she had. I thought that that was such a hard thing to nail. Also, this is something that she just routinely gets really well because she has really good end-user empathy, of understanding what is somebody going to think when they're reading this and what are the hesitations they'll have, or what follow-up data will they need to read to actually fully believe in what she's saying. And so, when I read it and gave her feedback, I said, "This is great and I feel like you've really nailed this balance." There were three or four times where I was reading this and I started to say, "I don't think so, I don't think so." And then in that next paragraph, she answered the thing that came up for me. So we were talking about some of her frustrations, and I think a lot of things came down to just wanting to feel safe, wanting to feel seen. It's not that you want to be seen as good or be told that you did a good job. But you want to understand that the people around you know what makes you you, and what makes you good has been specifically acknowledged and specifically valued.

SPENCER: That's really interesting about really seeing you for who you are and what you're doing well. I think there are other benefits as well for being really specific like that. For one thing, it makes it a lot more credible. It's very easy to just say, "Oh, that was great." But then if you say that was great because XYZ, it's actually much more believable. Furthermore, it also gives a reward for doing things a certain way, which is also encouraging the person to do it more of that way in the future.

JULIA: That's actually a great plug for giving negative feedback as well. I think folks tend to be pretty hesitant to give it, starting out. If I don't ever receive any feedback that's critical from my manager, I might be a little frustrated because I don't know what to improve on but I also just don't trust that I'm really getting their thoughts. And so, I think you need some level for folks to be able to believe that you'll come to them if you're frustrated with their work. I think one thing that's helped me give feedback in the past is just knowing, if my hesitation is coming from not wanting to be the bearer of a bad experience, me holding back is also making an environment that is a little less psychologically safe because it's leaving them out there to potentially dream up scenarios that are worse than feedback that you have in mind.

SPENCER: It reminds me of a situation I once had where I got to know a person. She did these things that made me feel like she thought I was really special and that we had a real deep connection. Then, I witnessed her doing it with two other people, this exact same thing. And I was just so weirded out because I realized that I couldn't trust my instincts about whether she actually liked me or what she felt like our friendship was or how deep it was or how quick it was, because it seemed like it was this special thing. And I think it's similar to someone who just always says things are good and never mentions any negative things. You don't actually know if they think you're doing terrible work or good work. It creates this deep ambiguity. Of course, you can go to the other extreme, too, like someone who's constantly negative and always putting down your work. That's obviously another extreme failure mode and that happens a lot too. But I think it's interesting to think about this less commonly discussed failure mode of just being overly positive and not giving enough genuine feedback.

JULIA: Yeah. I think on both ends, having it as specific as you can make it allows the positive stuff to be believed and the negative stuff to be scoped appropriately. I think this lowers the load of receiving it.

SPENCER: I would just add on top of it, being specific. There's a way that it can be specific that's not that helpful, like, "Oh, you should change this sentence to that sentence." That's really specific but it's not very helpful because it doesn't give the general principle. I think it's best to explain how you think it should be done but also make sure that you communicate the principle of how to do it in the future so it doesn't just apply to this one case.

JULIA: When you mentioned the one about that friend, I actually lost a friendship over this. I think I am that person, and that could be a curious thing to talk about.

SPENCER: Can you elaborate on that?

JULIA: Sure. I really, really like people, and in college, I had a lot of close friendships. That exact scenario happened: A friend who I was very, very, very close to, just didn't realize that he was one of the most special people to me, of everyone that I knew. But I think he would see me being close to other people, and in his mind, he could never feel secure that the friendship was as close to me as it was to him. Because he just said, "Oh, Julia has all these friends and knows all these people and I'm not anything special."

SPENCER: I wonder if that could in part be driven by him not being the sort of person that has that many close relationships. Seeing you have many friendships maybe makes it feel like it undermines the specialness. What do you think about that?

JULIA: I think so. I believe this happens on dates sometimes or a fight. If you go on a first date, and if you're someone who tends to have engaging conversations with a lot of people, you may have an engaging conversation with someone and yet still decide that it wasn't a very good fit. For that person, if they're not someone who is usually drawn into conversations very much, that might feel like there was a lot of spark. And so I think there are people with different background friendship experiences or just general baselines that can be really misread.

SPENCER: I've seen this dynamic quite a few times, usually with a female and a male, where the female is a very warm person who tends to be friendly to everyone, and the male is maybe not as used to getting really warm female attention and misreads it as, "Oh, she really likes me, maybe she has a crush on me," and then ends up being really hurt when they discover actually that the woman is not as interested in them, that the woman's just being nice and giving them attention because that's her personality. I think that that's an interesting dynamic to navigate, that positive attention is not read the same way by all different people. To someone who's not used to getting it or doesn't always connect in that way, it can be a much more profound thing, whereas someone else is like, "Oh, yeah, this is just how I always am."

JULIA: Yeah, this is my theory for why — I feel like on Reddit or other places — people will mention that once they get a girlfriend, all these women hit on them a lot more. And I've always felt so much more comfortable around male friends of mine who are partnered, because I think, "Oh, great, I can be as warm as I want to be. And it just won't be misconstrued because this person is not on the market."

SPENCER: Oh, man, that's funny and sad [laughs]. I find that dynamic so fascinating, that to many people, the idea of having a purely platonic, deep friendship with someone of the opposite gender just seems implausible. It's very strange to me because, for a very long time, most of my friends have been women (I'd say about 70%). And so, the idea that you couldn't have a deep, totally platonic friendship is just totally weird to me. And yet, I often encounter people that seem to have that attitude. What do you think about that?

JULIA: Yeah, it's a similar mystery to me. Growing up since I was quite young, my best friend was a guy, and I remember it was a source of constant teasing. Our siblings were friends, and they 'married' us or something when we were six. It was always like, "Oh, if we really want to rile up Julia and Craig, we'll insinuate that this is something more."

SPENCER: Yeah. It seems to me that so much is lost by preventing that kind of friendship, or just making it seem weird or abnormal. But then on the flip side, we also have to acknowledge that there are lots of cases where — and again, this tends to happen with a female and a male — the female thinks that they're in a platonic relationship, and the male actually has a crush. And then that also creates a really weird dynamic where the male keeps hoping that something's gonna happen, and maybe eventually, he does something that is totally unexpected, that can be very jarring and upsetting. So I feel like the other flip side of this is that sometimes it actually is hard to have those relationships.

JULIA: Yeah, I've been appreciating that as I have grown older. I think I felt pretty frustrated when I was younger. It felt like male friendships were sometimes cyclical. When I've had crushes on folks and it hasn't worked out, I think just because of the way that I tend to crush on people, it's 95% "Gosh, I just love this person," and 5% "Oh, there's a romantic component." And so, it was hard for me to relate when someone would crush on me, and we'd be great friends for a year, and then they would ask me out, and if I was not interested, they would just disappear. And I thought that just felt kind of confusing, and if the pattern repeated, it felt quite frustrating.

SPENCER: Does it feel like it undermines the relationship, like in some way it was fake?

JULIA: I'm not sure if it was fake but it felt like I had to rewrite history a little bit, like the previous year wasn't actually two people sharing a context and a connection; it was me having my conception of the friendship and them having a crush. You just suddenly feel like it's a distancing thing. But I think I've come to realize that it's an overly pessimistic view and I've often heard a lot of folks saying, "Yeah, it's hard to be around this person now," just because they're sad, or it's emotionally painful. I think that just didn't really click for me until recently. That's not necessarily rewriting the past year. It's just, "Hey, this is sort of a painful moment here," and maybe there's some element of shame or distancing or something but it doesn't necessarily rewrite the friendship you guys were building together in the beginning.

SPENCER: I think a dynamic that can play into this is this idea that many women seem to be taught that you're supposed to play hard to get to some extent. And then that is heard by men as, "Oh, if she's giving me ambiguous-seeming signals, that doesn't necessarily mean she's not interested. It means I just need to be more subtle, or be more likable, or be more funny, or whatever, and I'll break through that." Any thoughts on that kind of dynamic?

JULIA: You know it's funny because it really does lead to useful behavior, at least for me. Because I think I am much more likely to develop an interest in someone if they are subtle. It's like when you go to a friend's place, and they have a cat, and you march right up to the cat, they're hiding under the bed and they're not talking with you again. If someone just lets me slowly develop interest, that's always better for me. I think it's probably a dynamic that's responsible for a lot of negative side effects.

SPENCER: So you're saying when I was like, "Julia, you're awesome, let's be friends," that wasn't the optimal way to handle that situation?

JULIA: Oh, no, that doesn't count for friendship.

SPENCER: Oh, okay. Is it more for romantic connections?

JULIA: I think it's actually in response to broad gross generalization. But maybe if men tend to develop feelings faster than women or something like that, I think it's hard for me to develop feelings in the face of someone else having really strong ones.

SPENCER: Why do you think that is? Why do you think you want it to be more subtle?

JULIA: I think the fun part of developing a crush on someone is these slow little moments of wanting more from them than you have — like wanting a text from them now, even though you're gonna get one tomorrow but not right now — that anticipation of something that you want but you can't quite have. I think if someone is pursuing me a little harder than I'm pursuing them, then I never get to have the moment of wanting them more than I have them; I always have them slightly more than I want them. It's not fun [laughs] because you don't have fun moments of like, "Oh, gosh, I saw them this week. We just ended a date, but I really want to see them right now and I want to see them tomorrow, I want to see them now."

SPENCER: That makes a lot of sense. But I assume that wouldn't actually cause you to lose interest in the person, or am I wrong? Could that actually cause you to lose interest?

JULIA: I think it causes me just to not develop interest.

SPENCER: Oh, I see. So as those feelings are starting to vibe, without that kind of anticipation, it's harder for them to develop further.

JULIA: I wonder if this is at the root of people who tend to develop crushes on people who are off-limits because there's a safety — well, I mention the word 'safety' so it's clearly not just hard to develop affection in this environment — I think it's also needing to have a guard up, like if this person is trying to move faster with you than you want them to, you're always gatekeeping and that's not [laughs] a fun mindset to be in.

SPENCER: Yeah, that's something I wanted to ask you about. I have this model of the situation where a lot of times, a woman is not sure how she feels about a guy — maybe she has some interest but she's not really sure how interested or how far she wants to take it, or if she wants to date the person — there's like this quantum uncertainty, and as soon as the guy does something that's too unsubtle, the wave function collapses. I guess my mental model is that the wave function has to collapse to know — unless the woman's sufficiently certain — because there's a lot of risk in not saying no if you're pretty unsure. What do you think of that? Am I on the right track at all or do you think that's wrong?

JULIA: Something I would say in response is — at least for myself and presumably for a number of others — when somebody does something that makes me uncomfortable, my reaction is to freeze. I don't really relate to fight-or-flight at all. I respond in the pathway that somebody has greased for me and, as I'm walking down the pathway, I'm like, "Aaah, I don't think I want to be here." But it takes this activation energy to figure out how to retract yourself from a situation. And so, I think for me, there's a fair amount of risk. For example, I grew up in Chicago; I walk on the streets with my mom all the time. My mom is great interpersonally, has great boundaries, has a very cheery and non-bothered reaction to panhandlers or being asked lots of things. She seems to have no anxiety response to being asked to sign something on a clipboard, or being asked for money, or being asked for food, or anything like that. And I think she's able to navigate that because she's never in a situation where saying no is itself a hard or threatening situation. But I think if you are not very good at saying no, then a situation where you might be called on to say no, or a situation where someone might ask something from you that you don't want to give — even getting into that situation where that's possible — is stressful, a thing to avoid.

SPENCER: Oh, yeah, that's a really good point. So if you're the sort of person that could easily say no and it's no big deal, then you might not have to avoid these kinds of situations. But if you're someone who would struggle to say no, then you have to create this moat around what you get involved in because you don't want to get into a situation where you won't be able to say no and you know you should. It seems like one of the largest, most striking differences in the male and female experience is the experience that males and females tend to have in heterosexual dating. I'm curious to hear what you think about that. But from my perspective, it seems like men feel like they're supposed to be the ones who initiate in many different dimensions, not just initiate sex, but also initiate contact, ask the person out, maybe even do things like choosing the restaurant or that kind of stuff. And women, I think, are often taught that they're not supposed to initiate, that there's something bad about initiating. And some people say that there can be a life hack for women who don't mind initiating. They can actually get really great outcomes by just taking a share of the role and being like, "Hey, I like you. Let's go on a date," and the men are like, "Okay." And I wonder, to what extent do you think this "males initiate, females don't" happens for good reason versus bullshit, or versus there's some life hack around violating it? What do you think about that?

JULIA: It's hard for me to relate. I think I feel pretty out of the norm on all of this stuff. I think I just tend to have pretty low romantic feelings, or just tend to develop crushes really, really slowly. It's hard for me to give an intuitive guess.

SPENCER: And you wouldn't even initiate — at least on the romantic dimension — because you're not gonna have the [laughs] feelings, probably?

JULIA: Or the opposite, which is I think that I only have successful romantic situations when I do initiate because it takes me months to develop a crush on someone. With somebody else initiating, to me, they're going to be months ahead of me.

SPENCER: Oh, interesting. So it's not going to actually be productive? [laughs] Because on date eight, you're gonna be like, "I don't know. [laughs] I don't know if we should kiss or not."

JULIA: Yeah, exactly. But this does touch on a broader, instinctual thing I have that I think feels slightly contrarian. I think in general, there's a lot of criticism of, or tearing down, some broad social things that we see people do that feel dumb. And I guess in theory, these things are never dumb. They're just things that maybe aren't being done very well. An example would be small talk. I see a lot of people on Reddit or in rational circles that feel like small talk is a dumb social thing we've agreed to, and it'd be much better to initiate a prompt and have deeper talk. But I think of small talk as this breadth-first search that you use to figure out where you and the other person might have an agreeable topic to talk about, to have some sort of connection or a fun conversation.

SPENCER: I definitely agree that it can have that role. But it doesn't seem like it's always being used that way. Because if you're talking about, say, the weather, how is that going to lead to a conversation you're both interested in?

JULIA: Well, I think that's a great point and it delineates whether you're having the conversation to have connection or to have intellectual stimulation. I've been hiking a lot in the woods with my dog. When I pass someone and we both talk about the weather — usually we're like, "Oh, man, it's beautiful out there" — I will walk away and have this burning, happy love feeling for five minutes after that.

SPENCER: Whoa, wow. I think you and I are different sorts of creatures. Because I'm like, "Wow, that was awkward and boring."

JULIA: No, I think it goes back to the thing that we talked about with managers, which is wanting to be seen. I feel like a lot of connection or feeling good in a space for some folks comes from having a bit of overlap between your brains. Maybe you're talking with a friend about something, and they're asking you a lot of questions, and you feel like they really, really get what you're talking about now because they've asked you some questions and you've explained, and you've gone through that cycle a few times, and that feeling of like, "Wow, there's something on my mind and this person really, really gets, not just the fact that there's something on my mind. They really understand the thing that I've been grappling with." That's such a nice feeling, right?

SPENCER: I get that and I relate to that, but how does connecting over it being a beautiful day produce that?

JULIA: I think that you're probably not going to have as close a connection with a stranger if you're just chatting with them for a minute. But that's the quickest route to, "What is the thing that I'm thinking about, or I'm experiencing, that you're also experiencing?" I have a bunch of friends in cities like SF and Boston who were wandering the streets when Biden won and was officially declared on Saturday. And it's that feeling of — people are honking at each other — I think there was just a general sense of wellbeing in those folks, not just because they wanted Biden to win but because there's that feeling of, "Ooh, this thing that's been taking up even more space in my head than the weather is being shared and acknowledged," people are waving at me who I don't know. But we're having this moment of connection; the surface area of what's shared has grown and that feeling of connection is stronger.

SPENCER: I think that, say, I met a stranger in the woods just for a moment and we both witnessed something really bizarre or amazing, I think I would get that feeling. But I think when it's something that feels so basic and ordinary like the weather, it just doesn't leave me feeling particularly connected with them. But I like your point about breadth-first search, like a way of looking for something to connect over. And I think some forms of small talk work better than others and I think the weather is usually (at least for me) not a very good way. Maybe for you, it works for that, but I think there are other ways of small talk. When people ask people about what they do for work, that's not always a great question because sometimes people don't care about what they do for work, but it can lead to figuring out what someone cares about. I tend to like questions along the lines of, "What are you interested in?" or "What are you excited about?" They're a little more awkward to ask because they don't have the normal conversational pattern of like, "What do you do?" or "It's beautiful out" has, but I feel like they tend to lead faster to that common ground.

JULIA: Hmm... Yeah. I'm thinking again of that divide between finding connection because of talking about something interesting and finding connection because you're sharing a moment. I like your path for the former and I'm just now trying to think about it. I've been chatting with a lot of neighbors because I'm walking my dog and she's usually chewing on trash on the street. So usually a lot of neighbors will stop and say something, and a lot of those conversations really don't go deep or go anywhere interesting, but they do get somewhere very connecting, and I'm just trying to think through what the differences are there. Not to be a broken record, but I think that what happens often is pointing out something that's shortcutting to some sort of emotion that they can share in. Like this morning, I walked past a man and my dog had a hamburger wrapper in her mouth that I was trying to get out (to no avail) and he found that absolutely hilarious. And I think I was like, "Ugh, every morning," I said as I looked at him. And I guess there was this shared exasperation or the shortcut to an emotion without having the connection first.

SPENCER: I totally know what you're talking about because I've definitely had the experience. But maybe for me it's more commonly something like: I'm in the streets of New York City and someone does something completely crazy, and another person and I look at each other and we're like, "What the heck is wrong with that person? Why would that person do that?" I was walking on the street in New York once, and this guy takes his coffee up to this giant canister of liquid nitrogen that they're using for construction, and he turns on [laughs] the liquid nitrogen to cool down his coffee. And that's the kind of moment where I'll feel a real connection with the people around me [laughs], like, "What is this person doing?"

JULIA: That's incredible. I realize I love moments like that and I wish that there were some sort of concrete way to have them happen more often.

SPENCER: We just need wacky whimsical things all around us.

JULIA: I think move to New York or walk my dog, for me, both of those will.

SPENCER: You know what you need to do? You need to get a pot-bellied pig and walk it on the street. [laughs] I've actually seen that in New York City.

JULIA: Really?


JULIA: Wow. What haven't you seen in New York City?

SPENCER: Oh, man I've seen so much. [laughs] It's pretty disturbing. Not all of it good, I'll tell you that.


SPENCER: Another topic related to social dynamics that I want to ask you about is: How do we be better listeners? I feel like this also ties in to building connection with a new person, but just as much with deepening the friendships we already have.

JULIA: I guess my first thought to that is just to relax a little bit. I've had people ask me that question and I think that it often comes from this feeling of wanting to have better responses or wanting to land on the perfect response. I think people really worry about having the right response to things when they're listening. I guess you hear this advice all the time, where people say, "Oh, don't think of what you're gonna say next. Just actually listen." But that's not really helpful. If you're worried about what you're gonna say next, and you think it matters, having someone say, "Don't worry about it," doesn't help you very much. There's actually one thing I'd recommend people try, which is to grab somebody (like a housemate), grab your phone, and set a timer for seven minutes. Maybe sit back-to-back where you can't see them, and have your housemate close their eyes and just speak for seven minutes — just start speaking, keep talking — and you don't say anything. Their eyes are closed; they can't see you. You're not reacting. At the end of seven minutes, you can switch, and you can go. You close your eyes and you speak. Just start talking, whatever you want to say, and they can't react to you. And then talk about how it goes. I think it sounds horribly awkward. When I've done this with folks, they have been quite surprised at how nice it was, and how comforting they felt the presence of the other person. It's like they got 80 or 90% of the joy of talking to another person, or of feeling heard by another person. They got almost all of that without that person responding at all or nodding or having their face look receptive, or anything like that.

SPENCER: That's really interesting, because it removes all the feedback that we usually think we have to be giving all the time. And if the feedback is there, then maybe, in a way, you do have to give some of it. Otherwise, the person might think you're bored or whatever. But then under this set of rules, it's like, "Oh, actually, we can do without it." I think that's a really cool experiment and it might actually be more enjoyable than talking to some people. [laughs]

JULIA: Well, I would say that's a really true thing. I think it's what you said, that it's more enjoyable than talking with some folks. Because often people will interject, and they will either try to give advice, or jump on a piece that you said that they want to talk about. Often, folks will have a reaction that I think can end up derailing the connections you get from talking.

SPENCER: Right, like the person was starting to say something meaningful to them and then you jumped in and they don't actually get to that place.

JULIA: Yeah, or they jump in with a piece of advice. The thing that I tend to rail against is just giving advice in these types of conversations. Even if somebody is asking you for advice, I've almost always found it better to just listen first and to view your role. I think a lot of times when you want to be a good listener, you also want to help someone — like someone is stuck and it feels like someone's asking you for help getting unstuck and you want to provide it — I think one framework that might be helpful is, rather than trying to find somebody a solution, view their situation as a big tangled ball of yarn, and your role here is to help them sort through it or tack it out to try to figure out what's going on. Rather than weighing in on the situation, maybe a way you can participate is to ask more questions. Try to clarify and, through rounds of conversation, try to get to a closer and closer and closer understanding of what they mean.

SPENCER: I feel like part of that is asking clarifying questions. Is that part of your model of how to do that well?

JULIA: Yeah, I will tend to. If someone says they feel discouraged, I will often ask, "Discouraged in what way?" and just ask questions. Well, let me ask, what types of clarifying questions do you have in mind?

SPENCER: Well, one of them is when you feel like you don't fully understand what the person said. And so, you're just asking to clarify for yourself to make sure you understand, which I feel is important if someone's talking to you, to make sure you really understand what they're saying. Another type is when you feel like someone is starting to explain something but they haven't unpacked it yet. They're like, "Oh, yeah, I really wish that you would do something differently." And you're like, "Oh, okay, well, what is differently?" and you're helping them expand on what they just started.

JULIA: I think all of those are pretty helpful. I want to think through if there's anything, any caveat that I would put on that.

SPENCER: If you're not asking a clarifying question, what's the most common way you're reacting?

JULIA: I think I'm often subconsciously mirroring a fair amount (and I'm not even intending to do so). But if someone's exasperated, I might show surprise or empathy towards the fact that it's exasperating.

SPENCER: It's so funny you say that, because earlier I was thinking, I literally had the thought, "It's interesting interviewing Julia because she's like a mirror. It's like interviewing myself." [laughs]

JULIA: [laughs] Interesting. Yeah, I think I get into a weird spot when giving advice, ironically. I think a lot of what happens with me in a conversation is that my reactions will unintentionally mirror someone in a conversation (like in my face, body, tone, or questions). I think it's a helpful and genuine thing. And so, if someone says something like, "What do you do?" and if they want to incorporate that, then I'm not sure at what point is something manipulative. Do you get what I'm gesturing at? I'm having a hard time explaining it.

SPENCER: I suspect what you're suggesting is that you don't want to influence a person, that you want to help them figure things out. I do sometimes give advice. I totally agree with you that people try to give advice way too often and they often give it when people aren't ready to hear it or don't want it. But sometimes people want advice, and sometimes people need advice. And I'm really interested in how to give good advice, so I'd love to tell you about how I now think about giving advice and I want your reaction on how I'm giving advice. [laughs] This was actually partly born out of a bad experience I had, where a friend of mine was asking for advice about her relationship. She told me the situation, I gave her advice. I later decided it was terrible advice I gave her. And the reason it was bad advice, I would say, is I only had a partial picture of the information (I just had what she had volunteered). And later, I found out other facts about it that would have led me to give her extremely different advice. And I think she wanted me to give her certain advice. It was like she actually wants to do X and she wants to be told that that's okay to do. I think there was something going on there, like using it as a form of validation. But it was actually not good for her. So that experience really got me thinking about what it means to give good advice. The way I think about it now, I try to build a little mini causal model in my mind of the person's situation. As you say, a lot of times people don't want advice, and a lot of times, I'm not even trying to give advice. But if I feel like this person wants and needs advice, then I'm asking them a whole bunch of questions to help me build up this little mental causal model of their situation, where there's some little part of my model that's not yet developed. So I'm asking a question about that and that helps clarify that piece of the model. Then there's another piece of my mental model that's still not clarified so I ask them about that until I feel like I can hold in my mind like, "Okay, what's going on with this person is they really want their boyfriend to express love to them, and their boyfriend does express love, but the way they express love is not the way that they can receive love. And so they keep feeling like their boyfriend doesn't love them," until I have it to the point where I can actually explain the causal dynamics. And then that's when I actually try to give advice, but the way that I try to do it is by proposing two or three different options and saying, "Here are a few things that you might try. Which of these do you think would be most helpful in this situation?" So I'm bringing them back into the loop. They're still going to know so much more than me about the situation even though I have this little mental model. There's gonna be a lot of gaps in it and so I need their feedback on the advice, essentially.

JULIA: Yeah, I really like that. A few pieces: I really like the idea of just viewing it as, keep trying to build a closer and closer model. I think that prevents the issue that some folks come into, of just giving the advice when it comes into their head, rather holding on to your reactions and trying to learn more and more until you have a much closer view. If advice is coming into my head, I may say something like, "Is this the sort of situation where XYZ would be helpful?" and rephrasing it as a question or as an attempt to understand the situation more, which I think is what you're doing there at the end of saying, "Here are some things that have come up or that I can think of. Which of these feel suited to this situation?" You're both giving advice but still doing that model refinement of trying to understand the situation better and understand if this is the type of thing where your suggestion would be appropriate.

SPENCER: Exactly. And then even if they say, "Oh, you know, that might work," then I think you can even go further and say, "Okay, great, do you think you're actually going to do that? What kind of challenges do you foresee with trying to implement that?" That can even flesh it out further. And then they're like, "Well, actually, maybe I might encounter this problem." Now you can update your causal model further and fill in even more detail.

JULIA: I think when what you're doing is offering an idea and letting someone react to it, where they're still in control of the conversation, it's helpful. I've seen someone do this sort of thing but in more of an interrogative way like, "Would you do it? Why wouldn't you?" and there can often be a base fear underneath where you can get into a weird defensive mode.

SPENCER: I think that's really an excellent point. You don't want to make people feel bad for not wanting to do the strategy [laughs] that you suggested or to pressure them into it. It has to be really organically like, "Oh, yeah. That actually seems promising," and then you have to be attuned to that and not push any of the strategies on the person. But if they do seem enthusiastic about it, then you can help them debug a little bit more and explore what could go wrong with it. I've seen coaches — fitness coaches, and life coaches, and therapists — do this thing which I find is a really bad way to do things, where they'll give a strategy that they think is good and the other person will be like, "Sure that sounds good." But then what they don't realize is that the person actually has no intention of implementing it. [laughs] It's like, "Yeah, that's a good idea." But then if you ask the person, like make a bet, "Would you bet $5?" And this person says, "No way." I just think really good coaches and good therapists will then do that further thing and be like, "Okay, do you plan to do this? Why or why not?" without pressuring the person, just trying to surface any further issues. I would also say that another thing about trying to build this causal model to help the person is that it actually ends up seeming a lot like active listening because you have to listen really carefully. You have to notice subtleties about what the person's telling you and you have to ask clarifying questions. So the first 20 minutes or even two hours of the conversation might feel just very much like active listening even though the person actually wants advice, and you tend to try to help them.

JULIA: Yeah, I think of it really as active listening plus a way that you can be additionally helpful, maybe trying to name things eventually. If someone's describing something and you think, "Gosh, that sounds really isolating," you might be able to say, "Yeah, it seems like it would be really isolating." And sometimes giving people words, I think of it as standing at the ready when someone's trying to untangle something. And you're reflecting back what you're hearing in an active listening way. But then also occasionally handing out those tools like, "Does this help?" "Nope," or, "Does this framework help?"

SPENCER: Right, just helping them unwind that ball of yarn. Another thing about the causal model view of advice is that, in order to build a causal model on science, we generally have to do experiments. You have to go do something in the world — obviously with this kind of advice, you're not going to go out and do something in the world during your conversation — but there's an analog of that, where you give the person an experiment and they tell you the result. For example, you're like, "Oh, that's interesting. What if you were to say this to your boss?" and then the person's like, "Well, that wouldn't work because of (blank)." That's a kind where you just did an experiment and they just invalidated one of your hypotheses in your causal model, right?

JULIA: Yeah. It dovetails with another thing that I feel pretty deeply, which is that people are generally quite reasonable in their own reference frame, if you come to what you think is a very close understanding of the issue, and it feels like they really, really should just say this to their boss. And even if they agree that this is what they should say to their boss, then "Are you gonna do it?" and they say no, then that just means it's not a good fit for them. I think people's inclinations tend to be really close reads of their situation. I almost never find myself in a situation where I think, "Gosh, you idiot, just do the thing. This will be so much better." Or, "You are just not seeing this clearly." I think it's almost always the case that there's something deeper underlying that is exerting pressure on the situation that we haven't been able to name yet.

SPENCER: Right. Maybe you as the listener just don't understand all the forces. I mean, there are a lot of reasons why a person might not do a thing that is very reasonable even if that thing would solve their problem. (That's just assuming it would solve their problem, which it might not.) For example, it might just be against their identity, or it might be against their values, or they might actually have some kind of mental block around doing that sort of thing that's very difficult. Maybe they had a traumatic experience in the past and to do that thing might seem simple, but actually it would require facing this difficult trauma, and they are not prepared to do that right now. So I think we have to be very respectful. If someone says, "Yeah, I'm probably not gonna do that," it's important to recognize that they're probably right, that that's not the right path for them. And then I would also add to that — something that I've really updated on in the last few years — this idea that, most of the time, when we have a behavior that we consider bad or that we want to change about ourselves, there is some benefit we're getting out of it. And it doesn't mean that it's worth it. It doesn't mean that we shouldn't try to change it. But I think it's important to acknowledge that, "I actually am getting some benefit out of this." And being honest with yourself about what that benefit is, can be helpful, because it can help direct you towards, "Okay, if I'm gonna change this behavior, I should actually try to make sure that I still get that benefit some other way." Otherwise, it might actually be much more difficult to change it because I'm losing this benefit.

JULIA: I think that makes a lot of sense. I'm curious if you have suggestions or thoughts around ways that people can think through these sorts of things. A lot of these things are things that we're pretty aversive to thinking through in general.

SPENCER: When you say these sorts of things, can you clarify a little bit?

JULIA: Like habits or things that you maybe want to change — like if you're just procrastinating, if you have a lot of shame around those behaviors, they're just not pieces of yourself that you like — I think that's sometimes just a hard thing or an aversive thing to think about. But also if they've been helpful for you in ways that are hard to articulate because of some deeper anxieties or some deeper things, I feel we've gotten to this place where we're maybe not noticing them or not thinking about them for a reason. So I'm curious, does this resonate at all and are there any sort of suggestions that you have for folks that want to think through that?

SPENCER: One of the suggestions that I give for this is to try to really zoom in on the moment, when whatever the thing you want to change happens. Let's say it's like, "Oh, I tend to like to blow up in anger at my friends," or something like that. It's like, "Okay, well, next time it happens, try to pay extremely close attention." Maybe you can't catch it right before but maybe you can catch it right after it's happened, and be like, "What just happened?" Break it down moment by moment. And that really zooming in on the moment — the moment when that branching tree of possibilities goes the wrong direction — often has just a huge amount of information about what's really happening. And if you don't pay that attention, it's often mysterious. You almost feel like you're just going to inevitably do this thing again and again, because you don't have the insight that enables you to realize, "Oh, it's actually when this exact thing happens and then I start feeling this way, and then I say this thing, and then that triggers the other person, and then it spirals out of control," or whatever it is.

JULIA: Interesting. I feel like I've heard a version of this sort of advice and I remember trying to put it into practice with some procrastination I was doing. And then I was just sitting there I'm like, "Yep, I'm aware that I feel like I want to navigate away and load Reddit. That's what I'm doing right now." I just remember feeling some hope and then feeling it fizzle as I was like, "Gosh, I'm noticing it and the desire to tap away or to avoid the thing is as strong as ever."

SPENCER: Yeah. Understanding it doesn't make it go away. It just makes it possible to develop strategies around it. [laughs] I totally agree with that. And then this is kind of a common critique of things like psychodynamic theory, where at least some practitioners in psychodynamic theory — I don't want to paint too broad a brush — will do a very strong emphasis on your childhood and how your parents treated you and your relationship with your parents when you were young. And understanding things that went wrong in your childhood and how they're impacting you today can be a useful tool, but it doesn't automatically solve a problem, even if that was caused by this issue in your childhood. It doesn't make your behavior immediately snap into place like, "Oh, now I understand why I do this self-defeating thing. Okay, now it's solved," no. Now it gives you some understanding and maybe that's helpful. But there's still more you have to do. And so, for procrastinating and this desire to click away and go on Reddit, let's suppose you realize that it's happening when you're doing boring work. Is there some way to make the work a little bit more engaging? What if you were listening to music you really enjoy? Would that actually now feel a little more satisfying, and maybe you're not gonna have that desire? Or maybe it's more like when you're doing stressful work. And it's actually a way of avoiding the stressful experience, because going to Reddit makes you feel more relaxed. That's a totally different causal mechanism. So I think understanding is the first piece. And the reason that I emphasize the causal understanding — both with debugging your own behaviors and trying to create new habits, and when you're trying to give others advice — is I just think that the space of possibilities is way too large. So if you just guess and try things at random, it's actually really, really hard to solve problems because there's so many things you could try and there's so many possibilities of what could be going wrong. But if you can dig in and really understand that what's going on here is A causes B causes C, and that leads to this bad behavior, now maybe that doesn't give you just one intervention that immediately solves it but lets you hone in on a few things you could try.

JULIA: Yeah, interesting. It makes me wonder about the usefulness of having somebody else running through this with you because, while you were talking, I actually realized that I did figure out why I was procrastinating, and it's almost always when I'm doing something that I don't think is actually important, like a task at work that doesn't feel prioritized, basically a task that I don't think I should be doing but I have some external pressure to do it. It's like the stress response of the clash between knowing that it's really important to prioritize pretty strongly, especially in startup environments where you can focus on a lot of things. So it's like the pressure of, "This is wrong," but also the external, "I need to do this, but I don't want to, but I need to do this." And until this moment, I had forgotten that I noticed that because the right thing usually in that case is to not try to keep pushing through, but to step back and go and try to resolve the external, see if there's a way that you can just not do this thing if it's truly not prioritized.

SPENCER: Yeah. That's a great strategy, if you can do it and if it works for you. But then again, it's like, "Oh, well, sometimes people just have to do things that they think are dumb." And so, is there another strategy that in those situations, you can still get yourself to do it?

JULIA: For me, the anxiety or procrastination response comes from thinking like, "Oh, I shouldn't be doing this. My usefulness at work is doing the prioritization and I'm being a bad employee by doing this." So I think if I go back and clarify, and I really need to do this anyway, I think a lot of that procrastination would disappear. Because then it becomes, if this is a certainty and if I have to do this, then it becomes important to get it done because then it's off my plate. But I think the thing that was interesting for me here is that, as you're talking, I remembered that I'd realized this but I just sort of dropped it. I think I was like, "A good strategy for when I'm procrastinating is to focus to see if there's this element in what I'm doing, if I have some subconscious belief that this thing is actually not the right path." But I don't do that. [laughs] I just procrastinate. I had forgotten that. So it makes me think about maybe the presence of a third party like a therapist — or someone who's holding some of this in your head for you — might be helpful.

SPENCER: Yeah, they're giving you time to remember it and they're giving you social pressure to actually act on it. I think therapists could just be incredibly valuable, even if they're not teaching you useful life skills, just for things like, "Hey, you're gonna spend an hour (or 15 minutes, whatever it is) a week thinking about what you're trying to improve and having someone to listen to, to talk about it, and give you some social pressure to to put things into action," so I totally agree. But I love it that you already developed this little causal model, and this actually illustrates the next thing, which is that like, okay, you have a little causal model of what's going on. Now, you may need a little causal model about how to get yourself to do that thing, that exercise of when I have this experience, noticing it, and being, "Actually, this probably isn't something I should be working on." So now you need this additional behavior change on top of that. I also just want to point out that, going back to when we were talking about giving advice, you can do these kinds of (quote) "experiments'' where you ask the other person you're trying to help, "What if you did this? What would happen?" And that's kind of a way of invalidating your hypothesis and updating your causal model. I think the same thing happens for yourself. It's like you have a little causal model of the situation of why you're doing this behavior. It might be wrong, it might be right, or it might be right but not that nuanced and not accurate enough, and you need to add more parts to it to make it more accurate. Trying to solve the problem using the causal model actually teaches you, "Oh, wait, this causal model is missing something," or, "Hey, the causal model predicted that that strategy would work and then it didn't, so I know I need to update it and there's something wrong with it." So I still think of it in an optimization loop.


SPENCER: This same kind of loop occurs when building a startup. I'd love to hear some of your thoughts about the optimization loop of how you build a great product.

JULIA: I think that a theme when trying to think through the personal stuff of interpreting what you're seeing as a potential update to the model — noticing that, gosh, you thought this was the model and you tried something, it didn't seem to remove the problem — that sort of update, I think, is the loop that I see really good people doing. Or when I talk with folks at a startup and I'm like, "Oh, this is going to be successful," I think what I'm usually seeing is people keeping this optimization loop at the very strong forefront of their brain and filtering all actions through it. So usually it's like saying, "What is my current model? What do I need to learn for me to become more confident in that model, or at least for me to take the next step in that model?" and that may be verifying that something doesn't matter. Maybe the explanation is your model, or maybe it's this other thing, and that may be verifying that this other thing doesn't matter. I think, just in general, the folks I've seen that I've had a lot of faith in tend to be, when they are really filtering the information they're seeing and framing what they're doing in terms of running this closer and closer, more and more precise optimization model to try to have a really tight understanding of their product or their users (or things like that).

SPENCER: I think that's a great way of looking at it. I have a particular perspective on this and I'd love for you to critique it. I call it "looping the question" and it's a little hard to explain. Basically, I think if you're building a project — whether it's a startup, or a nonprofit, or you're just trying to make a cool thing, you want to put it out there in the world but you really want it to have an impact or be something people really like, or something people will really benefit from, or something people are willing to buy — I think that you start with what I call the 'meta question.' And the meta question is, "What do I need to know that I don't know?" And it could have a lot of possible answers. But first, you're trying to answer the meta question. And once you hone in on the answer to the meta question — which could be, for example, "Do potential users really have this need?" or "Does this way of phrasing this product actually get people excited?" or "Do people really need to have this particular feature or not?" There are many possible answers to the meta question — but once you answer the meta question, now you have an object level question like, "Do people need this feature?" or "Do people understand the value proposition of my product?" So now you go from the meta question to the object level question. Now, how do you handle that? Well, you have what I refer to as your tool belt, and the tool belt has a whole bunch of tools for answering questions. One example is you could go interview an expert on the topic. Another example is you could go and conduct user interviews with your users or with potential users. You could also survey people. If you have existing customers or users, you could analyze the data you collect on them to see their behavior. There are many, many tools in your tool belt and I find that entrepreneurs are usually familiar with only a few tools (at least new entrepreneurs) and so, they tend to use those over and over again. Like they've got a wrench and a hammer, and they think you can use those for everything, but it's like, oh, no, actually, there's like 20 tools, and which tool to use depends on what question you're trying to answer, and also what stage you're at, and what kind of data you have access to, and so on. And then, as soon as you answer the object level question, like, "Oh, do I really need this feature?" or whatever it is, then you immediately want to go back to the meta question again, say, "Okay, what now do I need to know that I don't know?" And I call this "looping the question" because it just happens again and again in a loop — meta question to object level question to the tool belt — then you answer the question to your satisfaction and then you go back to the meta question. I'd love to hear your reaction to that model.

JULIA: Interesting. Let me just talk through the sort of things that came up when I was listening. The first was, I think that it would be really easy to think through, "Gosh, I have these different tools in my toolkit," and feeling comfortable. A lot of things in your toolkit will sound like they've answered the question for you. You can talk with an expert on remittances, for example (since that's what I'm familiar with), on people sending money back home. You can talk with them and hear what they need, or you can hear them describe the pros and cons if you're doing market research or something. You can hear them discuss it. Or you can have a survey of your users and you can at least see on a survey what your users say. I think it's really easy to get the wrong answers from a lot of those tools. And so, I'm very much a 'talk directly with users and change something, and see what their response is.' I pretty much only use that tool in a lot of cases because I think I hear a lot about startups in the framework of making sure that you're getting information. But I think the hardest thing is not getting information, but ignoring misleading information, or at least getting information that you can trust. So an example of this would be: I'm doing a bunch of debrief interviews right now for this project I'm working on. It was for the election, so we just went through a cycle and I'm talking with a bunch of folks. This is a new product and I don't have a good sense of the users right now. And there's a lot of push right now in my org to send out a survey and ask how people felt about this piece and that piece and like the eight different pieces of our product or something. If I interview that user, and I say, "Hey, how did it go?"and they give me a 20-minute rant on how the reports didn't work for them, and then they mention "Oh, yeah, and the app was clunky," I should walk away from that being like, "Wow. Four people like this user, or four organizations like this organization. If we want to retain them, we really want to have absolutely top-notch reports." If they did a survey and I asked them to comment on all these different parts, they might say that the reports were terrible, the app was bad, the trainings were good, they might just rate all these pieces. I think I have this plan in front of me of these four things to focus on next, when really for this user, there's just one thing.

SPENCER: That makes a lot of sense. And I'm going to try to describe what I think we disagree on. I think that there are many useful tools and each of them is pretty hard to learn. They're not trivial to use and they also work in different situations. I think that there are many really useful tools and each of them can be quite challenging to learn and easy to use badly. Furthermore, for any given question, it really does differ which tool is the right tool for the job. It's very easy to use the wrong tool in the wrong situation. Whereas I think you're suggesting that only a few of the tools are really good (like the other tools are kind of junk). And so I think, if I'm not mistaken, the main crux of our disagreement is on how useful are the other tools. I think you and I both agree that talking to your users is incredibly valuable, right?

JULIA: Yeah.

SPENCER: Absolutely, and that's a very powerful tool. And it's one of the most useful and used tools and I would agree with that. But you might say, "Well, is talking to experts really that useful?" And I would say on certain questions, it's actually incredibly useful, but maybe it's like a smaller sliver of questions than many people might think. Talking to experts on remittances may not help you build your remittance app that well, but there might be some question where it's actually absolutely critical and it's very hard to answer through any other tool. Surveys — I run a ton of online studies, and I've spent years honing my understanding of how to survey people — and I think there's a lot of pitfalls with surveying. But I think sometimes surveys can just be dramatically useful and it's just about matching the survey to the right question that really can answer. I'm curious to hear your reaction to that.

JULIA: I think that's fair and it'd be really interesting to have an index of the various methods, and what their pitfalls are and where they can be safely applied.

SPENCER: I would love to build that index. I have little parts of it intuitively in my mind because someone's like, "I'm trying to answer this question," and I'm like, "Oh, you should use this." But I haven't actually [laughs] added into a big map. But I think there are a few things to say about that. One is, if you want to actually see how people will behave if this happens, A/B tests are often just the best possible way to do that, assuming you have users. If you don't have users, you can't do an A/B test. But if you have users and you're like, "How would people behave in this situation?" just go put them in that situation, because it's most true to life. So that would be one example of when a tool would do a really good job. When you want to develop a theory of what's going on, I think that that's when experts can be really helpful. You're like, "I don't understand this aspect of user behavior," or "I don't understand how we would even go about getting people to behave in this way." Talking with an expert can give you a bunch of ideas on mapping that kind of behavior or that kind of space. For example, we build products in the mental health space and let's say you wanted to understand something about borderline personality disorder. It really might be helpful early on to talk to a bunch of experts in borderline personality disorder to try to really understand the dynamics. Why does this disorder happen, and why do people behave these different ways when they have this disorder? And actually it might be really hard to build a good product until you have those dynamics mapped out.

JULIA: That makes a lot of sense. I'm curious if you can think of an example, where experts would be helpful in more of a...

SPENCER: A business-y context?

JULIA: Yeah, I just have a lot of skepticism and I suspect it's just born out of not thinking of the right situation. But yeah, I have a hard time coming up.

SPENCER: Totally. Let's use the B2B satellite as an example. You want to sell your product to, let's say, large financial institutions. I think that's a great example where talking to an expert — and by 'expert' in this situation, I'm not talking about an academic, I'm talking about someone who's been in that world and seen hundreds of products get sold to financial institutions or ideally been on both sides of transactions, that would be the best-case scenario — and they'll be able to break down a huge amount of information about what you need to know to successfully make those sales.

JULIA: Seems fair. I guess, my brain's like, "Oh, advice. Advice is great." I think I broadly agree here.

SPENCER: Maybe you're also pushing on when you should talk to academics, because maybe that was what you were thinking of originally.

JULIA: I'm not sure I have a pretty strong intellectualized theory here. I think I just have had a lot of experience in the work that I've done where talking with an expert has almost always been misleading, even to the point of like, "Hey, is this area safe to go into?" like literally, "Can I travel here?"

SPENCER: What kind of experts are you referring to?

JULIA: Let's see. It's sort of now codified into a gut feeling. I need to kind of roll back.

SPENCER: Are these people who are academics who published papers about it, or just people who have been in the industry a long time?

JULIA: I think people who have been in the industry for a long time. Part of the reason that the startup I was at was successful was it was ignoring some of the standards of what matters or what people care about because we were finding that users on the ground seem to care about something much different. I guess if I'm really talking with these folks about what matters in this space, that is something that it's probably better to talk to users about. So I think this could be a case of just having misapplied. It could be that for my experience or the type of work that I was doing, talking to users was appropriate 90% of the time. That just happened to be the space I was in.

SPENCER: Yeah, that's really interesting. And I do think that using these tools is very subtle. And, in your situation, maybe you're just doing the right thing by talking to users, but it also leads to this question of, "What is an expert really an expert in?" Because I tend to take the view that a lot of times, experts are not experts in the thing that people expect them to be. For example, someone working in the finance world for a long time is probably actually an expert in some things about the space and actually not an expert about other things. But they might have strong opinions. Let's say they've been on both sides of large B2B transactions. They will probably have a lot of really good advice about how to help those transactions happen and pitfalls you're going to experience if you try to make them happen. But let's say they were the CEO of a company and were never really involved in those transactions directly because they're too high-level. They might actually not fully understand the dynamics, but they might think they do because they're the CEO of a large finance institute, of course they understand. So I think that is a real subtlety here. Another example is that, when I think about experts, I like to think about what feedback loops they experience because it's very hard to truly be an expert in something when you don't have a feedback loop. It's not impossible, but I think it's a lot harder. And so someone who's involved in a lot of large financial transactions, they're getting a feedback loop, they're seeing some deals fall through, they're seeing some deals go forward, etc. But you can have a lot of situations where an expert seems to be an expert, but they have no feedback loop. As an example of this, imagine someone who gives people short-term therapy but they have no follow-up. They just give someone six sessions of therapy and let them go. They never find out what really happened to that person. Do they really know whether they're helping that person long-term or not? Their actual feedback loop is more around: Does this person come back after session one? Does the person seem to have a positive experience during the sessions? I would imagine that person would have a much tighter feedback loop on how you help a person calm down during a session and feel comfortable sharing, and not a very good feedback loop on how you help someone in the long-term.

JULIA: I think that makes a lot of sense. It makes me wonder how much the cache of knowledge I have about (I don't know) just the things that people generally tell you about health and nutrition, and just general life.

SPENCER: Let's unpack that a little bit more. I would say experts typically have these two types of knowledge: the stuff that they've learned that's the standard stuff experts learn, like a doctor is gonna go through a standard medical training. If their school is good, there's gonna be certain things they teach them about disease and treatments, and whatever. And then, they have the feedback loop stuff that they see. And both of those can be a real form of expertise. The first type though, you might get from talking to any pretty good doctor in that particular space; you're going to get the standard knowledge taught in med schools about that space. That other thing you're gonna want to pay really close attention to is what is the feedback loop that this person is actually experiencing in their expertise. There's a big difference between being in charge of a firefighting squad and having to decide on, "Do we go into this building or is it going to collapse on us?" and making the decisions everyday versus someone who's sitting in the headquarters [laughs] deploying the squadrons. They have very different feedback loops and that's something I think about. For example, really good doctors are ones that can do well outside of the standard training that a good doctor gets or ones where they've had a tight feedback loop on that particular kind of disorder, or disease, or surgery, or whatever. At least that's my opinion.

JULIA: Yeah, it makes me think of, one, again, I wish I had an index here of areas where this matters and how to choose or when to be skeptical. But it also makes me wonder if there's a way that things could be structured much better to try to keep the feedback loop attached to the practitioner, the person, a way to structure things that doesn't lose that feedback loop.

SPENCER: Yeah, this frustrates me so much when people are doing something that they could be so much better at if they just added a feedback loop. I'm like, "Add a feedback loop." But there are actually many things where it's abnormal to add a feedback loop even though it's so valuable. As an example, most therapists, I think, do not have a feedback survey. And this just seems devastatingly bad to me. In my opinion, you should have an anonymous feedback survey that you just send out to every patient, ideally every week (it could be really fast, it could take two minutes to fill out or something). But you could send it out to every patient, and ideally it's anonymous (with maybe the option to deanonymize if a patient wants to tell you who they are, fine, but they shouldn't have to), and then saying, "Hey, what did you think of our last session? What could I have done better? What was helpful? What was not?" That kind of thing. I just think this would be amazing and help people improve. If someone's already a good therapist, I think it'll make them even better. But if they're bad therapists, they really need this. That's just in therapy but I think there are so many things. For example, writing, I think there's one reason why being in a writing group can be really valuable. If every week, other people are reading your writing and giving you feedback, versus you're just writing in a vacuum for two years with no information coming back to you on, "Is this resonating with people? Are people feeling good about it or bad about it?" Any thoughts on feedback loops in life?

JULIA: I'm curious if there are any categories of this that you've discovered that are beyond just getting feedback from other folks. I think there's asking for feedback from folks. There's maybe like recording yourself playing piano and then playing it back or something like that. But I'm curious if there are more material ways that you've figured out how to add some sort of feedback loop into your life?

SPENCER: Well, I have a pretty funny one, which is, I have a goal of posting one idea on Twitter every day. I'll usually try to do it first thing when I wake up or before I go to bed, I try to come up with that idea, or, failing that, I'll try to think of one later in the day. This idea — then people will react to it either by just reacting directly, or commenting, or whatever — and if it's an idea that seems to resonate, then I'll consider turning it into a Facebook post. If I turn it into a Facebook post, I tend to get a lot of comments with people either critiquing the idea, or suggesting improvements, or riffing on it. And then I'll say, "Oh, thank you for that comment. I'm gonna go update the post based on the point you made," and I'll actually hone it through that feedback and make the post better and better and better. And then the final stage is, if I like that enough, I'll go port it to my blog where I do a final pass and I have someone who will read through it and will run a grammar check over it, etc. So that's the final polishing pass. For me, this is a wonderful chain reaction where I'm getting feedback, like it's a very little kernel of the idea that gives me enough information to bootstrap it up to the harder-to-do version where I get even more feedback, etc.

JULIA: Interesting. It seems like having different layers of feedback or having different circles of it — not just asking someone how it was, but having different audiences and refining it — that's really interesting.

SPENCER: Yeah, and there's different kinds of feedback. One piece of feedback is just, "Does it resonate?" If nobody seems to be interested in it, it's like, "Okay, that might be a great idea but I can't get people to care. So it's gonna be hard, and it's probably not gonna be that valuable because people won't care about it." That's one type of feedback. And a completely different type of feedback is like, "What's wrong with this idea? What are the flaws in it? How could I improve it, or generalize it, or whatever?"

JULIA: Something that's interesting here is that I often think of feedback as cheap, but feedback from the people you want feedback from, the specific kind of feedback you want is kind of expensive. People in the startup world often want to expand. If they're not gaining traction on their main user base, they may want to sort of expand it to like, "Oh, we'll take on this client, or we'll move into this geographic area, even though we know this isn't a great source of depth for us, we want to keep moving." And there's a huge hidden cost of that like, "Okay, but your feedback is now going to be coming from a group of people that are not your main users or the people you want to be your main users." And so, it might be overlapping. If it's completely different feedback, that's almost helpful, because then at least you can parse it out. But often, if it's just slightly, subtly different, it's going to be really hard to not be a little bit misled by that. And so I guess, as your one cautionary thought that I was thinking of on your Twitter, Facebook, blog post cycle is thinking, "Who are your Twitter responders?" There are ideas that might really resonate in some groups and not in others. I think, similar to the toolkit of using a tool but getting the wrong feedback from it can be pretty damaging, I would worry about getting the right feedback.

SPENCER: That's such a great point. I would say there's two subpoints there. One is, is it the right people? If feedback is coming from people, are those the kind of people you really want to optimize for? This is kind of the dark side of creating a feedback loop, you're going to be pushed towards it. If that's not really the right target audience, then that could be a problem. The second thing you're saying there (as I interpret it) is that, even if it's the right people, is it the right sort of feedback? Because I think a danger that can happen — let's say, posting on something like Twitter — is maybe you just get pushed towards saying things that get people riled up because you get a lot of attention. If your optimization loop was like, "How much attention did I get with my tweet?" that could actually go to dark places. So you have to make sure that the sort of feedback is actually honing in the direction you want. Because I tend to post about ideas that are not political — they're not criticizing any one group or anything like that — I think it tends to avoid this sort of optimization loop of negativity and outrage and things like that. It's more like, "Did this idea resonate with people? Did they find it valuable?" I try to keep it more positive, and, "Is it useful to my audience?" And I also have studied my audience on Twitter. I've actually run a series of polls to figure out who they are and what their demographics are. I kind of view them as the core audience, pretty close to the core audience that I'm trying to write for anyway, so it kind of works for me. But I totally agree. I think that's a really useful downside to consider. Julia, this has been such a fun conversation. Thank you so much for coming on.

JULIA: Yeah, thank you so much.





Click here to return to the list of all episodes.


Sign up to receive one helpful idea and one brand-new podcast episode each week!

Contact Us

We'd love to hear from you! To give us your feedback on the podcast, or to tell us about how the ideas from the podcast have impacted you, send us an email at:

Or connect with us on social media: