2x Grammarly’s Subscribers | PM Mock Interview (ft. Chess.com Growth Officer)

Katherine was really, really strong. Your ability to kind of jump across a variety of different dimensions of that I thought was impressive. Imagine you're the head of product growth at Grammarly and your job is to double the amount of paid subscribers that Grammarly has.

Hey, Katherine. Do you go by Katherine or Cat? >> Uh, Katherine's fine.

>> Okay, fantastic. Well, I'm excited to do this interview with you. Likewise.

>> Um, I'm Albert. I work at chess. com as their chief growth officer and have had a fun career working on a bunch of different consumer products and uh trying to figure out how to grow them.

So, I think this is a perfect topic for us to to riff on today. >> Yeah, totally. Nice.

Nice to meet you. Our um I'm Katherine. I'm currently a principal growth PM at Adobe focused on BBY growth and spent many years working on growth from startup to now big tech.

So yeah, I think this is going to be a a fun journey. >> Yeah, love that background. Um, have you used Grammarly before?

>> I have and fantastic. >> Yeah, it's >> Go ahead. >> I used to work there, so I think this would be a good company and product for us to do our growth question for if you're up for that.

>> Yeah, totally. Let's go for it. >> Okay, so let's imagine that you work on growth at Grammarly.

Like you're the head of product growth as an example. So the vast majority of the user base uses the product for free. Some small percentage of it is paid and that's how they make most of their money is off of those subscribers, right?

So let's imagine that you lead product growth for Grammarly and your job is to double the amount of subscribers that the company has. Um I don't know the number. So if you want to use 1 million, you can.

So let's say you're growing from 1 million to to 2 million. I would just love to talk with you for I don't know 15 30 minutes on like how that might be possible and I'll walk you through various steps of it. But first I don't know if you have any questions about that that prompt and then we can get into it.

>> Yeah totally. Um I think my first question would be when we're talking about subscribers are we talking about overall subscribers or are we only looking at paid subscribers which is directly transferred to revenue. So I think that's that kind of really determines where we're going to focus on across the growth journey.

So I would just love to get clarification on that. >> It's a great question. I would say let's focus on the paid subscribers.

But with that said, paid subscribers are a subset of the overall user base, right? So think holistically as you as you think about this problem. >> Yeah.

Yeah. No, thanks for the clarification. That's super helpful.

But yeah, I think when we're looking at the growth journey, obviously we'll start from the initial acquisition and uh maybe this is a good segue into just how I would look at how we would break down the growth funnel. Um and this is I think this is a really interesting product because it's a typical um growth funnel for premium model. So obviously we'll start from acquisition and this is the how do you even get the free users to sign up right and then moving on to freeto paid conversion.

How do we actually get the um users to start paying and transfer to revenue? And then I think an interesting part of Grammarly just based on playing around it I do see that there is a team expansion PLG motion as well. How do we nudge these individuals to um get their team on board right can and also working B2B we generally know that these team users have better retention and have more prop propensity to grow over time.

So that's a interesting bucket that we'll look at. And then I think eventually we also uh want to move on to retention. So how can we keep the um users continue to engage and really prevent turn.

So I think that's kind of the general high level breakdown that I will look at across the growth journey. And typically I think at big companies each of these stage is a pillar on its own. Um there's like a whole team usually supporting all of each of these stage and it has a lot of different tactics.

Um so since I'm I'm assuming I'm starting from ground zero um ideally I want to get a high level understanding across all of these stages and what's the current landscape um what's the current status quo and I think the way I would look at it is typically from three lens one is really understanding the user. I think I'm a big advocate for just really understand the user, understand their needs, who they are. Um, and then look at the data, both quant data and any quad data that we can get our hands on.

Um, and also like what's the experience? What's the current experience looking at like what are the potential gaps um based on the user psychology. So that's kind of like general, that's how my brain will probably start working.

>> That's great. No, and I appreciate you using the the framework that's common across growth, right, which is to look across the user journey and and the various stages. So, I think that makes a a lot of sense.

Um, yeah, let's start by understanding the user. Uh, I can give you a just a very high level construct, which is that Grammarly supports both students and professionals. And I think recently they thought about expanding to other languages, but largely it's a it's an English, you know, writing type of product, right?

And soers that are, you know, have those those needs to improve that that English writing in particular are the ones that are most attracted to trying Grammarly. But yeah, any other dimensions that come to you that you want to talk about? >> Yeah, I think um one uh I think student is an interesting one.

I think I'm actually interested at student versus business pro. What's kind what's the breakdown? Cuz um I think there is some budget that that's related to that.

So like for Bispo, think of myself who who is a PM who is working in document like working on documentation all the time and I really care about efficiency and clarity in writing. My intent might be different than student who is probably also looking at more how do I prevent um plagiarism? How do I become more efficient?

Right? Because I'm turning papers all the time. So I think these are uh different use cases and that can determine maybe different motivations with how we speak to them.

And I think another thing I think it'll be interesting is looking at non-native English speakers and really what's their local, what's their current market because I I would imagine a lot of people who are probably working in Asia or South America but work with English speaking speakers professionally. That's also another big use case. So I will look at these use cases differently and also like what are the different types of people that are behind these different use cases.

>> Makes a lot of sense to me. >> Cool. Um yeah, you can keep going.

I think that's great. >> Yeah. And I think um I don't know how much we need to dive into the users, but I do think there are some metrics that we can look at um to understand.

So I think for example for the non-English speakers because we have data across all of their writing, right? We can see if they're switching between English or Spanish or like Chinese between English, Japanese between English. I think that's an interesting thing.

Um I think for B pro versus student obviously like what's their style of writing what are the surfaces what are the applications to your point I think that's another huge things are they using slack or are they using word dark or are they using um canva for example right like all of the what are the different applications that they're using um so yeah I think these are just some examples that uh metrics examples that I will look at to break down the users >> those are all great and yeah I can fill in a little bit because I have some context on this is that you know roughly speaking let's imagine that students and professionals are like 50/50 in terms of the user base but their use cases as you mentioned are different right I think students tend to write more long form and they are more writing for the correctness and you know plagiarism use cases that you correctly called out so I think that's great I think professionals tend to be a little bit more uh short form it's email it's Slack it's these types of constant communications throughout the day and that also reflects its way to the subscriber side in that professionals tend to buy these like annual plans and they use it across lots and lots and lots of different SAS applications whereas the student use case is a little bit more constrained sometimes more short-term as well right because you might have finals or a paper or something so I think that's a great call out and that the two user groups are pretty distinct from each other >> nice yeah thanks for confirming that and it's great to know like further insights um yeah So I think that's the user angle and then um I think another angle that I talk about is looking at the actual metrics along the funnel where are they dropping off. So this is typically where you're looking at like what's the what's the friction what's the drop off point um along the journey. So I think some I think I don't I I would just go through the whole funnel.

It is pretty but like yeah I think it's >> you can make your own assumptions on this stuff. So yeah, don't let me bias you in any way. I think we're mostly trying to get a sense of how you think through this problem.

>> Yeah. So I think yeah, starting from acquisition going to that some metrics associated with that are like site visits, initial downloads, Chrome uh the extension downloads to the actual user sign up, how many account creation are we generating and then free activation. So I think related to that is actually usage or edit usage.

I would assume like within extra number of days probably seven days and then um paid conversion I think that's probably a big bucket and when we are looking at paid convers uh paid conversion um the funnel that I will further drill down to is if they um the initial usage within seven days how many of them actually start the pay see the payw wall meaning if they clicked into any of the unlock now or like starf free trial and then from the payw wall how many of them actually start the trial and from the trial how many of them actually um converted because there is you can potentially you I mean you can cancel within the first seven days right so I think that paid free to paid conversion bucket is generally what a lot of tests can happen um because I do think that's also like a golden period like within the first how do we catch them um at the beginning and then moving on to subscription retention then um there is the turn rate and the turn window like I would imagine probably the term window is pretty early on within first 30 days something like that. So um these are kind of the metrics that I would look at just to narrow down what are the problem areas. >> Fantastic.

Yeah. And to just bring it back up to that like top level question again about like doubling subscribers, right? So you now have the the user journey, you have some metrics that you would look at.

I don't know presumably you might try to do some benchmarking or some sense of like based on its actuals like where are the weak points that you might want to try to improve right is that the logical thinking here >> yeah totally so I'll match the um the the the data points with the user and also by looking at experience and understanding what's the current growth levels and generally identify what are the bigger opportunities are at >> awesome well let's take um maybe we can take a couple of potential solutions. One of them, um, you could use any user group you want and maybe one of them could be literally you given that you're a Gly user. Uh, happy to go in either direction, but I I want to encourage your creativity as well.

So, >> yeah, poke at a couple potential solutions without me like digging into actual Grammarly metrics. >> Okay. Yeah.

No. Yeah, totally. Um, so I'm I'm making some I'm going to start.

This is where I will start making assumptions. Um, and I will assume Grammarly is actually pretty strong in acquisition and initial activation. So there is a lot of free users and the free users are using the most basic probably just the original uh the basic version.

Yeah. >> Yeah. [clears throat] The correction.

Um, but the premium conversion is pretty low. uh I think you mentioned it is a pretty small percentage and we also see high churn happens within 30 days. So I think I'll hone in on these two stages and um I think when we're looking at freeto paid um some some uh just by looking at the experience some gaps that I notice is that as a free user I actually don't even I'm not really even fully aware what's the paid value.

So, I would just keep using the free the free corrections, but I don't even know what what are the pro uh the pro features. So, I think one opportunity area is to just further highlight what's the pro value to these users and get them experience it in real time and can so they can experience the before and after and experience the aha moments early on in their journey. So, I think that's one area and um yeah, I can I can go in specific tests.

>> I think that area is great. I will have some follow-up questions for that, but if you have a second uh idea to toss out there, that'd be awesome, too. >> Okay.

Yeah, cool. I can I can I can start with just specific tests. So, say like the the the test here is basically give users a pro taste.

Um so um as I mentioned the problem right now is that the user is unclear what's the pro value. They are not even getting to the pay wall to see to start the trial because they don't even know what they're paying they're going to pay for. So um one example one test example what we can do is that to show what pro can do before and after in real time and kind of give them the taste of a pro value and limit that to say like five times.

So um and before I do that I'll also do some data analysis to understand what are the top features that is positively correlated with expansion with paid and imagine after that exercise one is rewrite the whole paragraph which is the most um the feature that's most successful in converting users. So um right now if I look at the experience um I know we don't have a visual in front of us but there is some sort of pop-ups that shows you hey pro suggestions and then it's kind of a blur um preview. You don't really see what's being changed.

Uh my idea is that literally show users what it can what will be the after and then they can they have one click just to show if with one click they'll a they're able to see what is the after experience and they get free five free tries and on the very last one um we show a badge you have you're running out of your free credits and this is um do you want to continue to upgrade so I think that very last screen, that very last try is the most important touch point where we really want to capture them there. Um, so that's something I think we can mark out quickly for you. >> Makes sense.

Yeah. I don't know if you you want to do that live or not, but before that, let's uh help me paint the picture of like the the user story here. Like what what type of user are you imagining for this?

Um, what application are they writing? How would they actually encounter this experience that you're you're describing? >> Yeah.

So I think if uh imagine I'm the user I'm the product manager who is say like writing up a product requirement and um I yeah like I want to get something done really fast like I'm already late on my product I should be getting the I I need to send this out today right um so I think I'm working from Google doc in this case and Google doc my Google doc is already connected with Grammarly and as I was writing game. I think um I start seeing the different types of nudges and I think I do really want this nudge to specifically jump out that says rephrase your rephrase your um whole paragraph within one click and um I think what we can do probably in that nudge is also what's the end goal like how how can we help you improve is it improve the clarity or is it improve the tone I think for me uh for as a PM The importance is the clarity because that's requirement that I'll send off to engineers. So I think maybe I think this right now we're visualizing what are the >> that's why I asked the question to understand.

So I think maybe the nudges improve clarity of your requirement with one click and then um after that one click I see the whole paragraph updated and um and it's like a shorter more concise version and I'm think great and then immediately I also see one credit already used or one free trial already used and um as I move on through the documents I can probably easily hit five tries and then at the very last one then it shows you used up all of your free tries. Do you want to upgrade now? And I think we do have a free trial, right?

Like unlock like upgrade for $0 um to to to continue to use this feature. >> All right, fantastic. Um so let's imagine we go with this idea.

We want to experiment because we want to understand the relative effect of this change versus the you know pre-existing experience. So walk me through like how you might set up an AB test. Um what metrics might you try to measure?

>> Yeah. So I think um starting with the metrics that we want to measure, I think the core KPI here is the freeto paid conversion rate and also uh related to that is the uh ARR coming out from this experiment. I think the secondary ones um will be looking at trial to paid conversion rate and um the free use the the free taste usage because I think what can happen is that a lot of what we what can happen is that say a lot of people start um start the trial but we notice that this cohort of people who start the trial through this feature actually turns um actually cancels their uh trial right after or they term they have high turn rate early on compar higher than average so I think we need to continue to like that's why we need these sort of guardrail metrics on top of the freeto paid conversion rate and ARR >> yeah what's the kind of biggest risk to this in your mind that uh you know if implemented that doesn't work why why wouldn't it work >> yeah I think an example is that um they do use the free they they do use up the well I think there are two issues one is that um say they did use up the five and they did convert to trial um but the trial actually didn't conver convert to um a free a paid subscription and um I think the hypothesis of that is that it was a one-time usage it was a short-term usage and they don't see the value of continue to use it after the trial ends but I think with that there can be different tactics um that's focused on the continued usage and make them see the continued value.

And another issue is that they just use the five free taste and they never even convert. They see the they see the pay wall but they don't even um start the trial experience from that. They don't like I think that's um an even like part of it is short-term usage but also I think it's similar problem but I think we just need to resolve that through different touch point >> in a way it's similar to like having a free trial in the first place right you're kind of pushing the upgrade payw wall like farther back so users can experience more and you try to find that sweet spot right where as a free user you you do want ideally to you know love the value you want to build a habit around it such that you want to upgrade but you also don't want to give away so much such that you're really harming your business.

So finding that sweet spot can be hard. Um, okay. Any other thing that comes to mind when it comes to experimentation like the setup of it, how many variants you might take, like the users you might target?

I don't want to give you too many answers, but just like anything that comes to mind in terms of the setup itself. Yeah, I think um the setup itself on the technical side maybe there are issues on I don't know if there is any sort of technical blockers in terms of unblocking the the feature the the paid feature itself um and how do we show and when we design the experience how do we show the before and after in a clear way that's that that shows the value because I think sometimes when it happens is that because the one-click experience is so easy for them. They sort of forgot what they had before.

So, how can we actually show an experience of clear before and after? I think that's that will be valuable from design perspective. And also I think something else we talked about is that how do we incorporate the goal um how can we actually help the users in the experience itself too is I think we talked about in this sense it makes more like it's um because I want concise uh the efficiency and the clarity that's one thing that's something that we want to call on the experience but it could be different for students so I think the finding the right cohort for this um >> yeah to to that point right like we talked earlier about the difference between students and professionals and imagine you run experiment and this is like a great solution for professionals and like a really crappy one for for students.

Just how do you think about that trade-off of like only shipping a monetization you know tactic to a particular user cohort as opposed to everybody? What are the kind of pros and cons to that? Yeah, I think um from product perspective I think generally you want you want whatever that is relates to one specific cohort scalable to the rest of um the cohorts.

I do think in this case for student versus um business pros it could I think the experience is or the nudge itself is different but I do think that idea of showing the before and after of a one-click easy experience I think that baseline is is is should be built as scalable so that when we are building it for student later on it might be a different nudge but it should be able to use the same product platform like the same um component. >> Yeah, I think that's smart. Makes sense.

Could be a different copy or different number or something, but the general construct could be applied globally. That would be great. Yeah.

Um Okay, awesome. Maybe adjacent question to all this like let's imagine you build this or I guess you know one of the challenges growth teams often have is they try ideas like the one you're suggesting but you run into a core product team that owns these you know suggestions or whatever and they're like you know why the hell are you messing with our features or you know h how how do you think growth teams and and core teams can best you know collaborate and integrate in a way that's healthy. >> Yeah.

No, that's a that's a great question because I think yeah, it's a we encounter that a lot in both I think especially at big tech companies. I think actually one of the reason why I suggested this test is that this is this is bas ramping up a complete new feature, right? We're not ramping up a complete new experience in a sense that this is what the product is already doing.

we were just not showing it to um for users at the beginning. So I think from technical perspective it's already lower it it should be lower lift for the product team. Um so I think again but like that's one angle but maybe take a step back what we want to go for generally as a growth team is that we want to go for the high impact loweffort ones.

start with these and really start showing impact before we asking for more feature changes from product teams. And I think that's generally the motion that we go for is that we start with MVP tests that shows impact once we get the balls rolling and then we can move on. We start small and then we move on to bigger and bigger tests.

So I think this one to some degree is um again I don't know what's the technicality on grammarly side. I don't know how things were built but from my perspective this is something that it should be easier to implement and um I think I will also work with the engineers to understand what are let's not build a whole component right off the bat if if there are some sort of hacks that we can do as an initial test and if we can create a cohort that makes the most sense in this case like this pro in US um in a specific application maybe we just focus on Google doc um and we start small and to see and also that will help with the test design right because that narrows down the noise in the design as well so I think that's usually how how I would at least start the collaboration with a core product team um [snorts] and then once you prove the value then we move on to a bigger task and more like a more collaborative road map login all that >> yeah I I find like the bigger company that you work at the more like stakeholders that you're kind of dealing with, but you also have a lot of users, right? And so you can experiment and you can segment and you can test these like little cohorts in a bunch of different ways so you don't step it as many toes, but you can still learn quickly.

Yeah. So yeah, that makes sense. >> And I also think sometimes um one one principle that I personally follow is sometimes you do step on people's chest a little bit and you just do it if you can do it and apologize later.

uh >> tell me more about that like how's that worked for you? Yeah, I think sometimes you do need to take uh I think as a growth team at least like my personal opinion is that I I would do the most to the extent of what I can do to get something alive and um with the integrity with with the integrity of uh data measurement because I think that's the most important thing at the end of the day is that how do we understand how do we measure the impact of tests um but I think once we get that like once we get the numbers it's easier to get the balls rolling Makes sense. Yeah.

I think typically companies will have some sort of like lines that you definitely don't cross around, you know, user privacy, security, stuff like that. Yeah. You know, yolo, whatever you are.

But the the concept of quickly learning, trying stuff, being bold in certain cases, but quick, right, is is important for growth teams. Yeah, that's great. Um, speaking of being bold, right, I think let's imagine that instead of you're trying to, you know, 20% or even double subscribers, you're trying for like a 10x growth thing here that typically is not going to come through like some AB experiment off of existing feature, right?

So, help me think through like how that would change your thinking uh whether there's a different part of the funnel or whatever, right? Like how does that impact your approach to solving this problem? Yeah, I think when um when we're really looking at that kind of growth, I think in general my my mind will go two ways.

Either we create a new product or we really increase awareness. I think um to to to reach a new cohort of audience that you in the p that you typically typically haven't done in the past. Um, I think one example that I could and since I'm more on the PM side, I think we can think about how we can sort of le create a new line of product almost and um associate it with whatever that's happening in the social world because I do think um something that's interesting about Grammarly is that I was looking at this weekly insight that I got which is really interesting that tells me about how I write like um what's my writing style?

Is it more confident? Is more formal? Is it more um casual?

Things like that. And that almost makes me think about the whole MBTI task that's happening. So like it's sort of relating your writing style to your personality style.

>> Oh yeah. Uh >> um I think that's something and because what's happening in I think this whole MBTI has become somehow it has become a great conversation starter for everyone now all the conversation that I go to like people are talking about what was your MBTI and I think people are just getting more interested in learning about themselves. So I think there is something about Grammarly that helps you understand your personality style and also how you can grow.

So it's almost like a personal growth um um tool similar as what Dualingo is kind of doing. It's not just about a tool, right? It becomes part of your lifestyle.

I think there is something that Grammarly can consider is that how do we how do we build a new this mascot or this line of product that is more actually about u growing as a person uh versus just a writing tool. So I think there is something there. Um that would be interesting.

Yeah, I think that's one angle. And I think another angle is um for kids talking about how do we tap into a a new cohort, right? Like how can we help teach kids better writing?

And that's like a completely different like um audience group is K12, right? I think right now the students are probably more like high school or college students. If we're say if we actually recommend this product to a parent and help the parent um help their kids to write better, I think that's another line of product that we can think about.

>> Makes a lot of sense. Um sounds good. Yeah.

And even in Grammarly's trajectory, right, I think they initially started out tailored towards students. I actually could be wrong on this, but I think it was students and it was a solely paid product for a while until they realized like actually a ton of you don't need to be a a student or someone that's like solely working on pedagogy to take advantage of a writing assistant tool. This could be useful for professionals.

Then they open it up to be a premium product and that dramatically increased the you know the user base and the the TAM if you will. And then you know certainly recently like they rebranded as superhuman and they acquired KOD and they're becoming much more of this like AI native. Yeah, >> productivity suite which again is an expanding of of the market and the go to market motion and things like that.

So I do think companies have these like scurves in a way right where you can you can try to grow and do these like local tests to to make it work but at some point right you find your your second act and you find ways to dramatically grow your audience and that's >> it's interesting and uh yeah it's hard to do importance >> lots of investment. So awesome mock interview. Uh Albert, you've asked this question a ton out in the wild.

Uh I'd love just like a summary for what you think is important. Like what should people know about this this this problem? Yeah.

>> Yeah. And overall, I would say Katherine did an incredible job. Uh clearly a very experienced and just a natural conversation throughout, which is honestly half the battle is like good vibes, a good conversation, easy to talk back and forth, so it doesn't feel too rigid.

So I want to commend you on that. Um, as to like this specific type of question, I think this is like a kind of a typical growth leader question in particular. So, less so for for junior growth folks, but um the reason I say that is because I gave a very big overarching goal of like doubling subscribers, right?

And when you work on a product at scale, there are usually many different levers to grow the the end result of like a revenue or active user or subscriber metric. It's usually not some pinpoint solution. And unlike core product, growth product, people usually think in terms of user journeys and in terms of metrics.

So I thought Katherine did a nice job of starting there of laying out the funnel of, you know, you usually acquire users. You know, Katherine mentioned about free to paid conversion as another step at products like a Grammarly or a Dualingo. you often have a pretty long like free user journey in and it of itself which means like you want to like activate and onboard users effectively and then keep them retained for a while.

So that's usually part of that growth model if you will and then when you get to the monetization side you convert them into subscribers and in a PLG PL like productled growth productled sales type of model these um single consumer buyers like think of Slack or Adobe you know as examples right um they can be expanded into team and even enterprise accounts as well and so that funnel can be very very broad and depending on what type of company you're interviewing for you'll want to Make sure you go into that interview with an understanding of what their growth model is and what those stages are so you can speak to that eloquently. Um I think the second step of understanding the user is also where I would recommend going and it's more than just breaking down the user segment into a bunch of different like you know micro components but it's what are the the main first what are the highle ones which I think student and professional was good but also what are their specific use cases within the product you're working on what are their pain points what are the key differences um so talking about for instance students is scary about plagiarism and you know having long form you know I filled in some of those blanks right but painting that picture of that user story in a really concrete way and talking about how those distinct user segments are actually distinct from each other and not using the product in the same way. Sometimes I see like when I interview people, they'll name like all sorts of different user segments, but their relationship with the product isn't actually that distinct, right?

And so when it comes to how you're actually going to build solutions tailored toward them, it's not that helpful, right? And so you need to name the segments that actually matter for that product. If you don't know, you can throw out some assumptions and just clearly ask that or, you know, put that question back to the interviewer to help you out, right?

Because I think if you don't have that foundation, it's going to make the rest of the problem way harder to to solve. Then I would say especially in growth interviews they tend to be very metricsoriented. So I thought what you did was really nice where you talked about you know activation rates could be measured in like seven days and paid conversion can be like you know there's uh sometimes a free trial and there's trial conversion etc.

Um I think in in the interview because we're both growth people we just threw out the terms at like rapid fire mode and it was fine for like this type of thing. Um, but I can definitely understand where, you know, if you're a growth oriented person, maybe doing growthy stuff, but your interviewer may might not be a growth person, you might want to slow down a little bit and describe what these metrics actually mean and how you might calculate them and why they matter, etc. Um, I think in our mock interview, I just zoomed by it because I I do understand that stuff, but I think just being cognizant about who your interviewer is such that they can um clearly grock what you're trying to say there, right?

That's important. Um then when I think you jump to the solution part and you kind of said like the assumption is gramly strong in acquisition activation but maybe there's opportunities in premium conversion. I think that's fine but I think it could be strengthened if you tie it back to like the users and use cases and metrics a little bit to tell like a cohesive story, right?

Because you kind of just jumped straight through to like I'm just going to focus on this part of the funnel, but you just lost a little bit of like the why you chose that to begin with. Um does that make sense? Okay.

And then on the solution part I thought um yeah like I'm very familiar with this specific problem obviously. So I it was again it's intuitive to me. I think when it's less intuitive to the interviewer telling when I ask you the question about like paint this in in terms of like a user story like you know you're a product manager you're working on a PRD you need to send it out today you're anxious about that you know you're writing in Google Docs like making that come to life in a like mentally visual way really helps the interviewer clearly understand what you're what solution you're proposing because the solution in itself even if you describe that solution really concretely, you also need to explain to the interviewer that this solution is important and they can understand like why it solves a like a thing in their life and believe that it's going to work for them, right?

Um, so telling that story in a cohesive way I think is just really really good. And when I asked that question, you did a really great job of doing that. So, I think you you had that.

>> Well, I really appreciate that wars me to take a step back because I I think that Yeah, that's the part where I realize I'm missing >> missing I'm I'm jumping ahead a little bit. >> Yeah. No, it's all good.

[laughter] Um, and then we we jumped into the experimentation part. I thought you were strong here. This is the general framework of you have a target metric or core KPI as you described it.

This is like the main purpose of running this experiment. Um, if that's wrong, then your hypothesis is essentially invalidated, right? You have a main thing that you're trying to do.

But experiments are often, you know, it it tests your depth of thinking in that especially when you work in growth, you're actually going to be quote unquote wrong like, you know, 50% or more of the time, right? And so you you should go into it thinking of it as a hypothesis, which means like a lot of times your idea is not actually going to be right. And so you should have an understanding of like that's why I asked like what are some of the risks of these?

What are some of the guardrail metrics? Like if this hypothesis is not true, what should actually happen to other adjacent metrics that are not your target one, right? But things that might actually go down.

And so you want to measure that experiment holistically where you have the target and then the the guardrail things, which I think you did well. Um I also liked some of the nuance questions which I I don't typically ask. um like junior or like mid-level candidates, but I think more senior folks are are great at this.

the things around like let's imagine your experiment is successful for a particular like type of um cohort of users like how do you manage that tension of like and I thought your answer of building a platform that is more kind of global but allowing for some personalization in in current certain spots was like nuanced and and good and I think speaks to your experience in this space. Um, and then other types of curve balls that people might get tend to be on the okay, you know how to do this growth thing, but are you like uh are you going to be good at collaborating? Are you going to know how to handle like relationships and tricky situations?

So sometimes there's a bit of a bent of like behavioral questions and things that that can flow naturally from this type of interview because essentially it's a it's a product sense interview at its core but it's very growth oriented in that I ask for metrics and experimentation but then it sort of like naturally flows back into the okay let's actually imagine you're in this role and you're going to run into these like thorny things. It could be the CEO doesn't like it. It could be, you know, maybe you moved your metric way up, but you moved some other team's metric like way down.

Like, how do you handle that? Like that those types of conversations are tricky. And they're typically isn't unfortunately like an answer that works in all scenarios, but it's helpful that you take a, you know, your your relationships oriented, but also like you're trying to drive impact.

You find ways to strike that right balance, right? Or as an example, you mentioned the ability to segment in a particular like user cohort in a particular country, like you know, find ways to test that stuff really quickly. Um, so you're learning quickly and not investing, you know, multiple years to do something and step on toes in the wrong way.

Um, yeah. And then lastly, you know, I I thought the question around like the moonshots 10x thing is also interesting because it in in certain ways is kind of like a jarring question to ask because you've gone so far down a very specific solution rabbit hole and then you're told to just like forget about that and take a completely disjoint approach. Um but it's good and it's actually like not uh so uncommon in the day of the life of a growth leader like sometimes you are are supposed to fly at you know 30,000 ft and think about big picture next product things that you need to do while tactically you're also working with your team on like you know very specifics of on experiments.

So, I do like to sometimes this doesn't happen in the same interview, but I think it's nice when the candidate can show their breath of like bungee cording between the two altitudes. >> Awesome. Um, so if this were a real That was perfect.

Literally checked all the boxes. I didn't have to ask any questions. Done like a true super senior person who gives me what I want before I even ask for it.

That was amazing. Uh, this is the last question. So, just Yeah.

Um, if this were a real interview out in the wild, um, you would make a hiring decision. So, what would be your hiring decision and why? >> Definite yes.

[laughter] Um, no. I mean, Katherine was really, really strong. Um, usually there's like a rubric with this type of stuff, right?

And so, I thought in 30 minutes is not really that much time. It's actually more common, right, that you there's like I don't know five to eight different interviewers that will test like a particular dimension. Um, but just your ability to kind of jump across a variety of different uh dimensions of that I thought was impressive.

And uh yeah, no, you did a great job. >> Yeah. >> Awesome.

>> Yeah. >> Cool. All right.

Thanks everybody for watching and good luck in your interviews.