What is a Stakeholder

[00:00:00]
>> So far, I've just been talking about what a PM is and isn't. Let's get into, what I think PM should be doing. This is roughly ordered, is this true? Not totally, it's ordered in a way that I think I can draw a narrative through most of it.

[00:00:18]
But I try to keep the most important things up front and basically the most basic building blocks and then go up from there. So, we're gonna hop into stakeholder management now. Why did I put this first? Because product management, as I've said before, it's about people, and it's always about people, and if it's not about people, then you're probably not doing your job correctly.

[00:00:43]
Because ultimately, at the end of this, always should be a person. Even if you're building, I don't know, a piece that goes into a device that's then sold to someone, that goes into something that then ends up in a user's hand. You need to be concerned about everyone in that process and all of the people, the people factor, involved in all of these different pieces, including the end user of whatever you're going to be doing.

[00:01:05]
If you're not thinking about that and you're only thinking about the dimensions of your piece that goes into that, I don't think you're doing your job as good as you could be doing it. It's all about people, and so we're gonna talk about stakeholder management. And a stakeholder is just another word for person that cares about whatever you're doing.

[00:01:24]
So, let's take Google Maps for example. Fathom that you are a PM, that is the product owner, of Google Maps, Maps likely has dozens of PMs that work on it. But let's just simplify for a second and say that you are the only PM working on Google Maps.

[00:01:42]
Think through for a second everyone that cares about what happens on Google Maps, everyone. Well, let's talk about end users first, first of all, there's many types of end users for Google Maps. There's the navigating user, which is probably what most of us thought of first. But think of all the different types of navigating users, you could be navigating on a car.

[00:02:03]
Which is gonna be way different than navigating walking, those are two different users. Even if those users like, I'll walk one day and I'll drive the next day, that's just me changing my perspective on them. But I still care about both of those personas, I am both of those personas.

[00:02:19]
And it can be multiple personas at once, right? So, cars, biking, transit, rideshare, walking, those are all different types of users and you need to care individually about every one of those personas. I open Google Maps all the time just to find the hours of businesses, right? That's a type of user, that's a valid use case of your app, that's a stakeholder in your app.

[00:02:41]
How far away something is, I will check how far would it take me to drive to Fargo from here or something like that, right? Some of this was looking at geography, I've also zoomed out where the hell is Armenia? I don't actually know where that is. And I know was zoom out, my math was a cost there, I didn't know that.

[00:02:58]
Finding where their friends are, I've definitely used that before. Third party apps that embeds maps like Uber embeds Google maps. And if I'm the PM of Google Maps, I really need to care about that end user because now if they get upset at Uber because of their maps, Uber is gonna be upset at me.

[00:03:19]
And that's, those are all stakeholders that I have to care about. So, in that light, third party developers using Maps, you have to care a lot about their experience, both with the development of their app and the end user experience thereof. Business owners, they have to go in and manage their hours, their menus, do they accept DoorDash, are they kid-friendly?

[00:03:42]
And those are frequently very non-technical users because they're gonna be like mom and pop bodegas. So, that's another critical stakeholder in what Google Map does. And then you have to care about all of your internal users, your team has a stake in your app, your manager, your manager's manager, teams in which you depend, right?

[00:04:02]
So, if you are depending on Google infrastructure, that team, they're trying to solve your problems and therefore they have a stake in your app as well. The DevOps team that manages that, the Google shareholders off, obviously. We live in a capitalistic society, so they also have to care about your app as well.

[00:04:24]
So, my point here is, you have a lot of end users, you have a lot of stakeholders. And whenever you're building something, you need to think quite a bit about that. So, this is kind of an exercise of breaking down personas and stakeholders and you can kind of think of in terms of how I'm talking about stakeholders and personas being similar concepts.

[00:04:48]
Why do we care about any of this, why do we care about all the people that are using this? Because we're building products for and with our stakeholders, and we need to be incredibly mindful of the things, the directions that we choose to go. Because inevitably, to some degree, you are always trading off this stakeholder versus this stakeholder.

[00:05:08]
Even if it's like it's a slam dunk, this is the best feature that we can build, you are still trading off that your team has to maintain this feature long term, and you need to consider that, right? Like if you build a whole new service that has a new database and it has a vector database that then talks to a large language model, that's just three new things that you just threw on your infrastructure.

[00:05:30]
And now your infrastructure team cares, and your team has to care, and your team has to get paged on it. And if you're not thinking about this, you end up with an app where you're getting paged constantly and no ones doing anything about it because you didn't manage your stakeholders correctly.

[00:05:46]
So I think this is best rather than, I'm kind of talking in the abstract to just kind of take this and make this more, try and make this concrete. So, let's say we get a directive to make more money for Google Maps. Just pause for a second here and just think about, all right, you PM of Google Maps, you need to go make more money, how are you gonna do it?

[00:06:09]
I'll tell you what I thought of, sponsored listings, right? So, right now if you search for something on Google Maps, it'll show you one or two sponsored listings, why don't we increase that to five? It seems like a pretty, now we have more ad slots to sell. But let's talk about what did I, either explicitly or implicitly, just trade off to go from two to five sponsored listings?

[00:06:34]
So let's say I'm searching for something like coffee shop and I'm going to allow Starbucks. Which is a mile away to be listed higher than a half mile away this little boutique one that would have been on my way. Explicitly what I'm doing there is I have a ranking algorithm that would have put, let's call it Herkimer, cuz that's the one I live next to and I love their coffee.

[00:07:01]
That would have put Herkimer first and my ranking algorithm would have shown that first, right? Instead I'm gonna put Starbucks higher than that because they gave me some money to do that. I am explicitly giving the user a worst experience, I had an idea of what I think is the best thing to get this user and I am giving them something inferior in the name of taking money.

[00:07:22]
I have traded off that stakeholder for a different stakeholder, is that okay? I don't know, right? So you kinda have to ask yourself, what's my impact here versus who am I affecting? But that's the obvious one. Let's talk about people that are already paying for sponsored slots, right?

[00:07:43]
So, if Starbucks is already paying for that, let's say that I have another Seattle's Best Coffee, which is actually still Starbucks, but let's ignore that for a second. [LAUGH] That Seattle's Best comes in here and they buy one, and they're two miles away. And they paid even more money, so now they're ranked even higher than Starbucks.

[00:08:03]
I have now devalued the existing sponsored slot, right? Which probably means I'm driving down my price on those slots as well, because now it has become more cheaper, right? Because there's more of them. But, I mean, the other thing you can look at, is we can help net new businesses, maybe we have a brand new coffee shop opening.

[00:08:24]
And they open, they put a sponsored listing in there and now they can get the word out, they're nearby and they can open that. Obviously, we're harming even more businesses, like now Herkimer doesn't pay at all for this. And now they are the sixth result on Google, no one looks that, we have science that says people don't don't look six results down, right?

[00:08:46]
So that really sucks. So, I mean, actually, this was actually a real story for me when I was writing this. This was actually just Brian ranting right here, this whole section right here. I was on Instacart, did I say that here? I was on a popular grocery delivery service.

[00:09:06]
[LAUGH] And I was gonna give a gift to a friend of Tequila, right? And I was so frustrated with their search results. I went through and I counted, there was 20 results. 50% of those results were ads, and 40% of those results were not tequila. I would have searched for hard seltzer or bourbon if I had one of those things.

[00:09:32]
But I was just so pissed, I was like, I couldn't find a bottle of tequila. I'm not searching for tequila just cuz I like, I wanna get drunk, right? I mean, sort of, but that's not actually why I was doing that, right? So, I really felt that popular delivery service had sold me out as a stakeholder, right?

[00:09:51]
So, in the situation should you do it, right? Do you think this is a valid way for Google to make more money? I'm gonna argue that based on my just product sense, my intuition here, I actually don't think this is likely to increase profits that much. And certainly not at the cost of how much it devalues the experience of both existing users of the existing businesses and for the end user of people using it.

[00:10:13]
So, yeah, the metrics matter here, we have a whole section on metrics and we'll get there, yeah.
>> So when it does come to those more gray scenarios and then again, maybe we'll get to this later in the workshop, but are there ways to refine that gray area?

[00:10:34]
So, I'm trying to think of how would you go about quantifying the impact to the UX and distilling that down into, here's how it's going to impact our bottom line, potentially. And why showing more ads might increase the bottom line by one and a half percent. But it's going to degrade the UX in such a way that we actually lose money, there's a lot of factors-

[00:11:05]
>> Yeah.
>> In that calculus, so [LAUGH].
>> So, I'll give you the PMing tools to do it, but at this point, you're starting to get into game theory and stuff like that.
>> Yeah. [LAUGH]
>> So at that point, I involved the data science team cuz I am not a data scientist, right?

[00:11:18]
They're gonna be able to help me do, and probably the accountants and finance people, right? Cuz they're gonna be able to do projections and stuff like that. The tool, as a product manager, that I use, and I talked about this in the prioritization but it actually applies upon a lot of places not just in prioritization.

[00:11:37]
There's tools, one of them is called the RICE method, which is Reach, Impact, is it, kind of gotta remember, lemme just scroll down in here. This one here, Confidence and Effort, right? Using that, you can assign numbers to all of those things and you kinda get a score at the end of, if we do this, we reach this many people, we're this confident that we're gonna reach them.

[00:11:58]
It's gonna have this kind of impact on them and it's gonna cost us this much to get there. And it kinda helps you rank, do this first, then do this, then do this, and don't do this, right? Or you can just think about as impact versus effort, which is just the really simplest way of thinking, is like, is that high impact, high effort?

[00:12:18]
Or is it high impact low effort or is it low effort, high impact? Those are kind of things you can use to think about it. But in this particular case, because of all of the money and stakeholders and what is a three-way marketplace, how does that function? Generally, you need data people to help you out.

[00:12:35]
Or at least I do because I'm not smart enough to handle those kinda things.
>> Sure, yeah. Well, that's a good kinda coarse grain, next level down, how you can justify this is, my intuition leans one way and here's a way that I can frame it up that helps support my intuition.

[00:12:52]
And then, yeah, if you need another level of detail, then good.
>> And another thing I'll just like kinda throw out here is that, I'll experiment frequently, right? So, if let's say that we do think this is actually gonna be a good idea. I will say, we're gonna try this, we're gonna try it for a month, and we're gonna turn it off.

[00:13:14]
And then we're gonna look at all the data and figure out, what did we actually do to the products, right? And so those kinda low cost, quick experiments, cuz Google Maps has such a user base, you could probably get really good results really quickly on that. That would tell you really strongly, do we stand actually to make more money here?

[00:13:33]
That would actually be a perfect use case for experimentation cuz it should be pretty low cost to just say, instead of showing two, show me five now and see what happens. Did we lower CSAT? Did we increase revenue, did we? Are we sending people to the businesses that we think they should go to?

[00:13:50]
Those things, right? So, experimentation is also super valuable here. How do you frame this to your data science team to get the best answers out of them? Certainly one of the things is I try to have my data scientists in many of those kind of product planning meetings so that they have all of the context that I have.

[00:14:07]
I try to involve them as early and as frequently as they want to be involved. With generally, I work with our awesome data scientists, and then we're kinda data science platform. So, they're generally very interested in the things that I build. I have a user research team right in front of me, I just go ask them, it's like, hey, how did you not streamlet today, right?

[00:14:29]
And then I get to build those things. So, involving them, finding awesome partners and involving them as much as you can is generally just a solve all for many of the problems that you have. That's not always possible, like at Microsoft, I didn't have a designated user researcher.

[00:14:43]
And so I had to go to the user researcher Oregon and make proposals and stuff like that. In those cases I try and make a pretty detailed product spec, I try and walk them through, here's everything that I want, here's my goals here's my non-goals. Here's the metrics I'm looking to affect, here's the metrics that I'm hoping to hold steady.

[00:15:00]
And try and give them a pretty solid basis for them to go and make some predictions and analysis and those kinds of things out of. So, I just try and think like they do and I try and answer their questions both before they have them and after that.