Leave a comment
Get the GH Bookmarklet

AMAs

Oliver Young is a Principal Product Manager at Sonos, where he is responsible for a large swath of the Sonos app experience and driving better, stickier user experiences.

Oliver has been a product manager for nearly a decade at Twitter, LastPass, Bluefin Labs, and Jive Software. During that time he has built consumer software, business software, ad tech, partner APIs, and internal tooling. He has yet to build hardware but is optimistic that given enough time at Sonos he’ll get that chance as well. He is especially proud of making LastPass free, shipping dedicated NBA, Premier League, and TV experiences for Twitter, and killing his own baby: Twitter TV Ad Targeting. 

He has managed small teams of product managers and lead large teams of engineers, QA, researchers, and designers. And boy has he fouled some things up along the way! Ask about his failures as well as his successes.

Experiment-driven software development is Oliver’s passion and he has successfully implemented such a system at LastPass and is in the middle of a similar transition at Sonos. Not surprisingly he is a measurement zealot and believes if you can’t measure it you can’t fix it. Many years ago Oliver was a Sr. Analyst at Forrester Research. 

You can follow him on Twitter: @oliveryoung

  • JK

    Jonathan Kim

    almost 2 years ago #

    You've run tons of experiments for a bunch of different types of businesses. Are there any unorthodox things you now track/measure that most people aren't, and, if so, how did you discover them?

    • OY

      Oliver Young

      almost 2 years ago #

      Hey Jonathan, thanks for the question! Love you Appcuties.

      I've used lots of unorthodox measures over the years, but honestly I don't think any of them would be useful outside of the specific product they were designed for. For example at LastPass we paid a lot of attention to the number of items in your vault and the diversity of content types. It was a huge indicator of product adoption and retention, but would only really apply to password managers! At Sonos we look at the number of active users per household (since the system all about listening out loud in a home).

      • OY

        Oliver Young

        almost 2 years ago #

        Incidentally I think this invention of metrics that are specific to your business is some of the reason companies like Snap get beat up when the file and S1 with what appear to be crazy metrics to the market. They were built for internal use and have a lot of meaning, but don't really translate across businesses.

      • AS

        Adam Sigel

        almost 2 years ago #

        Can you speak more to the importance of multiple users per household, and how you measure this? Are you ID-ing users based on app installs, or some other method?

      • OY

        Oliver Young

        almost 2 years ago #

        "importance of multiple users per household"

        This is based on our core value prop: "listen together". We want to fill that home with music and if more than one person lives there we want everyone to feel empowered to fill that home. No rocket science here, just a matter of taking your values and playing them all the way out.

        As for the mechanics, anyone can install the Sonos app and control the system in their home which is a pretty good indicator of a new user!

  • KL

    Kimberly Lombard

    almost 2 years ago #

    What are the key differences between managing a team that doesn't roll into you (eng, design, QA) vs managing a team of PM's that does report into you?

    • OY

      Oliver Young

      almost 2 years ago #

      Hey Kimberly! This is an easy one; it's harder!

      The PM role is a unique one in that other than other PMs no one ever reports to you, yet you can't get anything done on your own. So you spend your time trying to convince people to do what you think is right.

      As a PM you start with a lot of "decision making capital"; your teams and colleagues will go with you nearly all the time to start. It gets harder if you lose their trust though bad decision making or poor communication. Frankly this was how I lost my first engineering team. I didn't do a good enough job of selling them on the vision. Sometimes you're leading a parade, turn around, and realize no one is behind you. It's a sickening feeling.

      6 Share
    • OY

      Oliver Young

      almost 2 years ago #

      I also think this is the reason -- in my view -- why PMs make great CEOs, but CEOs make terrible PMs. The PM cannot simply tell people what to do, but instead has to convince them. The CEO can just tell people, though of course the best CEOs convince you as well. I've seen a lot of CEOs get acquired, get turned into PMs, and really struggle with this transition.

      3 Share
  • MA

    Melissa Appel

    almost 2 years ago #

    It looks like you’ve done some B2C and some B2B. What have you found to be the biggest difference between the two?

    • OY

      Oliver Young

      almost 2 years ago #

      Hey Melissa, thanks for the softball!

      The biggest difference for me is the relative importance of a given user/buyer. On the B2B side, until you achieve some serious scale, any given customer is so critical for your sales targets you find yourself bending over backwards to win their business. Sometimes the dollar amount isn't even all that big, but it comes up at the right time of the quarter and is the difference between making and missing your target. That's really hard to deal with, and pushes PMs to focus on squeaky wheel product development, as opposed to doing the most good for the most customers.

      On the consumer side the importance of any single user is so much lower you have to think about the most good for the most users; there really isn't anything else to think about!

      What this means in practice is that road mapping, product development, experimentation, and the very tactics you use for B2C products are much more focused on the aggregate. On the B2B side you focus much more on gaining or not losing a specific customer or set of customers.

      3 Share
      • OY

        Oliver Young

        almost 2 years ago #

        "From my experience other big differences seem to be accessibility to customers and prospects for user research (easier for B2C because they are easier to find) and the tendency to assume that what works for you will work for all your customers (more of a potential trap for B2C)."

        I've actually found it much easier to get access to B2B customers. Any sales rep worth their commission is itching to bring a PM into a call. Those squeaky wheels will find you! For B2C you have to make the decision to go out and interview people -- especially since forums and app reviews have some serious bias in them!

        Regarding falling into the trap of assuming you are your customer, that can happen in either case. If you use the product there is a natural tendency to think your experience is representative. Fight it!!

      • MA

        Melissa Appel

        almost 2 years ago #

        That makes sense but it’s assuming that all B2B customers are big companies. A company like HubSpot, for example, serves mostly small mom and pop shops and they have many B2B customers, so they have to act more like your B2C example. From my experience other big differences seem to be accessibility to customers and prospects for user research (easier for B2C because they are easier to find) and the tendency to assume that what works for you will work for all your customers (more of a potential trap for B2C). Thoughts?

      • OY

        Oliver Young

        almost 2 years ago #

        "A company like HubSpot, for example, serves mostly small mom and pop shops and they have many B2B customers, so they have to act more like your B2C example."

        Totally! This is why a lot of B2B companies are starting to adopt the methods and tools of B2C companies. Scale matters!

      • MA

        Melissa Appel

        almost 2 years ago #

        Thanks Oliver!

  • TN

    Tri Nguyen

    almost 2 years ago #

    Hey Oliver

    Talk about that experience of killing Twitter TV Ad Targeting.
    How and why did that happen?
    What was your biggest learning from that experience?

    • OY

      Oliver Young

      almost 2 years ago #

      Hey Tri, thanks for asking! I love talking about things I've killed. At one point my boss at Twitter was calling me the undertaker . . .

      The reasons we killed TV Ad Targeting (which was a specific flavor of the TV Targeting ads product Twitter still runs today) were fairly simple:
      * It was too complicated
      * Not a lot of advertisers used it
      * It was costly to run and maintain
      * We were confident we would not lose the spend in the product, just shift it

      The reasons for keeping largely focused on:
      * It already existed
      ** There were folks using it who would need to be moved
      ** We already invested the upfront resources to build it
      ** The sales staff was trained, the material was built, etc.
      * This was a big chunk of the very unique IP that Twitter acquired BluefinLabs for. If you kill it, what does that mean about the acquisition?

      The biggest learning for me was that it's often much much easier to birth a product than maintain it or kill it. Once it exists it has a momentum all its own. I also realized that PMing a middling product is the hardest job out there. If the product were successful you'd either do nothing or focus on expansion of the use case. If the product were clearly failing you would have killed it already. The middle case where you are neither doing great nor sucking wind is the hardest; you can make the case either way.

      3 Share
  • AF

    Adam Fullerton

    almost 2 years ago #

    Could you talk about a product of yours where some element, or perhaps the entire thing, was considered a failure? What was the big learning? If another project since shared some traits with that failure, did it make you squeamish or more confident?

    • OY

      Oliver Young

      almost 2 years ago #

      My first real product was ultimately a failure, the Jive Apps Marketplace. Since not a lot of people will know Jive, it was one of the pioneering social productivity platforms, attempting to replace email and other older forms of corporate communication. If it were founded today it would be something like Slack. The Apps Marketplace was intended to allow individuals to augment the system by purchasing capabilities or apps, the same way you can augment your phone by adding apps.

      There were a lot of reasons the feature failed (which by the way does not mean we should't have built it, just that it didn't work out) but some of the ones that standout:

      * We did less validation of the feature and concept than we should have. It largely was a top down dictate and probably could have used some more research before we committed to it.

      * We were trying to do a lot all at once, including adding a way for individuals to purchase within our app, which is complex and required a lot of internal buy-in from folks like finance. That buy-in was not always there.

      * It's really hard to build a two sided marketplace from scratch!

      The thing that I look for in any new feature or project is how we can scale incrementally. Sometimes you have to take a really big swing, but most of the time you would rather take baby steps to get where you're going.

      3 Share
  • PH

    Pradyut Hande

    almost 2 years ago #

    Hey, Oliver! Glad to have you here.

    Creating satisfactory app user experiences depends a lot on data-driven marketing today. Could you talk about how Sonos goes about creating and running omni-channel user engagement campaigns to power more adoption?

    • OY

      Oliver Young

      almost 2 years ago #

      Hey Pradyut, I wish I could tell you! I actually had to Google was omni-channel user engagement campaigns are!

      Joking aside, I have relied on my marketing team to bring users to the front door, then as a PM focused on converting that interest into first an active user and then second a healthy user. Personally I think this division of labor is the right way to go.

      Where we must collaborate (along with Product Marketing) is figuring out what the right target user actually is. Sometimes that means adjusting your process to bring them to the front door, sometimes it means changing your product to better serve and retain those users, but in either case you want to be SUPER clear about who you all are targeting. If not you're going to be in trouble!

      2 Share
    • OY

      Oliver Young

      almost 2 years ago #

      One other experience worth sharing: when I was at LastPass there was a lot of expectation from management to bring a lot more users into the product. The company had been purchased for a hefty sum a few months earlier and the execs rightfully wanted to pour gas on the fire.

      What I found however was that our new user retention was much lower than I wanted. It's a complicated product so it's not a surprise, but I was very concerned that if we paid $x to bring a user to the front door my product was just going to waste that spend by churning that user right back out.

      Marketing was, in my view, too focused on user acquisition, and not focused enough on long term user value and if we were making a smart investment.

      Though I wanted to hold off on the big marketing campaigns until after we made some progress with our retention issues, I ultimately lost that fight. You'll have to ask the current team if it was the right or wrong call!

  • AS

    Adam Sigel

    almost 2 years ago #

    Hi Oliver, long time listener, first time caller. What's the project or idea you're most proud *didn't* ship under your watch? PMs get pitched all kinds of ideas and often face pressure to ship more/faster. What's the most satisfying "no" you ever got to deliver?

    • OY

      Oliver Young

      almost 2 years ago #

      Hey Adam, nice to see you! I don't think I ever get super excited about saying "no". Though you can get really sick of hearing the same idea over and over again, I'm in the camp that says you should probably say "no" a few times before saying "yes" regardless -- if it keeps coming back it's probably a good idea, if it dies a silent death it wasn't mean to be. (To be clear this applies to saying no to your own ideas as well as those of others. Basically all of my roadmap at any given moment I've said "no" to at some point, even if it was my idea).

      It's been far more rewarding to me to kill something that already existed and you know needs to die to make room for other things.

    • OY

      Oliver Young

      almost 2 years ago #

      For everyone else, Adam used to work for me as a PM. He is just now realizing why I always said "are you sure you want to do that?" to all his roadmap items!

  • DH

    Dani Hart

    almost 2 years ago #

    Hey Oliver - very excited to have you here.
    Experiment-driven marketing is what we espouse a lot here.
    Can you talk more about experiment-driven software development and how/if that interplays with marketing/growth teams running experiments simultaneously?

    • OY

      Oliver Young

      almost 2 years ago #

      Thanks for the warm welcome Dani!

      To echo the response above, I think this is where the communication and agreement between marketing and product about who the target user is comes into play.

      At LastPass we were trying to expand our audience beyond the highly technical, security minded early adopter and decided to target "do it all moms" (note: I didn't name the segments!). What this meant in practice is that we needed to change some things about both the product and our marketing campaigns to attract and retain these users. That coordination was critical, and once we flipped the switch to start the marketing campaigns it was critical to see how these users fared compared to our traditional persona.

      Probably not too much of a leap to say this is where systems that can talk to each other are critical. We spent a lot of time making sure we could follow a user from the marketing campaign to the registration to actual app usage and retention. It was not easy and frankly there were (and probably still are!) huge holes in our visibility, but when it worked it was a much better world!

  • AS

    Adam Sigel

    almost 2 years ago #

    How much emphasis do you put on competitive analysis? What information do you value most and what decisions does it impact? Also: (how) does your answer change if you're at a large company in an established market, versus an early-stage market entrant?

    • OY

      Oliver Young

      almost 2 years ago #

      I think it's imperative for a PM to understand what is going on in their product space and in the market in general. Sometimes competitive analysis is critical and you spend weeks or months dissecting what competitors are doing. Other times you purposefully ignore what a competitor is doing to focus on your ideas and solutions. I've been part of a very early stage product where we discovered a competitor and I personally did not install their product until after I had fleshed out my own thinking to avoid being influenced.

      When I was at Twitter I was on the Twitter TV team, where we focused on the platform as place for TV conversations, actors, show runners, and all the bells and whistles of TV. At the time Tumblr was making big gains in TV and Instagram was just getting into the game. I spent a *lot* of time comparing our experience to theirs, especially when it came to which shows were active on which platforms. That competitive drive was a big influence in how we evolved Twitter and the ways we engaged partners.

      2 Share
  • GH

    Glen Harper

    almost 2 years ago #

    Thank you for joining us today, Oliver.
    From your time with and managing product teams across companies, is there some universal truth about product management that's emerged to you, especially in terms of some characteristics of people, processes, culture etc - without which such teams are just doomed to mediocrity, or worse, failure?

    • OY

      Oliver Young

      almost 2 years ago #

      Thanks Glen, very excited to be here!

      The most universal truth I've found for how teams/products/PMs succeed or fail is "keep it simple."

      It's very easy to get sucked into making a really complicated product and, boy oh boy, I have been sucked into that trap in the past! The problem for me (and I assume all PMs) is that you know all the nuances of the use case, the market, the competitors etc. and it's really hard to focus on the most compelling pieces and set the rest of the noise aside.

      How this plays out is that you create overly complicated products with requirements that cannot be (easily) implemented, features that can't be tested without a dictionary sized set of test cases, a marketing story that can't be easily understood, and a sales staff that can't figure out what to actually sell. The whole process grinds to a halt and you and your product fail. It's a hard lesson . . .

      3 Share
  • JP

    John Phamvan

    almost 2 years ago #

    Hi Oliver

    a. What tools are you using at Sonos for experimentation & analytics right now (especially interested because we're talking about hardware here)?
    Have you added any tool(s) to the stack recently? If yes, why?

    b. Where does your data live, ie what is the "source of truth"?

    c. What collaboration tools does the team use ?

    Thanks!
    John

    PS: If its easier to answer this from the perspective of the company before Sonos, pls do that.

    • OY

      Oliver Young

      almost 2 years ago #

      Hey John, we are just starting the journey toward software experimentation at Sonos, so a lot of this is in flux, but the piece we've added recently is tooling for A/B testing in our apps. We had a hard choice between a "buy" and "build" approach and having built in the past I was a strong advocate for buying. I'm very excited to have someone else worry about the mechanics of running an A/B testing too so I can focus on building a delightful multi-room home audio app.

      At past companies I was really pleased with the RedShift/Segment/Amplitude analytics stack. While I rely on data experts to focus on where the data lives and how we ensure it's flowing right, the people who created Amplitude dashboards on my team were me and my PM team. We all felt VERY comfortable with the tool, could get insights quickly and with high confidence. As a result we used it constantly. You could be certain that there was at least one Amplitude tab open in every PMs browser at any given moment. I would often spend an hour or two chasing a question down the rabbit hole, from chart to chart, analysis to analysis, and come up with some valuable insight I could bring to the team or use to influence my roadmap.

      For collaboration I'm a big big fan of Slack, despite the fact that it can be a huge attention suck. Dropping a chart into slack and talking it through with an entire distributed team at once is super effective.

      1 Share
  • AA

    Anuj Adhiya

    almost 2 years ago #

    Hey Oliver - so cool to finally have you on!

    If Sonos was a totally offline product, how would you measure retention?

    • OY

      Oliver Young

      almost 2 years ago #

      Hey Anuj, thankfully I've never had to deal with offline products! It sounds miserable.

      If I had to I would focus on whatever data I could get about usage, however I could get it. I'd try to start with the add-on purchases that mean you're using the original purchase (think razors and blades), but my guess is you end up spending the time and money to survey the install base. Personally I think that expense would be well worth it if you have no other way of knowing if the product is used or used effectively.

  • MD

    Mark Anthony de Jesus

    almost 2 years ago #

    Hi Oliver

    1. What are 3 key things you feel a Product Manager must consider while prioritizing new features or feature improvements?

    2. If you lack enough data or have no data, how do you push features that are based on gut / intuition / leap of faith assumptions?

    • OY

      Oliver Young

      almost 2 years ago #

      Hey Mark, tough questions! But let me take a swing at prioritizing; I do like to stack rank things . . .

      Three key things to consider when prioritizing (beyond the obvious of cost, and expected impact):

      1. How much is it going to take to care for this feature going forward? Once you birth it, that baby is yours. How much is college going to cost?

      2. Does this take you in the strategic direction you want to go, or is it taking you someplace else? Someplace else may be just fine, but you want to know you're going elsewhere.

      3. Are you just the latest PM to be pitched this idea? I find that most ideas at a company have been kicking around for a while, and especially if you are the new guy it's easy to think 'yeah, this is great!'. Sometimes that idea has been tried and it actually failed (again not a reason to say no, but you want to be aware!) and sometimes other PMs have rejected for really good reasons. Try to get the history!

    • OY

      Oliver Young

      almost 2 years ago #

      Regarding lack of data and going with your gut, the big thing for me is how you will measure eventually, even if you can't measure ahead of time. For example at Sonos I'm working through a user flow right now where we have a strong belief that users are falling into a trap in our UI. We get the feedback in surveys and app reviews, but I actually don't know how many people fall into the trap or how often.

      Ideally I would instrument, wait, and then make a call. But I have enough confidence that we're both instrumenting and building a solution simultaneously. What's key is when we turn on the experience we'll be measuring against something, even if that something is new. We'll be able to assess if it was the right solution to the problem, as well as check if the problem was pervasive in the first place (thanks to A/B testing). If I'm wrong that it was a problem I'll be the first to say so.

      What makes this one easier is that A) its a relatively small change, so the cost of failure is not big and B) it aligns with our overall strategic direction. We can talk ourselves into the approach given our overall strategy. When the thing your gut says is big and contradicts your agreed upon strategy it's a much harder sell!

  • JF

    Javier Feldman

    almost 2 years ago #

    Hola Oliver,
    Do you have any recommendations for how to best instrument for collecting and acting on customer data?
    Would this recommendation differ by business type?

    • OY

      Oliver Young

      almost 2 years ago #

      Hey Javier, I typically leave this stuff to the engineers for the "how", but the acting on it is usually the more interesting part anyway! My bias is to collecting more data than you think you'll need. If engineers and data storage/processing were free I'd instrument every single button and parameter in my app. But of course they aren't so grab the most important stuff first, and over rotate on data collection when you are in the decision making phase of the feature/project.

      As for acting on the data, the best I've been able to do is hire curious, data-minded PMs who will want to get their hands dirty and chain myself to my desk to look at the data. This is why I love easy to use analytics platforms -- any piece of data is going to ask just as many questions as it answers. You want the freedom to chase those new questions.

      I don't think it varies much by business type, though I found that B2C companies are usually more diligent about collecting data since there is no other way to see how your features are used at scale. B2B has a very handy channel called Sales to tell you what's working or not (though you shouldn't always believe them!).

  • DO

    Danielle Olivas

    almost 2 years ago #

    In your experience, how important is company culture in regards to growth?
    What things have you seen done well to contribute to successful teams?

    • OY

      Oliver Young

      almost 2 years ago #

      Hey Danielle, I don't think it's possible to say "company culture is not important" but I also don't believe that every part of a company has to be thinking about user growth all the time. Some folks can and should be thinking about monetizing the users you have, retaining them, driving stickier engagements and so on. What is important is that the folks responsible for growth have the resources and funding to do their job and the cross-functional support where it matters.

      When I was organizing a PM team we had four PMs doing day to day work. One was focused on monetization, two were focused on core user experience, and one was dedicated to new user growth and retention. For our team that was the right mix.

  • JD

    James Dunn

    almost 2 years ago #

    What do you think it is about product management that most people still don't understand?

    • OY

      Oliver Young

      almost 2 years ago #

      Hey James, I actually got into this discussion just the other day with a bunch of PMs here in Boston. The impetus for the conversation was https://hackernoon.com/we-need-fewer-product-managers-50e47dfd95a0.

      I think what the article misunderstands and what a lot of companies don't get is that the role of the PM is to make the hard decisions (or at least recommend the decision!). Too often the role is reduced to requirements writer, or scrum master, or executive whisperer and so on. All those things are critically important, but are in service to the goal of making smart decisions. If you give a PM all those responsibilities but don't let them make decisions you've not empowered a PM, you've hired a project manager.

  • SK

    S Kodial

    almost 2 years ago #

    What is your preferred toolset to inform surveying/customer research before building new products?

    • OY

      Oliver Young

      almost 2 years ago #

      Whatever works! I don't mean that to sound flip, but pretty much anything is better than nothing. Sometimes it's hallway conversations, sometimes it's app reviews, sometimes it's feedback from sales, sometimes it's formal 1k respondent surveys. Yes please to all of them! The only question is how much time and money you have. Sometimes a simple tweet does the trick . . .

  • YY

    Yin You

    almost 2 years ago #

    Hi, I am always wandering one question. How Sonos as a connective smart home sound system manufacturer would compete with the Bose and other giants like Amazon and Google in the field of speakers. As product manager for user growth, how are you able to do to engage potential customers and acquire more market share?

Join over 70,000 growth pros from companies like Uber, Pinterest & Twitter

Get Weekly Top Posts
High five! You’re in.
SHARE
57
57