Coming up next
TEI 277: 24+ speakers share strategies for product managers and VPs to have higher success with Chad McAllister, PhD
And this is where product leaders and managers make their move to product masters learning practical knowledge that leads to more influence and confidence so that you can create those products that customers. There's love and organizations are developing robust data science capabilities. You've probably taken note of this. And they're adding the role of data scientists to the ranks as the importance of data science increases and organizational strategy. Now says and operations it is also impacting our work product managers leaders and innovators and product. Managers Gers are being asked to work. With data scientists now are still kind of the forefront of this and figuring out how product management and data science intersect is becoming in a kind of growing topic to explore this topic. We are joined with to pass guests who've been working at this intersection back in episode one. Seventeen Flea Anderson shared how she was building a Product Management Council at Pitney Bowes and episode zero five five Ritz burnoff shared. How product managers can navigate organizational the challenges for the past year? They've been working together and helping product managers work with data scientists. If this topic is impacting your product work. Yet it will in the future and this is information that you will need. Remember if you WanNa go revisit key concept we take the notes for you. You'll find those notes at the everyday. INNOVATOR DOT com slash two do five eight. That's also a great way to introduce others to this podcast and also help others find this. PODCAST would be wonderful if you take a couple minutes to itunes and and leave a review. That really helps this rank better and helps other product boundaries. Find it now. Let's listen to the insights from FELICIA and rich. Well welcome back you too. This is really exciting. FELICIA and rich. You are joining the podcast once again you both have been guests and today we're talking about the same topic that you have both been working on this notion of data science will get to that in just a moment. I flee shut your head of in this stage of a transition. What's going on since last? We talked doc Chad I recently joined epic. Legal Services Technology enabled firm as via product. And what are you going to be doing for them. It's going to be working with their data science teams to figure out. How do we use data science to improve our products and deliver more client value so very much aligned with other things? We're talking about today absolutely which has to do with why we're actually talking so and you some background in this which rich you're connected to that. So so how did you get connected to Felicia and working on the same topic of data. How data science kind of impact product management? So I've had the great good fortune to work with Felicia at kidney bows for several years on a variety of things around project management training and probably marketing a bunch of other topics and this had come up as something specifically that her teams and the data science teams really wanted to explore so they asked me to lean in a little bit providing training and the public version. That is now in a blog post on my site that depletion. I both worked through a few times to make sure it's it's acceptable for the full public and we'll pry. I catch some of those topics as well and I'll make sure people in the show notes can get to the original posted. You put together there too and we we got connected because of your work there at BP Brittany. I'M GONNA mess this up kidney bows Pitney Bowes for some reason I had the engine maker my head all of a sudden so you never know where my mind's go during these times guys because you were responsible for the all the product managers kind of their care and feeding this council which I thought was such a great effort to be doing for partick arctic managers in leading so. That's a good experience there. Maybe both have good examples. You can share from various work about. How does the data science play Klay these days into product management? What have you seen why? Why even care about the intersection? At least if you WANNA start. Sure well there. There are many many use cases we had at Pitney Bowes. I'll talk about a one part of the business tomor services and where we couldn't when did use it there. The car services business is a big piece of that is fulfillment delivering returns for retail clients. And that means things. It's moving parcels across the country for retail clients and there was a very rich Set of use cases. We're you'd like ideally as a consumer to know win as my parcel going to get to my front porch and therefore a logistics carrier or event Leabeau's would like to predict where are the parcels throughout the network and like a stable which of these parcels is at risk of being delayed for. This is a place for data. Science can really help with this very complex problem. Deliver that customer need of win better prediction of win that parcells going to be at my doorstep. There's another case in the same kind of area. If you WANNA give you one. More airplay. So imagine. Part of that. Business is taking parcels from inbound inbound from international Places I say parcells coming in from China or the UK or other countries when they get into the US they have to clear US customs customs US Customs as well beyond what the retailer can control. Or what a vendor like Pitney bowes can control and it's a very bercy somewhat unpredictable process of of it but a the company processing those parcels on the on the flip side of customs would really like to have a good forecasts of winner these volume one of these parcels going to hit my dot for process. This isn't that a parcel by parcel level. But a volume level so that they can optimize staffing and other resources to process that as quickly as they can again with the goal of delivering the parcel as quickly as is possible unpredictably as possible to the consumer. So they're just to kind of Pieces of the puzzle but in that kind of environment there were just so many different applications applications and data science so it sounds like a key thing. There is collecting data. Obviously to help us understand. What's going on with shipments where things are but then doing whatever the form of data science doing some analysis to help us predict? Wouldn't we might have problems. What are the risks involved in packages? Actually showing up when we think they're going to show up and then one of these larger volumes hitting so you can have people alerted and ready for that so the product part of this has to be pulling the data collecting the data. And then you're doing doing something to analyze that and then put that back into the product to communicate information to people so right right on it can be customer facing and some some places that information was using used in customer facing applications but also there's a lot of internal use of that information are outs managers ears and other folks using that data optimize the business processes internally so when he found use cases on both sides CAC YEP rich. You're a part of this AD. Also I know you've been doing other things too. You have examples you want to add. I do and I think that the one of the trends we've been seeing is that we used Christie is data analytics to give ourselves internal insights that we then hard coded into our applications so most of my work is with folks who build software as their primary revenue source certainly clients and partners the mind likely both where they're using it internally as well but we've really seen I think moving to the four machine learning and natural language processing and other kinds of high end data analytics where we're now building an into the products themselves so so people shipping software N.. Consumers are going to see that are GonNa make judgments or recommendations that really are visible to the rest of the world. The challenge sounds of course here. Is that these kinds of data. Analytics are never one hundred percent perfect. They're better so you've got to think really hard about what all the edge cases in problems are are if we happen to make a bad recommendation or the data's missing or something else rather surprising happens so I see product. Managers have to struggle with less certainty Than we did in the days when everything was hard coded and every time you ran it you got exactly the same answer now. Things might be changing a little with the advent kind of a and what you talked about machine learning natural learning and the like the edge hate cases interesting to consider the reminds me of statistics and the type one type two error thing right and the error of thinking something to so when it's not or vice versa..