Transcript

Paul Ford Can we just hit pause [Rich snort laughs] and point out that your niece’s first word was Google?

Josh Schwartz Is that true?

Rich Ziade [Sadly] Yeah. They have a Google Home. So it’s, “Google, what’s the weather?” “Google, this,” “Google, that.” [Others chuckling] And mom and dad—

PF Honestly, it’s a very easy first word cuz it’s [like a baby] “Goo-gle”.

RZ She says, “Goo goo,” yeah. But it’s still a little bit creepy. I mean we’re— we’re— we’re talking about—

PF [Crosstalks] No one was happy with this—

RZ— a one-year-old.

JS I will say, as a traffic expert, if you write an article about that it will do well.

PF Really?!? [Laughs boisterously] [music fades in, plays alone for 18 seconds, ramps down].

RZ Paul.

PF Here’s the big question— the big question in life; the number one question; the ultimate question of the internet: what is it? . . . Take a guess.

RZ Who’s visiting?

PF That’s a good one. But the other one is— starts with, “How do I get ‘em?” [Music fades out]

RZ Oof.

PF That’s the big question: “How do I get them to look at my thing?” And then once I get ‘em, “Who the hell are they?” And then once I figure that out, “What do—”

RZ “What am I gonna do to them?” [Laughs semi-maniacally]

PF Yeah, “What do they need? So, to help us understand this fundamental aspect of the internet, we have the good fortune to host, here at Postlight, Josh Schwartz, who is Head of Product at Chartbeat. Welcome, Josh.

JS Thank you, thank you, great to be here.

PF Can you tell me what Chartbeat is, start there.

[1:31]

JS So, Chartbeat is an analytics company that is used by the editorial side of the organization at a media company. So if you are a writer, an editor, an SEO person, a social media person, and you wanna understand who’s reading your stories; how they’re getting to them; how they’re engaging with them, you use Chartbeat on a daily basis to try to understand what’s happening, both in real time and also kind of over time.

PF And for a long time, it’s— it’s been the industry standard, right? Like I’m— The New York Times famously uses it, and sort of all the— all the sort of big players because you can just look at it in the moment. It used to be it was every week or, you know, every day you’d run your analytics.

JS Yeah, I mean, I think that’s one of the things that we’ve been really proud of is it’s— it’s a— it’s a product where data comes to you. You login and just immediately there’s stuff that’s engaging for you [right]. So you can take someone who has a day job that has absolutely nothing to do with data and get them data that’s very relevant to them without really anything happening on their end. So that’s been kinda the magic of it.

PF What does it tell them?

JS So the first order of thing that it tells people is kinda this— this in the moment, “What’s happening on my site right now.” So especially if you’re a person who uses Chartbeat all the time, you get an intuition for, “What should my site look like right now in a typical day,” and then, “Is something weird happening,” right? “Am I getting an anomalous spike from a referrer? Does one of my top stories have a typically high or low engagement? Is traffic way off and maybe there’s— there’s something wrong with my site? Or I— something wrong with, you know, with my social media post?” So it gives you that sense of kind of what’s happening.

RZ I mean that stuff’s addictive, right? But what are people wanna do with this stuff? So, you— you— you know, obviously having those insights are great. It’s like real time Nielsen ratings [mm hmm] but what— what next?

JS So, you know, I think the— the key thing that— that we think about is the process of engaging a reader, right? If you think about, you know, a typical reader on the web, um they’re probably not navigating directly to your website, they’re probably using some other platform. And out of the million links on that platform they could’ve clicked on, they somehow clicked on yours and they landed on your site. And now you have this just brief moment where you can kind of turn that user from somebody using Facebook or searching something on Google or using Apple News or whatever the case is, to suddenly somebody who’s engaged with your site and your brand um and so that momentary like hook of the user, I, you know, we think of as kind of one of the key things that we measure. So, um, when you’re looking at Chartbeat um what you’re seeing is not just, you know, how many people came in but how those people are engaging. So, are they actually reading the story? How far down the page are they reading? If they’re on a homepage, what links are they clicking on off of that page? And so on. And we try to surface opportunities to use that data to kind of improve that kinda user engagement. So if everybody who reads the story is dropping off two paragraphs in, there’s probably a problem on the page. Getting a report about it the next day isn’t good enough. That story is— is old at that point. You need that data right at the moment where you can actually make the change at a time that matters.

[4:42]

RZ So is that— is that what happens? I mean—

JS Absolutely!

RZ So people like— a— an editor is going in and just saying, “Gosh, this— this third paragraph is just making people sleepy,”?

PF Well or it could be— it could be a bad ad too.

JS Yeah! Absolutely. Or one part of the tool is a headline testing tool. So people can do multivariate test of their headlines, right? So, we have organizations where basically every story that makes it to their homepage gets AB testing. And, you know, the— the delta between a good headline and a bad headline is just remarkable. You know, you can see a 30, 40 percent lift between a— a good and bad headline, and that’s just a huge difference in the number of people who actually read a story.

PF What else works? So, good headline versus bad headline. What else gets you the people to click on things?

JS When we think about writing a story, the first piece of advice I give people is you have to ask this question of, you know, which audience is this story for? Right? A story even about the same, you know, piece of news, if it’s designed to do well on social media, it needs some sort of emotional saliency, right? The things that do well on social media are things that people want to share and wanna react to because those are the, you know, the metrics, of course, that social media platforms use to— to amplify things. But if something’s gonna do well in search, it needs to be, you know, a factual response to the type of thing that somebody searches for. So that story is gonna be completely different.

RZ You can have the same story taking five different shapes?

JS Oh absolutely! I mean we definitely see people write two or three stories. One tuned to their homepage audience; one tuned to their social [interesting] audience; one tuned to their homepage audie— [trails off]

[6:12]

RZ Are— are writers depressed? [Josh chuckles]

PF Ye— yes.

RZ Like I’m a painter, right? And I— I’m just like, “This is my work of art. It’s a— it’s a babbling brook.” And then you give it to the— to the gallery—

PF Nah! You know what?

RZ— and the gallery says, “Mm, a little more red. I’m gonna just take out—”

PF But that’s— people used to make Hallmark cards too. I mean I’m sure they still do, I just haven’t received a Hallmark card in a long time.

RZ The thing about a journalist who spent two weeks on a thing.

PF Listen, everything is bad. You wanna work that out right here? [Others laugh boisterously] You wanna fix that here, on the—

RZ I’m not pointing blame by any means! This is an observation.

PF This is about analytics.

RZ Paul, we’d have to get into your— [laughs]

PF Um. Well this is— this is tricky, right? So, analytics do drive content in 2019.

JS Yeah.

PF That’s just— that is the reality that we’re in.

JS I [stammers] think that analytics, for me, are a tool, right? Whether it’s for journalism or analytics for, you know, measuring product effectiveness. Analytics shouldn’t be telling you what to write. Um but analytics should be telling you, for something that you’ve written, how to make it perform best, right? That 2000-word piece may make it out to, you know, to— to a million people, a couple million people, but if its job is to be read, you know, by 10,000 people but the right 10,000, like, make the story work well, right?

RZ Yeah.

[7:21]

JS You write the 2000-word piece and people only read 25 words of it then— then you have a problem. And that’s what the data can— can help you kinda figure out.

RZ Also, writers wanna be read.

PF Mm hmm.

JS Yeah.

RZ Let’s call it for what it is. If you told a writer, “I’m gonna change two things so that another 200 percent read it.” They’d be like, “Yeah, of course I want 200 percent more to read it.”

PF What about non-textual forms of communication: video? Podcasts? What do you do about that?

JS So we measure video in pretty deep detail, in the same way that we measure text content.

PF Ok, so, what do I do— if I wanted to start something new. Like in the next three months, what— what— what media platform would Chartbeat say I should build?

JS What do you mean by media platform?

PF Well, I mean, should I just do like— first of all, should I just be on Twitter and Facebook and say, “To hell with it”? And—

JS So, you definitely shouldn’t do that, right? [Paul laughs] I mean I think you need, you know, this advice in some ways feels trite but you— you need a— a coherent brand [mm hmm]; you need really good stuff that people actually wanna read; and a notion of who that audience is; and you need to nail it. You know, you need to write a lot of stuff; you need to look at what people are engaging with; and you need a unique voice.

PF Quality is hard, right? It’s a pain in the ass.

JS Yeah.

PF Who wants to do quality? [Josh laughs] It’s like—

RZ Medium! [Laughs]

PF Yeah, Medium does. I mean it’s like you gotta go buy— it’s like that point in your life where you’re like, “I should probably get a sofa that doesn’t fall apart in two years.”

[8:42]

RZ Or buy one every two years [Rich and Josh laugh].

PF Yeah, there’s that.

RZ Which is media today [laughs].

PF Most of the media industry is like, “Let’s just go get that sofa down in Brighton Beach and not even go to Ikea.”

RZ “And we’ll come back in a year.”

PF Yeah [Rich laughs]. Um ok so tell me your— what’s your full title?

JS Chief of Product Engineering and Data Science.

PF Like, you walked in and you found it a certain way and now, today, you have a leadership role, very clearly, and it’s— it’s different. What’s changed in those six years?

JS I came in into a data science job. My— you know, that’s where my background is. And, you know, I thought that my role was kind of gonna be doing science on data. And—

PF So they would literally be like, “Josh, here’s some data—”

JS “Here’s some data, could you science it up?”

PF “Do some science.” Ok.

JS Yeah, yeah, totally. And— and there were a few, you know, bits of science to be done but by and large, um, you know, first of all: there actually wasn’t the data that I thought there would be, right? There weren’t the systems in place [mm hmm] to actually do analysis on. And then, also, you know, I kinda realized that actually there was other stuff to do and that I really loved the— the product stuff. That kind of marriage of like how do you take what our data is capable of and turn it into something that can solve a problem for a customer is— is pretty cool. So, one of my favorite analyses that we ever did was, you know, doing a kinda meta-analysis of all of the headlines we’d ever seen. So, you know, we’ve seen hundreds of thousands of different headline tests at this point and they provide this really interesting data set because you’ve got a given story, and you have multiple different headlines for it. So, a thing that you can do is you can take a rule that people use about how to write headlines. So, people hypothesize let’s say, you know, getting numbers in headlines really worked. And what you can do is look at all of the times when somebody ran a test and one version of the headline had a number in it, and one version didn’t have a number in it, and you can say, “Ok, how often does the number version win?” And— and once you’ve done that analysis, then you can think about building that back into the product.

[10:37]

PF So now you have a hypothesis about how headlines work, and you wanna get something into the hands of media people who are of a range of skills, as users, [yeah] what do you do?

JS When we inject statistics or models of any sort, we do it in a way where you don’t even necessarily know that there’s a model behind it. So, if a story is doing well on the homepage, we’ll color it red or green;if something is trending up, we’ll put an up chevron next to it, or a down chevron next to it if it’s trending down; we’ll change the opacity of things if— if we think that a number is really important, it’ll— it’ll just have higher contrast. But these sorts of things where I don’t think a user, if you asked them to say, you know, how do I know that this number’s important they would ever really call out that we actually were changing the opacity of numbers. But it just kinda brings to the fore in people’s minds when somebody’s using the product, you know, the data that we think is important.

RZ This is design.

JS Yeah!

RZ You’ve got this world of information, this world of data, and you’ve gotta distill it down so that someone that’s gonna glance at this thing over five or six minutes can actually make sense of it and you can guide them along. I mean, it sounds like your role is beyond just getting good data and gathering it but also presenting it.

JS Absolutely! I mean I, you know, I— I think design is incredibly important, you know, in all— in all products but especially if you’re making a product that people are gonna use all the time but it’s not their main piece of work, right? People are willing to tolerate a clunky UI if using that UI is their core job.

PF Mm hmm.

JS But if somebody’s trying to, you know, get a quick glance at their data and go back to their day job, it needs to be really easy to use or their just not gonna take that glance.

RZ Yeah, I mean, Google Analytics, it takes me a minute. I stare at it for about five or six minutes to really absorb it.

PF [Crosstalks] But that’s because it’s not— it’s not our everyday job.

[12:29]

RZ Yeah but, I mean, what I think what Josh is saying is it’s not, very often, it’s just people logging in and saying, “How’s this going?” It’s not their job.

PF Right, right, right.

RZ Um and that’s— that’s hard, right? And so that’s a design challenge. It’s very add— it’s kind of fun and addictive, especially if it’s alive. It’s always fun to stare at that stuff.

PF No, Chartbeats moving. Whenever I drop in to Google [yeah, yeah], when I go into Google Analytics, though, I’m just like, “What have they done, this time?” [Yeah] There’s always new rectangles and things flying around.

RZ There’s a lot to process. Yeah. Yeah.

JS But if you’re a power user of it, it’s great, right?

PF Of course.

JS You just can’t be a passive user.

PF Same is true of like the ad buying tools for Google. You go in casually like, “I think I should buy an ad,” and that’s it for the week. But if people use those every day for ten hours a day, they don’t have any— any questions about what’s going on.

JS Right.

RZ Does Chartbeat suggest what to do?

JS We have always tried to stay away from that line. I think people really have a good sense of news judgement and also are really hesitant to use a system that directly tells them what to do.

PF Sure.

JS You need to take people right up to the point where you highlight things that are important but if you go too far you lose your users.

RZ Mm.

JS And I think that— that, you know, is a really, really interesting kind of, you know, line to walk.

[13:43]

PF You know, one of the hardest lessons to learn, too, is that if you have a hit, you don’t get to go back and do that again and have another hit. Like it’s just— you have to create an environment in which things can pop but you can’t just say, “Go—” And— and this happened all through the 2000s, people would go, “Can you go viral?”

JS Yeah.

PF “I need you to go viral,” and it was just like, “What the hell are you talkin’ about?”

RZ Also, I mean, even certain patterns that are clearly attractive or have their moment flame out. Like those 30 second cooking where they compress two hours of cooking into 30 seconds, those were everywhere at one point. Like you’d watch, you know, chocolate cake get made in 30 seconds and now, I think, people are just done with it and so—

PF Oh I’m sure or there’s someone out there who’s figured out a way to make a chocolate cake in five seconds. You know it’s— people get—

RZ Paul, those are sped up. Those weren’t [others chuckling] actual people cooking for 30 seconds.

PF They get numb to form, right? Like—

RZ Yeah, exactly.

PF Like somebody’s like, “Oh—” And people— the great flaw of the media industry, in my opinion, is it finds a form that works at all and then it just spreads like wildfire. Everyone’s like, “Well we’re putting in a video. Here we go.”

RZ Yeah I think that’s just the— the— the industry and the environment will dictate that. I mean, [that’s right] “8 ways to blah blah blah”, I mean was, I feel like that’s just exhausted at this point.

PF Do lists still work?

JS Yeah, I mean, you know, but they’re used a lot less than they were a couple of years—

PF Right, it’s not everything is in list form anymore.

JS Yeah.

PF Cards never caught on. Remember cards?

JS No, yeah, cards— cards really—

[15:05]

PF Vox was gonna be like a stack of cards— there was gonna be like hypercard for—

RZ Cards? Oh yes, yes, yes.

PF Yeah, yeah, yeah.

RZ The internet was heading towards a, you know, um a day when nobody knew how to read. It was just—

PF You know, bless— bless everybody for trying though.

RZ Yeah, I mean, look, this— it’s an environment that has enormous pressure around it. I mean, that’s just, they have to play — [laughing] play around.

PF Does anything like that ever catch fire? Like the new— it just seems like we’ve got articles and lists and quizzes and we’re kinda gonna be done for awhile.

JS Yeah, I mean, I will say, you know, the— the article, I think, is still going strong [chuckling], right? You know?

PF It’s good. It’s good form. You can do a lot with it.

JS Yeah! I mean and, you know, most people who are going to a media site are going there to read. Right? [Music fades in]

PF But galleries though.

JS I mean they’re still there but they’re— they’re mostly there for the ad revenue [music ramps up, plays alone for six seconds, ramps down].

PF Rich, let’s break in here for a minute, from all the [music fades out] conversation about product management and talk about Postlight.

RZ And how it’s mastered product management [laughs].

PF [Chuckles] You know? It’s a— it’s a process. Chartbeat is a good example of the kind of thing we would integrate with. Like you’d come to us and you’d say, “I’m standing up a new media property and it’s gonna be on this platform and I need some custom work and there’s gonna be ten kinds of users [yeah] and I need some good analytics and I don’t want you to build the analytics, obviously, we’re gonna go with Chartbeat and you’re gonna integrate with their API and it’s gonna at the CMS level.” That’s the kinda thing people say to us!

[16:31]

RZ Yeah, and— or we’ll say to them, cuz we’ll say [that’s right], “You don’t need to reinvent the wheel here. There are good tools out there for analytics, for CRM, for whatever.” Unless it’s so specialized and specific, which we love to do [laughs].

PF And we’ll take the time, and we’ll figure that out too.

RZ Yup.

PF So, if you need us, and we do a lot of work for media; a lot of work for finance; a lot of work for non-governmental organizations; and big non-profits; and all sorts of other companies in between, you know how to reach us. It’s hello@postlight.com. Alright let’s, let’s [music fades in] — let’s stop talking about ourselves [Rich giggles]. Ok [music plays alone for five seconds, ramps down].

RZ How much do you know about and how much [music fades out] would you like to know about that user when they come in? Obviously this wades into territory that’s been on top of a lot of people’s minds over the last year or two years is what do you know about me? And what are you doing? Because you know this about me?

JS So, in the early days of Chartbeat, the company made a decision that at the time I think didn’t feel that significant but actually really was— was pretty prescient, to try to build analytics without collecting any personal data of any sort of meaningful form. And that decision has gone through to the present day. So, you know, for example, we don’t store an IP address, we don’t set what’s called a third-party cookie. So we don’t set a cookie that you can use to track users between websites.

PF You made your 2018 so much easier than everybody else.

JS Oh no! Yeah, yeah, yeah [all laugh]. Um but you know it’s sort of a profound question in some ways like how do you make analytics that are really actionable and still have data about users, but without pulling that data that’s server side. So we actually do some stuff on the client side. So a user’s browser can— if you wanna know what— let’s say whether or not a user is new or returning to a site, one way you could do that is you could send data about, you know, a user ID back to the server and the server could decide whether it’s seen the user before. But another way you could do it is a user’s browser could just keep track of whether or not they’ve been to the site before [mm hmm] um and then tell us, “I’m new,” or, “I’m not new.” [Right] And so— so we push a lot of that stuff to the client and it means that we— we don’t have to collect it which we feel good about.

[18:46]

PF Did you, as product lead, had to spend a lot of time dealing with the GDPR?

JS It was largely my responsibility to deal with. I’m our Data Protection Officer.

PF Ok.

JS You know, for us it was a lot of legal stuff [mm hmm] and it was some system stuff but um— but it was a lot less than— than it would’ve been if we had a ton of personal data. And we don’t sell data or anything like that so, you know, we didn’t have to the sort of editing that we would have if we were— and had retargeting or something like that.

PF Do you have lawyers in house or do you go out?

JS Both.

PF Ok.

JS So we have in house counsel but for things like GDPR you have to work with— with specialists.

PF Are there specialists where you can—

JS Oh yeah.

PF What do you think about it? Do you think it’ll achieve its goals? Actually can anyone articulate GDPR’s goals? I’m not quite sure I can.

RZ I think it’s being more transparent with the user [ok] about what’s happening [right] to them.

JS I think that GDPR was a really good thing, you know, I think that there certainly are flaws in a law and implementations have been so across the board that I think it’s kinda created a mess for— for users and certainly it’s been expensive for companies but—

PF Mm hmm.

JS— the idea that a user should control their own data is a great idea, right? You know that we’re no longer in a world where a site can collect any conceivable thing about you that somebody can engineer a way to collect the data, great! Um that we’re no longer in that world but you actually have to um get somebody’s consent and show them what you’re collecting [mm hmm]. And that seems like it’s totally a good thing.

[20:15]

RZ Ok, so, let’s— let’s summarize for— for people listening: you visit a site, there’s this footer banner that kicks in, if you hadn’t seen it before, and it has a statement in it about what the site is doing, and then you can close it away, or you can do what?

JS So on— on a typical site what happens is you can go in and you can set your preferences. So you can— a lot of sites will— will say, “We use this long list of technologies, you can opt in or opt out of each one.” And there are checkboxes for every single piece of technology and they will or won’t fire on, you know, those vendors depending on what people click.

RZ I think the failing here isn’t— I think the spirit of it is— is very good. I think I under— I mean I understand the motivations around it. I would venture to say that 95 plus percent of people close that box and barely read the words in it.

JS The— the funny thing is that it doesn’t— it doesn’t necessarily matter because companies had to do the work to support the box, regardless of how many people use it, right? So, you know, if somebody goes to one of our customers and says, “I want all the data that you have collected about me,” that customer can forward us a request that says, you know, “I want this data” —

PF Right.

JS— and we have to comply! So that changes our relationship with data.

RZ Where you say “our”, you don’t mean Chartbeat, you mean anyone that is running a site that has had to comply with GDPR.

JS Absolutely. Um you no longer— if, you know, the second somebody comes to your website, have rights to all their data, you really have to think about what the data is, how it’s accessible, how you document it, and so on. And that’s a good thing even if only one percent of users is actually making use of those rights.

RZ Yeah.

PF I think what people may not know, too, is it’s like a fungal rot. Like a deal that the sales guy made three and a half years ago that was like— seemed sensible at the time [yeah, sittin’ around] but now is feeding payday loan ads throughout your site [yeah, yeah]. Like it forces you to pick at that stuff a little bit.

RZ Yeah. I mean, look, this was a compromise, right? I mean the real brute force way to attack this is, “You can’t do that. It is now illegal”.

[22:16]

PF Is opt in.

RZ Or opt in. Like just something a little more aggressive.

PF The reality is is it’s just more stuff [yeah]. Like you just— everyone on the other side of GDPR will be like, “Oh yeah we have to engineer for GDPR compliance,” and it’ll— it’s like dealing with user authentication [yeah]. It’s a horrible, boring task [Josh chuckles] that you must execute well on and then figure out how it fits into the overall product strategy.

RZ Also, we’re web— we’re browser users. I mean we’re— we’re technical— technically minded people. On a phone, it’s really weird.

PF [Chuckles] It’s really weird.

RZ It’s kind of like, “What happened [Paul chuckling] to my browser? What happened to my phone? I think there’s something wrong,” and it’s just— I don’t know. Anyway.

PF Talk— you know I’m curious to know— talk a little bit— you’ve got a— you’ve got a product that’s been around for awhile.

JS Yeah.

PF Um maybe about ten years old now or?

JS Yeah, we’re at ten years in April.

PF You came along and now you’re running a product on something that’s been around for a bit. It’s not green field; you’re not starting from scratch. Talk a little bit about how you plan.

JS So, I’m a big fan of thinking about the kind of classic, 70-20-10, you know, breakdown of 70 percent of your effort should go toward satisfying the— the needs of your core users; 20 percent should be looking at adjacencies; and 10 percent should be totally green field.

PF How long is a typical— like do you think in one month, six month, one year— like what are your— what’s your time?

JS The way that I at least like to talk about it is our one year plan, we try to paint in watercolors [mm hmm], right? It’s kind of, “This is where we think the world is— we wanna adapt to—”

PF Yeah, it’s a little blurry at the edges [yeah] but very pretty.

[23:48]

JS Yeah, yeah, yeah. And then we have a six-month plan that actually looks like a plan [ok]. You know, a six-month plan that says, “We’re doing this.” And then we do— you know, sprint plans that are like, you know, draw it pencil. Right? That [right] have, this is literally what each person is working on.

RZ I mean dashboards have been around for a very long time. What does a dashboard look like in five years?

JS Oh. Inter— interesting question.

PF [In low tone] Oh!

JS For me, it’s very likely that the dashboard of the future still lives on a desktop. Unless work moves off a desktop, um it’s gonna be there. A lot of what we think about these days is integrating data into people’s workflows. So, one of our most popular tools is— it’s basically a browser extension. And it’s something where you install it and then when you visit your own website—

RZ Mm hmm.

JS—  you know Chartbeat data is just overlaid there, right?

RZ Yeah, that’s really cool actually.

JS Yeah and so, you know, or— or Slack integrations, things like that—

PF Mm hmm.

JS— so— so you don’t even have to think about navigating to our product. The data is just there when you’re doing the task that you’re trying to do—

RZ You’re challenging the whole notion of dashboard [right]. You’re— you’re just letting the data and that intelligence seep into other—

PF [Interrupts] What’s better than not having to click? [Rich scoffs] The best thing in the world.

RZ I don’t want to touch my keyboard. I wanna go through a workday without touching it.

PF I mean when it’s— if it’s in the corner of your eye and you can safely ignore if you want to [yeah] or you can look at it if you want to [yeah]. That is the best user experience possible.

[25:09]

RZ Absolutely. That makes sense. I mean there was that trend for like a week with those heat maps, everybody had those— and they look so pretty. But their— their utility was kinda limited. Um but that makes a lot of sense. So, the dashboard! Is just everywhere beyond some destination.

JS Yeah. But I do think— you know, on the other hand, I think the dashboard itself is also sticking around, right? If you actually need a piece of information, you’re gonna go to a product and— and do it [sure], right? Um it’s— it’s different if you just wanna passively consume information as you’re going about your day, that’s a great opportunity to integrate [yup] but sometimes you just wanna know an answer to a specific question and the easiest way to get it is to login to a product and— and ask the product what the answer is.

RZ I wanna close it out with a question about product management. We talk a lot about product management here. What’s the one bit of advice you’d give to product management leaders?

JS Having the right goals is so important, you know, that feels a stock thing to say in some ways but the amount of time I’ve seen wasted from people doing really great work on goals that were bad [chuckles] is remarkable [mm] and then the value that we’ve gotten from, you know, even small bits of time where people were working on just the right problem and in a couple of weeks were able to have huge impact. You know, I— I think that, in some ways, is the— is the core function of product management is to figure out what people are supposed to do. Setting the right goals; making sure a team knows them; making sure that that work is actually achieving the goals is so, so important.

RZ Yeah, it sounds basic but it’s often a thing that slips through.

PF It’s always basic.

RZ It’s always basic.

PF I mean there’s nothing complicated, ever, at that level [yeah, true]. It’s just . . . humans.

RZ Yeah.

PF They get excited by really shiny stuff. I do.

RZ To this day.

PF I’m still excited by shiny stuff [Rich laughs]. Well, on that bright note: [Josh chuckles] this has been a great conversation, Josh. [Music fades in] Thank you.

[26:58]

JS Yeah, thank you!

RZ Thanks for doing this.

JS Yeah.

RZ That was great. Paul, I— I wish I had a dashboard that just gave me a real time understanding of our relationship.

PF Uhhhhh [laughs, Rich joins in]. It— it— I think for both of us, it just veers between like very steady state, tremendously loving, and then there are these Mount Everests of mutual annoyance.

RZ [Laughs] Um.

PF Anyway, yeah, the— the dashboard of our relationship is something I never wanna see.

RZ [Laughing] Ok. Fair enough.

PF That’s what I would say. We should get back to work but if people need us: hello@postlight.com. We are here to help! We like when people get in touch from the podcast cuz it allows us— it justifies us continuing [chuckles] talk to each other while recording.

RZ Talk in your ear. Ok.

PF Ok. Thanks, everybody!

RZ Have a great week.

PF hello@postlight.com [music ramps up, plays for five seconds, fades out to end].