Transcript

Rich Ziade It doesn’t have that—you know, the VR headset image that a lot of consulting firms use, so that—It’s like that robot woman. Do you know what I’m talking about?

Paul Ford Oh, they love it. Yeah, I know. 

RZ It’s like a robot woman looking out into space but she’s a robot but she’s also wearing a VR headset but she’s a robot, so she doesn’t even need the headset!

PF [Crosstalking] You know what that is? You know what that is? It’s embarrassing. Our industry is extremely embarrassing. I’m embarrassed by [Rich laughing] our industry a lot [music plays for 18 seconds, ramps down]. Rich . . . hey! 

RZ Paul, welcome to the Postlight Podcast, formerly known as Track Changes. 

PF Good to be here on the Postlight Podcast. Track Changes is but a distant memory! 

RZ It is. 

PF We’ve rebranded [music fades out]; we’ve got a new logo; we’re gettin’ the new deck done. It’s good to have a deck. 

RZ Yeah, I’m feeling our vibe. I like it. I like where we’re going. I like what we’ve grown up into. But I wanna talk about somethin’ today, Paul. 

PF Oh, you got something on your mind? 

RZ I do. 

PF Whachu got on your mind, Richard? 

RZ I love tech that gets rid of tech. 

PF Here’s the big idea, right? Which is that—and I think it’s more subtle than this but the big idea, “Boy! We always have done it this way, and that’s how we make our money and keep our control but this software over here, makes it a lot easier and faster but we don’t have as much control and we won’t make as much money from those parts.” 

RZ Mm hmm. 

PF “So we probably shouldn’t use it. We should be really, really careful before we bring that in because we don’t wanna mess up our world. Oh, wait a minute! The company that—down the block, that’s a little bit younger than us, just started doin’ it that way. They’re able to work a lot faster and they’re charging their customers half as much. Their customers aren’t aware that it’s premium. This is really bad! This is terrible!” 

RZ Yeah. 

[1:53]

PF [Chuckles] “We need to acquire a company immediately!” 

RZ Yeah, yeah. 

PF Right? Like . . . this is how this plays out. 

RZ Yeah, and—and—and you can resist it for as long as you want cuz you think—You get overconfident right? Of course somebody wrote the memo at Walmart 20 years ago [snort laughs] about Amazon and what was coming. 

PF Well—

RZ And Walmart’s doin’ just fine but still you get big and when you get big you ever see like a big, big ship turn? [Chuckling] Takes—

PF Slow—slow process. 

RZ Takes like an hour! [Laughs] 

PF Oh! Like a cruise ship in Brooklyn. 

RZ Yeah, yeah. 

PF You see one of those like, 40 storey cruise ships like runnin’ around in the harbor. It’s just comedy. Like whrrrrr [to imitate engine]. Like, you know, the other thing worth noting, too, like you said, Walmart’s doin’ fine. And it’s very easy in retrospect to write your case study and say, “How could Walmart have missed the growth of—” 

RZ Of course, everyone does it. Yeah. 

PF But it’s like, what you have there is not this story of like two competing brands. It’s the story of one firm that utterly optimizes its logistics pipelines at the platform level down to the silicone. 

RZ Yeah. 

PF And then you have another one that has an enormous real estate footprint with lots of bodies coming through the stores that they’re trying to optimize the hell out of anyway. 

RZ Yes. 

[3:08]

PF And God help anybody who works for either one of these organizations, especially in their physical plant. My God! 

RZ Yeah, yeah. 

PF They’re both very tough on human beings. 

RZ They are. They are. 

PF So, I mean I think you gotta—you have to refrain that cuz it’s really easy to be like, “Well, how could Walmart be so stupid?” Like, Walmart was exactly what Walmart wanted to be. 

RZ Yeah—

PF But there is this moment, and I’m seeing it—I’ll tell ya what: I’m seeing this in our own firm, which is that a lot of the platforms that we stand up for people really are starting to become more and more commoditized than even than they were four or five years ago. So I’ll give you an example, I just had a conversation with someone today . . . and it was like, “Look: you seem to want three things: you want a funnel with a pi—where people come in and they pay you money and that’s probably stripe, and you need content, and you need to put it in a place, and you need a place where editors can manage it and put art and so on and that’s probably WordPress. And [mm hmm] and you want to track relationships with the people who give you money and do things with them and invite them to things,” and that’s a CRM. And that pattern is—I’ve probably seen that pattern 20 times and the CRM is probably Salesforce but it could be something else. 

RZ Yeah. 

PF And what’s real is that because the work of setting those up and packaging them up and sending them over—A lot of times you would think, “Oh my God, that’s really bad for Postlight.” The reality is it turns out to be really good for Postlight. We can do more work that matters and less work that is, “Let’s set up and learn the new content management system.” 

RZ Yeah. 

PF It doesn’t actually—it doesn’t eat into our business the way you would expect to have things that are kind of out of box. 

[4:38]

RZ Well, I mean, you’re touching on something . . . and I’m not an economist. So, you know, you look back on—Actually, I’ll give you a concrete example. I was working at an insurance company that took in faxes from insurance agents, and there were data entry people . . . in the firm. 

PF Woof! Wow. Ok. You have to start over cuz I just died inside [laughing]. 

RZ Ok, it’s now—it’s 1863 [laughs] . . . horses are still the primary means of transportation. No, I’m kidding. 

PF I just—my hand shaking! [Laughing] 

RZ No, but—I mean—they had a network of agents, the agents got their 15%, “Bring me the fax, and I’ll insure their house.” That’s all they did! But something interesting happened, right? I was brought in and some of the work was essentially saying, “Uh, the internet’s here, guys, it’s 1999 . . . ish.” 2000. 

PF Early for insurance, early for insurance. 

RZ [Jinx with Paul] Early for insurance. I was like, “They should probably go to a website—Not the customers. We’re not at Geico scale yet. 

PF No, no. 

RZ The agents . . . should just go to a website. And fill out a form. And when they fill out—

PF About one third of them had AOL at that point. 

RZ But, no, actually, the agents—

PF Actually, no, cuz agents—

RZ They were starting to get it. 

PF They had to get on e—Email is really good for an insurance agent. 

RZ Email’s important. Yeah. You could attach stuff. 

PF Yeah. 

RZ So, anyway [yup], we get on there and then this kind of dark cloud comes sweeping over the entire organization which is, “What the hell are we gonna do with these 30 data entry people?” And it’s a classic [oh] right? Like job elimin—“Oh my God, we’re destroying jobs,” right? But ultimately—I’m not gonna get into the narrative of this particular company but lemme just put it this way: thousands of jobs were created on the other side of this kind of efficiency and innovation, right? Strangely. Right? You would think people would be left out in the street but what it actually does is it’s incredibly empowering because the organization’s allowed to grow and it just shifts. Like, customer service people who are handling things with agents are now not entering data anymore. Wonderful, wonderful thing. So innovation, actually, can be pretty empowering. It was a weird time because you could tell you were kind of spreading anxiety, unintentionally across an organization that was used to just coming in and gettin’ their—you know, their Dunkin’ Donuts coffee and just puttin’ in data all day. 

[6:58]

PF Well, lemme—lemme respond to this two ways, right? So one is if you care about efficiency, the other side of that is labor and it’s really tricky and you’re telling this very straightforward story. I don’t actually—We talk about this a lot more. It’s way more complicated. There are human beings on the other side, there are families. You think about that. And that does factor into your decision making but ultimately if somebody can turn on a computer and do a job in ten minutes that used to take somebody three days . . .

RZ Yeah. 

PF You have to account for that in how you are building and running your business. The speed, not just the labor. And you gotta assume—Like, we work with big organizations and we talk with them about this and they’re like, “No, we need to know what’s coming so that we can retrain people and help them.” So there are good ways to approach, right? So I think, you know, it’s not as simple as, “Oh, we disrupted it.” The second part is just like—you can’t escape this. Like, this is—this stuff—

RZ It’s gonna happen, right? You’re gonna die if you don’t address because somebody else is gonna show up. You heard of Lemonade . . . the startup? 

PF Yup.

RZ I think they’re like, “You know what? The hell with forms. Why isn’t it just a bot named Stan that you just talk to and say, ‘I wanna insure my jewellery’”? 

PF Stan can talk to you. 

RZ And then—yeah, and I think if I’m not mistaken, when you file a claim in Lemonade, they wire you the money instantly. Like right away. They’re like, “Stan, my bicycle broke and I had it insured.” They give you the money right away and they worry about the validity of the claim after the fact. 

[8:23]

PF Listen, it all reduces to one spreadsheet, right? Like, they have their model. The other thing I wanna say that’s important, right? Is that this is, unfortunately, a side effect of a successful career and I don’t know how to work around this . . . is that when you walk into the room, people get a little bit afraid. I’ve seen this now for ten, 15 years, and I mean, you know me, I don’t like that feeling. It’s a bad feeling. 

RZ Mmm. 

PF But when you’re the consultant and you’ve come from outside and you represent a new way and you can point to results, or you can ship the software faster. 

RZ Mm hmm. 

PF It ain’t great. It’s not a good feeling in the room. 

RZ Yeah. 

PF And working through that—Like, you push through. I go home and feel really bad but I don’t know what else to do because like, they came and asked me could we do this better and I said, “Yeah.” [Chuckles] Right, like that’s—You could save a lot of money and do it a lot faster. Like I can’t lie!

RZ And the truth is when companies do well they hire more people. And sometimes you have to do these things to do well. And yeah, does that mean you need to retrain? And do something else to meet the needs of this new growing organization? Yeah! But it also means what was an 80 person company became a 500 person company because of the willingness to continue to seek out those efficiencies and whatnot. 

PF I mean I’ve seen it go both ways: people cut to the bone and then that’s it and then I’ve seen people [well—] reboot and they grow. 

RZ Yeah. Well, I mean, look, there are investors—

PF Let’s—

RZ There are investors out there who actually are almost . . . they relish the inefficiencies, right? They go out there and they do what they gotta do. 

PF Look, we just walked into the greatest thicket in American history right there, right? Like, let’s keep going. Let’s talk about what software does. 

[10:00]

RZ Well, yeah, I mean, but let’s talk about where it took ‘em though. So now I’ve got agents filling out web forms to submit policies for their customers. 

PF In ‘99. 

RZ A big leap up. And that’s wonderful cuz, oh, we don’t have to worry about faxes. And I see the data right away. And I don’t have to have data entry. That’s nice. But here’s the other thing that ended up happening: I was having a conversation with the executives and they said, “You know, when someone’s credit score is less than 600, and their home is worth more than half a million dollars, I don’t want the policy.” I’m like, “Ok, [hmm] cool. Don’t take the policy. What’s the problem?” He’s like, “Well, we have to submit that business rule to engineering, to the IT group. The IT group puts it and logs it in as a ticket and then that ticket [mm hmm] and then that ticket has to go into a cue and it’ll be bundled up with the next software update which will be after testing, live and in the world, three months from now.” I’m like, “Ok! Software takes time.” It’s engineering, right? 

PF Yeah, but that’s a row in a database. 

RZ There’s wires. Yeah, well, yeah. Yeah! So, what we found was it wasn’t just that we made things go faster and you didn’t need data entry, what we found was: ok, now that it’s information coming in, can we make it easy for business to inject its decision making into the process without going to the software people. How do the software people create software that eliminates the need for the software people? That is a game changer, right? So, here’s what we ended up doing: we ended up creating—let’s call it a rules engine where business people—

PF That’s a good thing to call it for people who don’t know this world very well. That’s literally—it’s a rules engine. 

RZ Literally a rules engine. 

PF Yeah, that’s a thing. There’s—

RZ So you give this thing to the business side and these are people in insurance and there’s actuaries and people are thinking about risk all the time. And they just went buck wild. They started injecting all these rules into the workflow, into the process, such that a form wouldn’t even make it over if it didn’t meet certain criteria. Right? 

PF Hmm. 

[12:14]

RZ And what ended up happening was an incredibly optimized risk portfolio because they were able [mm hmm] to constantly refine it and that lag—it wasn’t just that they’d discovered that a bad credit score and an expensive house is bad for business, it’s that between the time they discovered it and it going live in software, hundreds of policies made it in. Right? 

PF Yeah. 

RZ That lag, right? It’s like information. It’s no different than information travelling, right? If it was—If I had to put it in a note and get someone, you know, the pony express to get on a horse and get it over to me, the plague is already spreading, right? Like it’s already too late. 

PF So theoretically it’s worth hundreds of thousands of dollars in eventual claims or in how you manage your risk, if you’re able to lock this down sooner. 

RZ Et cetera, exactly. When you look at Amazon, one of the things Amazon gets tons of criticism for is, you know, they’ve created this incredible marketplace, right? Everybody—you don’t know where you’re buying your stuff from anymore. It’s sometimes from Amazon, sometimes it’s from a guy in Illinois who sells backyard inflatable swimming pools. You just don’t know where you’re getting it from. 

PF Sometimes it’s directly from the manufacturer too. Yeah, you have to look at that small type. 

RZ It’s that small type, exactly. 

PF “Sold by Happy Co, ships from Amazon.” 

RZ Exactly. But there’s something that Amazon that’s actually widely publicized. I read it in the book about Amazon, I forget the name of the book. And they do something which is they have a whole team that just analyzes the products that other people are selling and how they’re pricing ‘em. All day long. And what Amazon does is they go in. They’re like, “Ok, now we fully understand the marketplace of triple A batteries and we will not obliterate it.” And they go in. And they come up with an Amazon basics product, and they jam it at the top result when you type double A batteries or triple A batteries and they decimate the other sellers in their own platform. That’s real—

PF Some might say that’s not the best and most appropriate use of their market position. 

RZ I really—you know, that’s another podcast, Paul! [Laughs, and Paul joins] But what this speaks to is they could hire a team and say, “Hey, a lot of people are buying this particular toy right now. Should we get in on this? Should we get our own inventory? And put it in our warehouses? And should we charge a dollar less? Well, how do we know how much they’re gonna charge? Well, we could always track how much they’re gonna charge. That’s software.” Right? 

PF Mm hmm. 

[14:42]

RZ “I could put a team out there and we could have a meeting once a month and talk about trends in products or I could write the tool. And we could start to get a real time picture of what the market looks like.” 

PF I could also do that with stock trading. I could [chuckles] do that—There’s so many things that are reduced to numbers in databases. And the minute that happens, I can start automating them. Like, the minute that rule can get put [that’s right] next to all the other rules. 

RZ Yes. 

PF We’re in a funny moment because if you think about a database as a set of rules and facts, right? 

RZ Mm hmm. 

PF The thing that you did with that insurance firm years ago, which is, “Let’s make it a lot easier to add rules, so that, you know, this is software building.” Like, we talk a lot about Airtable and No-code and all that stuff, it’s the same stuff. It’s the same like, “We’re gonna put it in.” What’s happening now is that large organizations see that they had this data all over the place and then they go—they don’t think, “We should make software to make it simpler in order to deal with and manage this data.” They think, “I know! I have a machine learning problem. I’m going to analyze all this.” 

RZ Exactly. Exactly. 

PF And turn it into some magic set of insights so that I have more rules that the computer made up that I don’t really know quite where they came from but that’s cool. And then I can go from there. And that to me is—it’s a lack of bravery. 

RZ It is! It is a lack of bravery. And not only that, it’s an overreach. The truth is the question you need to ask is how is software improving my reaction time? Right? 

PF Mm hmm. 

RZ And if you pause and look at the way software works today, why are you heading towards machine learning just yet? Relax for a second. Your reaction time is just shit. If you improve your reaction time, forget the machine learning part. I don’t need to automate it just yet. I just need to eliminate the lag. And the truth is the lag exists everywhere. Part of Postlight’s success is eliminating lag. That’s what we do. 

[16:34]

PF You know what’s funny here is we spent the first part of this conversation squirming and talking through about, you know, eliminating jobs with technology but machine learning is supposed to be the ultimate automated way to eliminate jobs with technology [mm hmm] or, you know, you farm it out to Captchas and people tell you—everybody has to identify a bus but there is that point in the middle—and this is very abstract, right? But what we’re saying is that there’s probably a lot of value in your people where they could tell you how they evaluate and do their work. 

RZ Yeah. 

PF And if you gave them tools to describe that and put it into a system. You could do a lot of the analysis and a lot of the evaluation that machine learning are supposed to do for you—are probably baked into the humans in a more—in a way that actually is gonna be editable and understandable and serve as a set of rules and a set of knowledge that you can use going forward, if you have good software to capture. And good software includes boxes on a screen and [yeah] some drop-downs and a database. 

RZ Yeah, and you’re touching on—I mean, you mentioned, you know, No-code before and tools like Airtable. And the truth is those are the kinds of too—There is no machine learning in Airtable. All it is is an empowerment tool. It gives enormous power to people, effectively, and to find efficiencies whereas if you walked over to IT and said, “I really need a tool to track all of the different candidates coming into HR,” they’d be like, “Yeah, no problem, gimme six months.” Or they spin up Airtable, and they get going fast. Right? And the truth is that kind of—

PF You know what’s funny though? Is they actually don’t. They spin up a HR system. Like, that middleground, IT doesn’t deal with that middleground. They like [hmm]—They like fully custom, big hosted, you know, in our purple zone, like, secure setups. 

RZ Yup. 

PF And then they like software-as-a-service that they can buy. And the stuff in the middle is really tricky because it requires maintenance. And this is the thing about your rules engine. You know, the rules engine you’re talking about has been runnin’ for 15, 20 years now and it still requires, it requires a skilled person to operate it. It doesn’t—

RZ That’s right. 

[18:37]

PF It’s not as simple as point and click. And so people really get confused in the middle because there’s—

RZ I think that’s right. 

PF There’s this tremendous fallacy in our industry where people decide that a human being is either an absolute genius who must know everything or they’re absolutely incapable of learning. No one ever wants to say, “This’ll take a couple of weeks for someone to get it.” 

RZ No, I think that’s right and I think it’s about empowerment and the truth is you could have really brilliant people who don’t have the tools to act on what they know—to act on the knowledge that’s in hand, and that’s still the case to this day. Right? Like, that is still the case to this day. I just insured a car recently and the process was pretty amazing. The guard rails that were put in place for me to go through this and I had to go eat lunch and I paused in like step five of the form and [mm hmm] they reached out to me, and they knew exactly where I was, and they were taking me all the way through it—

PF In the form, they’re like, “Hey, Mr. Ziade, looks like you could fill out the rest of the page and we could get this done.” 

RZ I got an email. I got an email that was like, “Hey! Where are ya?” And it was—you could tell tons of thinking was happening around the funnel to get me all the way to the close, right? 

PF Mm hmm. 

RZ And what was also interesting was after I closed there was still information they needed but they had already assessed that the risk was behind them now. Even though they needed a [mm hmm] couple more things, I got this almost threat—not threatening but it was like, “Listen, we need two more pieces of information from you, otherwise we’re gonna have to cancel the policy.” I’m like, “What the hell? You just took my money two weeks ago, why didn’t you ask me for this in the first place?” 

PF Hmm!

RZ I think what they’re doin’ is they’re closing the deal . . . and they’re asking for the little bits later. 

PF Well, no, they probably went out and hit a couple databases to get a little more information about you, right? 

RZ Yeah, my credit score’s good. Like they saw this extremely low risk. 

[20:23]

PF Yeah, no, no, but I mean, they got you on the other side and then they’re like, “Ok, now we can do the batch job. Like we don’t wanna do that in real time it’s probably really expensive.” They probably spent five dollars [yeah] to make sure—to lock you down. And then they’re like, “Oh wait, go get those two pieces of information.” 

RZ That’s right. And so I guess what we’re gettin’ at here is there’s tons of lost cycles that don’t need to get fixed with machine learning or AI [!!!] like that is sales. We are a consulting firm, AI and machine learning is plastered all over every consulting—technology consulting firm. 

PF A lot of people ask us about it now. They’re like, “What are you doing about ML?” And I’m like, “You don’t have any data.” 

RZ Yeah, yeah, yeah. You do things terribly today and there are better ways to do them which will result in huge benefit that has nothing to do with a computer thinking on its own. They just don’t. Right? 

PF No, I mean, it’s literally, like, half the time people are asking what are we gonna do about ML? It’s like, “Well, can I look up your customers by zip code?” You know? And [Rich laughs] they’re like, “No. That’s not—we didn’t get there yet.” And you’re like—

RZ Right, yeah. 

PF —Ok, well, you know, start there. Look: let’s actually—let’s close on some advice, right? So we’ve gone the entire gamut of digital capitalism in a number of different ways here and you’re in a world in which . . . here’s what we’re saying: we’re saying there’s a gap between you and your competitors and you can close it up with simple software that uses the intelligence of the people in your organization and then you can do predictive things if you gather that. That takes the form of a rules engine, a database, and so on. Very good for industries like insurance but also anything with a CRM where you’re trying to build relationships. I feel that this is relevant. But what everybody is selling now is ML and prepackaged solutions. So what do you do if you believe that your organization should really take—try to own that middle? And build something—Like, how do you sell that internally so that somebody gets excited and committed to it cuz it’s a very hard thing to communicate. 

RZ I think that it comes down to a question you can ask that I think resonates with business a lot which is how can I use tech to empower business people to rely less on tech? That’s the fundamental—

PF Mmmmm. 

[22:41]

RZ—question, right? It’s like, “How do I—how do you—how do I make it so you don’t have to come to me and I have to tell you it’s gonna take eight weeks? How do I empower you, right?” And that has very little to do with ML because what—I’m doubling down on the human here. I’m doubling down on the decision maker, the person who wants to put their thoughts into action without going to IT. 

PF Mm hmm. 

RZ That is the question. 

PF You know—

RZ And that’s not interesting and not sexy and it doesn’t have that—you know, the VR headset image that a lot of consulting firms use, so that it’s like that robot woman, do you know what I’m talking about? 

PF Oh they love it. Yeah, I know. 

RZ It’s like a robot woman lookin’ out into space but she’s a robot but she’s also wearing like, a VR headset, but she’s a robot, so she doesn’t even need the headset. 

PF [Crosstalking] You know what that is? You know what that is? It’s embarrassing. Our industry is extremely embarrassing. I’m embarrassed [Rich laughing] by our industry, a lot. 

RZ Yeah. 

PF No, but you know what? Ok, the last thing I would say on that, right? Like is so what you just described—What I would—the exercise I would do if I was tryna figure out how to improve things, I’d draw my funnel . . . really carefully, like how do I get customers? And I look for that—Like, how do I make a decision about this customer and qualify them? 

RZ How do I make the best decisions? Right? 

PF That’s right. 

RZ How do I make the right decisions? How do I bring in the right customers? Like these are things—

PF And literally where everybody is saying ML and they drop that into your funnel so that you can pre-qual—blah blah blah blah. Think: who knows how to do this? And could we kind of—like is there an algorithm that we could actually drop in here that—and not only that—like, not just drop in once but could we give people the ability to modify and manage this, so that—

[24:15]

RZ Empower. 

PF Nothing has changed with artificial intelligence where the computer is suddenly smarter than the people [!!!], it can just deal with more information [music fades in]. Right? Like it doesn’t know more than the people on the floor. 

RZ [Crosstalking] That’s right, that’s a great point. Paul, I loved—you know one of the things I’m most proud of about the new postlight.com is that there [yeah] are no metallic looking women wearing VR headsets staring out into space. 

PF I’m gonna vow to you that we will never do that in our marketing, ever, under any circumstances. 

RZ Ever. Ever. We are a digital strategy design and engineering shop based in New York City but based everywhere nowadays. We’ve got some great case studies up now. I’m very excited to talk about Goldman Sachs and the MTA. You should go check our stuff out at postlight.com. 

PF We should also say Goldman Sachs just released a new font and it’s amazing. 

RZ Is it really? 

PF It’s called Goldman Sans. It’s really good. It’s really good about [Rich chuckling] about numbers, you can download it. I gotta give it to ‘em: straight outta the park on Goldman Sans. 

RZ Hilarious. Hilarious. 

PF And, alright friends, hello@postlight.com is how you reach us and how you always reach us. We welcome all of your feedback. 

RZ Yup. 

PF We bring strategy, design, and engineering to deliver platforms and experiences that drive digital transformation, Rich! 

RZ Woo! 

PF That is what we do. Alright, let’s get to work. 

RZ Have a lovely week. Ok. Bye. 
PF Bye! [Music ramps up, plays alone for three seconds, fades out to end.]