Get in touch

Why discuss only one topic when you can discuss three! This week Paul and Rich dig into three unrelated but important topics that they’ve had top of mind. They share musings on going back to the office and break down Natural Language Processing (NLP). Then Paul reveals what he learned from watching a two-hour presentation by Salesforce.

Transcript

Paul Ford You got me in Benioff and it’s just like, ‘you know, what’s good crackers after bedtime?’ [Rich laughs] You know, it’s just like it’s, it’s a mess. [music ramps up, plays alone for 15 seconds, fades out]

Rich Ziade Hey Paul, which topic Are we going to discuss today?

PF I was thinking, you know, we should shake it up a little bit, I got three topics. [Rich whistles] Three. Let’s see if we can stop running our mouths long enough on one topic to actually get to three. So the first one back to the office, we’re going back to the office. You know what the New York State is got now? You know, what has actually you know the big winner in the pandemic? Not humans, humans were not the big winner. You know the big winner? QR codes.

RZ Meaning…

PF I mean, just all of a sudden their QR codes everywhere, because touchless logins and contacts are a big deal. People don’t want you handing over things.

RZ I went to a restaurant, a lot of restaurants have these menu displays where you point your phone at the QR code and a menu comes up on your phone, so they don’t have to hand you a paper menu. Which is nice. I guess.

PF Touchless transactions are interesting, but I mean, you know, it happened. Finally, they’ve been trying for 20 years. And they finally got that QR code across the line.

RZ So Paul, I hear you, I hear two narratives. Which one are you? There’s the ‘God, this is so great, that convenience and freedom of working whenever I want, wherever I want, is just so great.’ And then there’s the ‘I got to get the hell out of here. I need to go see people, I need to be in an office and not because I want to stare at my computer, I just want to interact with others’ like where where do you stand? Where do you live in this spectrum?

PF Well here’s what—I first of all, work can be done in a million different contexts. People always are trying to make this into this very—like somebody is in a wheelchair, they’re very productive, they’re really good at their job, it’s really hard for them to commute in every day, okay, so can remote work, give them a career and help them be really productive? Absolutely, that’s gonna take immense stress away from them, absolutely net gain totally worth working around some of the constraints that they have, because you get someone who’s productive and successful, and you get that value for the company. And it’s, it’s a great way to win. And so there are some real obvious wins for remote work that I think, you know, everybody has been negotiating with for years and years and years. And I think we’re over the threshold on that. I think that like hiring people in different cities and seeing how they fit. But also, let’s be clear, there’s culture, culture happens at a faster rate when human beings are bumping into each other, and there’s absolutely—like there’s Slack, and there’s video and so on, you can build a remote first culture, that’s, that is a big part of it. But it’s not, it’s just not as fluid. We’re not in Second Life, we’re not playing a game, it is not as fluid. You can’t drop people into digital environments and make them feel safe and secure in the same way that you can, when a client comes to your office and you give them a cup of coffee. It’s just different. I rode my bike in the office this weekend, because I know I’m going back in relatively soon. And I wanted to just, I hadn’t been there in a year, it was like Chernobyl. And my suit jacket is still hanging over the back of my chair, right? Like, it’s just we all got up and left.

RZ Let me, but it’s worth noting, though. I mean, you’re talking about a particular industry, which in Postlight’s case, we’re both the cofounders of Postlight. talking to people and interacting with people is important for relationships, but also one of the ways we’ve been able to get such great talent that wants to be here and not necessarily, you know, Google or some other big company, is culture, is the environment we created that they can be within, we can sort of have that there is some culture that lives inside of Slack and inside of Zoom and inside of other places. But it’s hard to be the experience and the connection to the to the New York City community that we’ve been able to build, right?

PF We exist in a place right? And that place is also like it’s Thursday evening, do you want to go see a movie with the person next to you? Do you want to stay at the office and play some games? Did you see that thing at the Met?

RZ I think it sounds like we’re making a case for ‘Hey, everybody back to the office.’ It’s worth noting we are hiring more remote people now. I think it’s a blend, I think Postlight has changed. It’s like a scar like, you know, it’s like you know, ever since I had that knee injury and that recovery I’ve been a little different.

PF Before you pursue that metaphor any further, right, like we were we were half remote on engineering before this all started.

RZ Yeah, we’ve been partly remote. 

PF Just you know, I want that mix. Right? And I want I want both cultures celebrated and frankly like, ‘Hey, how are things in Minneapolis? How are things in Austin?’ This is what bugs me is it’s like remote is this blobby abstract space. And then New York City is the other place and then Lebanon is the third and I don’t like remote being that abstract. I think like people live in their places, they’re in their communities. Like, to me, it’s interesting if somebody goes to a church, right, like, we don’t talk, I’m never gonna ask an employee that like, that’s you keep this stuff separate. But it’s like, to me, you live in it, you are part of a community, you go to a church, you have your family nearby, things like that. Those aspects of how people live are fascinating to me. And they, when you know about them, and again, I’m the CEO, I don’t ask about them. But as they come towards me, they give me a richer understanding of that person and sort of where they’re at in their life and what they believe and what they’re doing. And I like when that comes–

RZ You’re talking about connecting with people. I mean, that’s what you’re talking about. 

PF I mean, I know what that’s like in New York City. I love when that comes from outside of New York City. I’ve learned so much about Lebanon, in the last five years. Think about culture, like remote, to me is not a culture, it’s more like the aggregate set of signals coming from all around the world that made the internet magical in the first place, is something that can happen now at work. And we should celebrate that and connect to it.

RZ As leaders of a growing company where there are now layers upon layers between us and a lot of the people at Postlight. I can’t just sort of stroll over into someone’s kitchen and say, ‘Hey, can I sit with you for lunch today?’ [Paul laughs] The ability to do that, you know, just to cross paths with someone and they tell you about how they’re building their own kind of homegrown drone. It’s very cool. Like it’s a five minute conversation that happens in a spontaneous moment. That isn’t in a meeting that allows us to connect with people in a human way. And that’s hard to do. It’s impossible to do!

PF Remote lunches don’t work.

RZ They don’t work, man.

PF I am a good small talker. I will work hard to make that conversation moving forward. But the remote lunch has bested me.

RZ Yeah, it’s hard. And it’s just not a natural thing. I mean, that’s–

PF No, it’s that point where you’re going like ‘wow, this is awkward’. That’s that’s what you want to avoid. You want to avoid ‘this is awkward’. The other thing I’m going to tell you the thing I missed the most you and I over the last year, frankly, we’ve grown a lot closer even than we were before because this has been a hell of a thing to manage this company for the last year. We’re so nice to each other now because he can’t really fight on video and I miss just telling you you’re full of shit. I feel that that was an important part of our relationship.

RZ I’m really looking forward. I’m really looking forward to it.

PF I really can’t like in a video chat with you just absolutely tear you down and tell you what a complete pain in the ass you are. And you in the same way. You’re very tender with me now, you’re very gentle. Because, you know, this context just forces everyone to be a little snuggle bugs. And I’m just like, I cannot wait to just look at you and say ‘what the hell are you talking about? That’s pure liquid sewage. Throw the entire whiteboard away. We should never have started this company.’ Like those are words that used to just sing out of my mouth when we were working together before. [Rich laughs] And now I’m here to ‘did you get your coffee? Oh, good buddy. Let’s work together to build a coherent, cogent, wonderful community!’ That’s not us. [Paul laughs] I mean, it is for everybody else. Obviously, we’re gonna, we’re not going to be fighting with anybody else in the org, but I need to tell you, you’re full of shit with a whiteboard and a coffee in my hand. [Rich laughs] So we’re gonna get that back. I’m looking forward to it. And I’m sure everybody else is looking forward to it too. Alright, now look, I want to get to a different subject. Here’s another one. This is sort of like, I want to start I want us to start talking about buzzwords in our industry. So we can demolish them! I think there’s too many buzzwords and sort of what they really are. So—

RZ Is this a segment in the podcast?

PF It’s a segment, we’re doing segments. [Paul makes trumpet noise]

RZ We’ll give it a name. Name the segment.

PF Bash that Buzzword?

RZ Oh yeah, Swat that Buzzword.

PF Here’s one, so we ask people for ideas. Like what should we be talking about? And they were like NLP, man. We’re getting a lot of people asking for it.

RZ Okay, what is NLP? 

PF Natural Language Processing, alright? The idea is—

RZ Work back from a real world example.

PF Okay, great. I want to search through a lot of text. Now I can just match the words in the text. Right? Okay. 

RZ So if I search for ‘oxygen’, I’m gonna find the word ‘oxygen’.

PF That’s right. That’s a great example. Concrete thing. It’s a chemical element. But now hold on a minute. I want to search for ‘politics’. I’m interested in politics news. I have a million news stories. You know what we’re doesn’t appear in a lot of news articles about politics, is the word politics. What you see instead is ‘Biden White House announces blah, blah, blah’. And now I have a challenge. The challenge is, how can I figure out that when I’m talking about Biden, and let’s say there’s also a tennis player named Sue Biden. So now I don’t want to match that one. I want to match, I need to match Biden, White House. And I need White House. I need to figure out—then I want to I want to know that those are political terms. Okay, so now I’ve been I’ve implied two things. First of all, computers are not magical animals that can intuit anything, they have to do everything through pattern matching. Now I need two things, I need an the ability to break up the text, and then match the text against little pieces of text that I already know about, like Joe Biden, White House. 

RZ So it’s creating associations between words. Is that what you’re saying?

PF Well, first, I have to extract those terms. Entity extraction.

RZ Ahhhhh, okay. Alright. Now, now, you’re confusing me, Paul.

PF How am I confusing you?

RZ Well, okay, so you’re saying natural language processing is the ability of a computer to take some words, and understand what you mean by them to help you find other information? Is that the definition of natural language processing?

PF Yes, I mean, okay, if you go to the Wikipedia page for natural language processing, text and speech processing, analyzing the syntax, like there’s a lot of like linguistics applications, and—

RZ Let’s keep going through the journey of politics. I searched politics, and you made a great point, which is a lot of articles about politics don’t have the word politics in it. So the computers got to somehow make the leap, make the logical leap that politics is conveyed through a lot of different words and phrases, right? 

PF Okay. So what I’m doing is I’m chunking everything into individual tokens, and then I’m comparing it to a database of terms, like it’s not, there’s not a lot of science, it’s just a computer being fast.

RZ So foreign policy, Joe Biden, Vladimir Putin, White House…

PF Uh huh. Do I have any matches? It’s a database of terms, okay. But it’s not just a set of terms, the database knows a little bit about what they are. So if I say Joe Biden, it knows that’s a politician, and possibly it knows the, it’s the President and wait, whose database, what database, database is controlled by Amazon, Google, IBM, etc. So like, these are API’s and services that you can use. This is, when people are talking about NLP in 2021, sometimes they’re talking about wonderful open source libraries. There’s a great one called spacy for Python, and so on, where you can do all this stuff. Or they’re talking about services that are super mega powerful, where you get to use Google’s understanding of the world in order to figure out what the meaning and the sentiment inside of your text is. 

RZ So Google is kind of next level here, like this is their bread and butter, this is what they do. So there’s—

PF It’s actually kind of, it costs money to use that API, they price it pretty well, because it is good, you give it this blob of text that’s about the White House, Google will come back to you and say like ‘this appears to be about politics, we found five people in it, we found six locations, so on and so forth’. And so it gives you data back that is tagged with the semantics of your content that you can then use to build your own data set. So I have a million articles in my archive of news, I have a million forum posts about politics, I want to get all of the ones that are about the Biden White House, which might only mention, the Biden White House includes Kamala Harris, but may not mention Joe Biden, it might include you know, it’s, it’s time bound, and so on. The entity extraction will get you there. The next step over is sentiment analysis. 

RZ Okay, so hold on, let me restate entity extraction. Okay? All articles have many words in them?

PF Or voice or you know, whatever, like, whatever it is, but if you can reduce it to text, right?

RZ Yeah, reduce it to text, yes, yes. But of those words, a lot of the words are what I call connective tissue, they’re kind of useless. ‘The’ and ‘a’ and ‘them’ aren’t that useful. But then there are the words that carry real semantic meaning that get plucked out, that get extracted and treated like entities so that they can create associations, create relationships with other words.

PF Yeah, people, places, things, gets more subtle than that. But yes.

RZ Yeah, concepts, etc, etc. But that’s step one, right, so get all those words out of the article, the words that I can key off of, right?

PF Nothing magical. This is just a computer, its gears turning really, really fast, like billions of times per second, it doesn’t know anything, it just knows that your string of characters looks like the string of characters over here. And it goes like, well, this seems to be—as a result is able to sum that up and go, ‘this seems to be about politics’.

RZ Alright, so I said politics, a bunch of words have been associated with this article. Okay, so now what happens?

PF Well, I mean, you know, you can now you have data of your own through all of your articles. And now what are you going to do with it? Well, you can make the politics search function smarter inside of your own search engine. Let’s say you do this across a million pieces of text, forum posts, okay. Those are good because you really tend to have no idea what’s inside of them. Right? As opposed to a lot of times articles how you can you know, they’re in the politics section of the paper so you can intuit. Okay, so I have no idea what’s inside my forum posts. What do I do now? Well, now I can do recommended posts. I can do like, wow, this is a lot like this. So I can tell you, you might enjoy reading this. Now the problem is it, it’s never magical quite like that. Yeah, they’re pretty related by concept, but it might be a boring, boring experience. And so what you need is actually to build some knobs and dials and be like, okay, when people read about Kamala Harris, it actually turns out, they want to read five more things about Kamala Harris, we learn that through, you know, you need to be able to you need to be able to mess with this stuff as it goes forward. But that is the root of your, a lot of your recommendation engines. And there are other ways besides pure entity extraction, there’s all sorts of statistics, machine learning things going on in the world. But frankly, entity extraction is your starting point, right? Like use one of these API’s, figure out what something is about, before you decide you’re going to build some giant ml model of the universe and, and sort of blow everything up, like, start here.

Computers are really good at looking up data. What computers need to work harder on is understanding a piece of prose or an article or a book and derive anything from it. What you’re doing here, is you’re doing the hard work of sort of making sense of just a bunch of random words connected together. You send the term sentiment analysis, what does that mean? 

PF One of the other big features in those API’s, right, which is I know all kinds of things about words, I know that some words are more grumpy, and some are more happy. And so I’m going to automatically tag how someone is feeling when they write this by looking at those words, and giving and using statistics and scoring it and saying, this is an unhappy post. And this is a happy post. 

Okay, so this sounds like social media has to deal with this. They’re trying to, even though this is a really hard problem, like to see if there are, you know, users on their platforms that are threatening people with violence or harassing people.

PF That seems like it would be simple, like, I’m gonna flag all the swear words, okay. Then there’s all sorts of edge cases that spill out. So boob is a good example, you call someone in boob, maybe you shouldn’t do that. But that’s a lot, that’s a lot different than talking about breasts, maybe you should be allowed to use the word boobs. So suddenly, you’re in this this world where you’re actually trying to like, represent culture as a set of filters. And it’s a boondoggle. And in the meantime, clever and malicious users, like let’s say you want to, you really want to ban, you know, racist content on your platform, they’ll come up with secret little words that are a little in jokes amongst themselves that are just as racist as the words everybody knows, but your filter won’t catch them. And then they’ll have a good time for about a month until somebody brings it to your attention. And you realize that there’s just been a huge amount of racist garbage all through your system. So you actually, you got to be in there. 

RZ Also you could write an article about terrible online violence, and you have to distinguish and article that is writing about it, that’s trying to report it versus actual,right? Because you’re using the same words!

PF Or people are talking about things that were done to them, experiences that they had, and the filter catches that in the same way, or things that were said to them. So the context, computers cannot judge context, not really, they can kind of bundle it all together, but they can’t, you know, there, it’s just so simple. So sentiment analysis, though, is like, all of this stuff is first pass, its weighed to do things in bulk and aggregate, that it would be really, really difficult to get a human to do. It’s very expensive and difficult for a human to look at 2 million forum posts. It’s less expensive to add, let’s say a little flag to every single post where if people see something that upsets them, they can hit that flag. And then it filters up to a moderator. And actually, I think like the social contract of user generated content, allows for that, right? Like, okay, we will police this just enough. But once we hit that flag, we expect you to take a look at it. Well now, so entity extraction, good for search engine, good for admin tools at scale, right? So, ‘hey, these things are filtering up. People are upset about them. I’m seeing a lot of really grumbly words. I’m going to prioritize those’ right?

RZ Let me ask you this. Okay. So I mean, this all is making sense to me. It sounds like this is about publishers and social media. Why should a business, any business, care about this? Let me give you a business. 1-800-flowers. I don’t want to use an actual company. So I’m going to call them 1212 Flowers. [Rich laughs] Like can NLP help me? Because like, I’m a CIO, man, I go to conferences, and I eat the you know, the shitty snacks at the hotel, and they tell me, you know, are you on top of NLP as a CIO of a, you know, large corporation? Why do I care about it? 

PF I mean, first of all, doesn’t sound like it’s urgent to your business. I’m not going to sell you something you don’t need but let’s talk a little bit about why people might use it in your context. 20,000 is not an enormous range of complaints, but if they add up over the year, you’re talking about a quarter million and as you know, let’s say you’re business is growing and you’re looking at more and more, you don’t want to increase the size of your call center that costs you money. And you know, you may want to be reducing the size of your call center, awkward conversation. So what do you get with NLP? Okay, first of all, nothing is as good as human beings reading things and making decisions. So let’s keep that in the loop. Okay, don’t assume that the magic computer will come save you all the money and take away all risk, okay, that’s not going to happen. However, I can do a couple things for you. First of all, if I use NLP across all the complaints, and all the feedback that’s come in, I can start to draw an aggregate view of all the different things that are happening in your organization that cause people to contact you, all the different, you know, it turns out that we have a real rose problem here at did 1212 Flowers, it turns out that people are really happy with delivery over in this corner, and then there’s this wonderful delivery person or whatever. Okay, so I can like—

Let me ask you this, can I surface my call center people that are just being a little short and salty with my customers with this stuff?

PF Yes, frankly, you can you, what you do is your if you’re recording all your sales calls, which is now built into Salesforce, you can turn that into text, and you can look for coachable moments, that is absolutely something you do. You can search through and look for coachable moments. Over time, you could build an aggregate model of how people are interacting with customers and look for ways to coach the team. That’s all real. But that still requires you to zoom in, right? You got it, you got to go in there and look, you can’t just sort of hit a button and be like, ‘Hey, everybody, we figured out how to fix sales!’ So yes, absolutely, there is a process here where we can look at all the conversations that everybody is having with customers, and we can analyze them. And we can give people pivotal moments to help them understand how to how to improve things. And NLP is a tool inside of there. I gotta say, when you are teaching and conveying the stuff, you got to give a demo, like we’re talking in this very abstract level, it is hard to understand it until you see it because it’s both more powerful and faster than you can imagine and also knows much less about humans than a human does. So it’s just is very like—but I can go in I can find I can make you better dashboards about what people are talking about can help you prioritize your emails and your input and your in your call center by doing some sentiment analysis, because when somebody is just like smashing all caps or saying ‘hate’ a lot, that’s going to be the person who you probably want to deal with first. So things like that. It’s an infrastructural toolkit, rather than like this one size fits all, but it’s yet another buzzword that kind of gets thrown out like ‘you need NLP!’ Okay, well, what are we gonna do with it?

RZ Paul, you said you wanted to get three topics into this podcast? I don’t know. First off, is there a theme here, dude? We started with going back to work because we’re getting vaccinated. Then we went to NLP?

PF Noo, there’s no theme. The theme is just two really thoughtful guys just trying to help people get through this crazy mixed up time. That’s all.

RZ Okay. Alright. So the third is going to help us get through this. What’s the third Paul?

PF Heh. This is more of a confession than a topic, but I watched a two hour Salesforce demo. Over the weekend.

RZ Why would you do that? What happened? Is everything okay at home? [Paul laughs]

PF Things are great. And it’s because I watch Salesforce! [Rich laughs] No, I gotta admit, this is before they acquired slack. Salesforce is fascinating to me, first of all, because Marc Benioff and I have the same physical type, which you just don’t see that often, like just beefy, sort of pasty guys who are way too tall. [Rich laughs] So, so there is just an element of like, huh!

RZ Well, I mean look man, you are in good company. That is an incredible success story. It’s not even a story anymore. It’s just part of the world.

PF No! Well, I mean yeah, so this is someone who is not just in money terms, but impact term really, like 20,000 times more than I’ve ever been able to achieve. So that’s all good. I find it fascinating because I think first of all, I think Salesforce, Salesforce is aiming to be maybe not the next Microsoft but like the last Microsoft. It wants to be kind of the operating system for how people do day to day business. And it wants to take that away from the world of Windows, I don’t think it doesn’t see Apple, you know, in the same way, like it wants to make sure that like we’re you used to think I’m going to go into Windows, open my app and do my work. Then instead you’re thinking I’m going to go into Salesforce, use all the Salesforce tools on the web, and I’m going to sell services, talk to people through video calls, chat, get my work done and and sort of be this driving engine of revenue inside of my business, right. And they’re building an enormous company built out of lots of enormous pieces doing that. And look, I mean, we play, we are a tiny mammal to their giant dinosaur, right. And that is just just a fact of life. And so I’m fascinated by how they communicate. Now, when you and I communicate, we talk about what we do. We talk very much about the products we build and the things that we do, we actually get right in there. And we’d like to show people things. What we’re competing with, you know, consultants on one side and IT services on the other, where we’re just we’re coming in the middle and saying, ‘the thing you want is to build your thing, let Postlight help you’. Okay, let’s not do 18 months of discovery, let’s not presume that this thing can be done by just grabbing some off the shelf software and crossing your fingers, you need product work, and that’s going to drive your digital transformation. And here we are and good for us. And for a certain kind of buyer, especially someone who has been burned by services on one side and burned by consulting on the other. They’re like, ‘great, please get me my thing. I love you!’ And those are the relationships that have been very successful for us. But more and more as we go into the room, people are like, ‘Well, why wouldn’t I just use Salesforce?’ [Paul laughs]

RZ You’re validating sort of a theory I’ve had about Salesforce, which is, software is very secondary to Salesforce. It is not about the software, the software is part of the thing. But culturally, nobody is going bananas about the next Salesforce conference because they’re going to unleash feature after feature a-la new Apple phone. Nobody cares. 

PF That’s right. There’s no m1 chip in Salesforce land.

RZ There’s no like ‘we’ve gone to 70 megapixel camera from 50’ There’s none of that. I’ve never heard them talk about—I mean, if the closest I’ve seen is maybe the Einstein thingy?

PF Oh it’s their machine learning.

RZ Their little Einstein thing. But even that has been packaged up into a cartoon character. Right? And I think that is not a mistake. That is them saying—

PF They haven’t made many mistakes. 

RZ They haven’t made many mistakes right? And and so—

PF People find it opaque and boring. But so let me tell you what I saw. And I don’t think—

RZ Okay, so what did you watch? You watched a YouTube video?

PF I watched a two hour YouTube video about, I don’t even know what it was. Okay, so for Marc Benioff and others, on the roof of a hotel in Washington DC, during an incredibly strong rainstorm, so just like rains pouring down, just some some planning and that in the age of COVID, that just didn’t work out. It wasn’t even like they were announcing new features, because it was stuff that they’ve even seen before. It was just it was kind of their two year—

RZ So is it a product tour? Is that what you saw Paul?

PF Here’s the pattern, okay, the pattern is head of marketing kind of kicks it off, talks about values, Marc Benioff comes on. And he talks about the values of the company for a long time, for an almost uncomfortable amount of time for me, as someone who talks a lot.

RZ It’s not unusual for a company of that size, and the founder—

PF Here’s what we do in public, here’s who we are. Then he brought someone on who’s one of their trailblazers, essentially an admin, like a certified admin, and this is somebody, she’s a special needs kid, her husband was in the military. And she really focused on developing her Salesforce skills. And so he pivots from values to talking about success, success connected to the Salesforce world, right, like so. And then they talk about how they drive trillions of dollars, or $1.2 trillion into the economy. So start with values. Take your time, talking about all the things you do. Go to success. Talk about, take your time, talk about all the people you help. Clap for them, show trailblazers, show success stories of people in your network, because that’s who’s watching. Okay, I’ve communicated my values. Third, at about like the halfway point, like, you know, or like a half hour in. Now I’m going to show you one video about a product and we’ll interview and talk to some people and I’m going to show you a produced video about like, how Salesforce—

RZ It’s a demo? 

PF No, there’s no demo. It’s a promo marketing video from an executive showcasing somebody getting their vaccine in New York City, through a Salesforce managed program. 

RZ Okay, so they went right for it.

PF Salesforce vaccine cloud, like, it’s like a five minute video, beautifully produced. Everyone is very accessible. They actually very like clearly, they don’t choose the shiny as humans, they choose smart, accessible people who are like you and me. And it’s a really powerful signal by the end. Like I’m not inclined towards any of this stuff. But it’s like, you got Benioff and like a normal jacket saying, this is amazing, amazing, amazing. Everybody says amazing 5000 times. We have amazing values. And we do we’re ethical, and we take it seriously. And we talk we work in public and we you know, just want to see people succeed. Here’s what that success looks like. Now, actually, you know, and let me show you the engine that drives that success. But really, it’s not the most important thing. Let’s get back to your success. Let’s get back to the values. The product itself is an afterthought. And you know, I gotta say, I watched for the whole two hours and it wasn’t out of—

RZ Two hours?!

PF Two hours, man.

RZ I gotta say, you know, I’ll share from my perspective, you know, we are all about the product and all about quality. And for the first three years, four years, we just made fun of Salesforce, we thought it was just the dumbest, ugliest thing we’d ever seen. Right? And it kind of is, it’s not attractive, it doesn’t try to be beautiful. And if you lift the hood, it’s pretty basic, what it does. I mean, it’s just a lot of really good side relational database stuff. [Rich laughs]

PF It’s like, I mean, it’s an object model from 1999.

RZ Yeah, it’s pretty basic! Oh, and so it’s like, what the hell? But you can’t, you have to appreciate what they over-indexed on to create the value they’ve created, which is people don’t care! People don’t care! They just don’t care, you and I care, because we’re nerds. And we can’t believe there’s still like remnants of jQuery in the thing. But people do not care. What they care about is you named it Local Baker Cloud.

PF No, no, it’s Salesforce Vaccine Cloud, right? Your boss tells you to use it. And you know, what all Salesforce does that, I think is kind of interesting is it’s like, let’s try to make that as positive as possible for everybody. Right? Like, that was never Microsoft, Microsoft was we will, you’re going to be, we’re going to sell this in the enterprise, we’re going to make sure that anything that’s coming for it is quickly destroyed. And then you’re all just going to have to deal with it. 

RZ Yeah, yeah. And then the licensing revenue just goes on forever, or whether you’re gonna delete your data, right? Like it’s, I mean, I think that’s changed. I mean, Microsoft has changed, I think.

PF They are, but Salesforce wants everyone to feel good and successful who touches this thing.

RZ It’s a great observation.

PF I can tell you from objectively that they don’t, but, but nonetheless, I really do think they want it in a way that they want business to be this sort of big, healthy, positive thing.

RZ Okay. So I think that’s a lesson learned. That’s a lesson in humility. Don’t be too arrogant about the tech.

PF No, no, no values, success, product in that order. That’s how you bring people into your world. I want to go out and show you the cool widget I built. And it’s like, that was a that was a lesson. I’m just sitting there watching it going like, wow, they’re not showing the product, we’ve learned the wrong lesson from Apple, because Apple is a consumer products company. You have to get that product in your brain as soon as possible. Salesforce, Postlight, Software as a Service, you sell success to the client. 

RZ Yeah, that’s a great point, Paul, I mean, what an amazing journey this podcast has been. I mean, we covered going back to the office, natural language processing—

PF It’s going an amazing journey for our editor. [Rich laughs]

RZ Salesforce puts values ahead of software, which is a meaningful thing to say, we can have a podcast just on that topic alone, to be frank. [music fades in]

PF And look, I know that I’m talking about enterprise software and that people are already going like, ‘what, what happened to Paul?’

RZ Well you can’t deny it’s value and success, you have to understand it, right? And it’s something I’ve been probing lately to just understand, like, why aren’t you failing? What is happening? Why aren’t you failing? And you know, it’s we’re always learning I think that’s key here. And I think it shifts our thinking about what we project as valuable beyond just software. We are Postlight. Check us out at postlight.com, lots of wonderful case studies. We’ve got some new ones up actually recently, lots happening at the firm, love to talk, hit us up hello@postlight.com

PF Yeah. We’re gonna share our values. We’re going to tell you how you can succeed and then and only then, we’re going to talk about software.

RZ Yes, values first, success second, software third.

PF Alright hello@postlight.com Let’s all, let’s all do good out there.

RZ Have a great week, everyone. 

PF I love you.

RZ Woo! [Paul laughs]

PF Back to the office, soon! Bye! [music ramps up, plays alone for 3 seconds, ends]