See all Insights

Back From Dreamforce And The Future of Marketing Analytics


Chris Butler: Welcome to the Newfangled Agency Marketing Matters podcast. I’m Chris Butler.

Mark O’Brien: I’m Mark O’Brien.

Lindsey Barlow: I’m Lindsey Barlow.

Holly Fong: And I’m Holly Fong.

Chris Butler: Lindsey, this is your first time on the podcast. Welcome.

Lindsey Barlow: It is. Thank you.

Chris Butler: You’re all grown up.

Lindsey Barlow: Happy to be here.

Mark O’Brien: And only your second time having four people, is that right?

Chris Butler: Yeah, other than the podcast we did with David Baker, I think this is our first time with four on the table, and our new set up accommodates that quite nicely. So you all just got back from Dreamforce, and we thought we’d just talk about what happened there. I received a bunch of excited text messages from Mark sharing some of the things that you all discovered, and I think our audience would love to hear about those things. One thing I wanted to mention though is that every year you all discover something. I’ve actually never been, but Holly, this is your first time,

Holly Fong: It’s my first, yeah.

Chris Butler: Lindsey, this is your third time?

Lindsey Barlow: Mm-hmm (affirmative).

Chris Butler: And Mark, your fifth, probably?

Mark O’Brien: Fifth, maybe sixth.

Chris Butler: Fifth or sixth.

Mark O’Brien: Yeah.

Chris Butler: And it seems like every year there’s something that you don’t expect to find out about at Dreamforce that when you do you say, “Gosh, I’m glad I went.” And it’s this massive event. For those of you listening and don’t know about it, how many people are at this event every year?

Lindsey Barlow: It was-

Chris Butler: It’s like a hundred thousand?

Lindsey Barlow: It was more than that, it was like 170.

Chris Butler: 170.

Lindsey Barlow: Yeah.

Chris Butler: It’s insane.

Mark O’Brien: It’s ridiculous. It’s insane.

Chris Butler: I mean, it takes over the city.

Mark O’Brien: Yes. Well, the financial district is owned by-

Chris Butler: And every hotel in that area.

Lindsey Barlow: Yeah.

Mark O’Brien: Oh, yeah. Yeah.

Chris Butler: So it’s this overwhelming haystack, I guess. And what I’m kind of curious about is every time you guys go there’s something that you come back and say, “Well, that was essential. I’m glad we discovered that.” But how are you doing that? How are you deciding where to be and what to see when you go, because there are so many events, so many people there, so many companies representing themselves, so many sessions you could go to. How do you choose?

Mark O’Brien: It’s gotten a lot more difficult over the years. The first time we went was in 2011, basically because Blair said I had to go with him, and so I went with him. At that point we had stopped using Salesforce. I did not like Salesforce, all the typical complaints. I had to use Interface. I was ugly, that I was clunky. It wasn’t very user friendly. None of the fields made sense for the way I operated as a salesperson, and we got rid of it. And I went back to a spreadsheet, and Blair said, “No, this is the future of everything that you guys are doing and everything your clients will be doing, and so you gotta go.” And that was 2011. At that point I think there were maybe 60,000 attendees.

Chris Butler: Still quite big for, I mean for-

Mark O’Brien: Very big. Oh, yeah. It felt massive. I’ve never seen anything half as big as that still, but it was easier then, because the landscape was simpler. The most notable thing that happened that first year was that we figured out what this whole marketing automation thing was about. And they are at the Expo 4, signed a contract with ACTON, who’s still our primary partner for all these years. And the tracks were simple like that, deal with Salesforce analytics, have separate ports and dashboards and what’s marketing automation and content strategy and SEO, and all those tracks are still there very much, but the digital marketing analysis landscape is so much bigger than that. And it’s not only the head count, going from 60k to 170k but the material and … That’s kind of the hardest part of it. We were talking about this when we were all together last week. It takes a day or two to figure out what’s going on.

Holly Fong: Yeah, and Lindsey gave me a great piece of advice when I got there and was putting together my agenda. And she had kind of said, “Stick with a lot of the Salesforce talks,” because a lot of the other one ones are other businesses promoting themselves in some sort of tricky way. But when you kind of stick with the primary Salesforce ones, a lot of those are pretty educational. And then I think we also split up into what interests us as individuals, and we all do very different things within the company, so that was helpful too.

Lindsey Barlow: Yeah, comparing this year to previous years that I’ve been, I spent a lot more time with the Salesforce products, going to product key notes and seeing what’s in tore for those. In past years, I feel like we spent a lot of time at the Expo looking at the sort of ecosystem of businesses that’s grown up around Salesforce, but in the years since then, Salesforce has kind of taken on all of those things. So they have a marketing automation platform. They have an advanced analytics platform that they offer, and so you don’t really see those other companies there. So it’s very much seeing what’s in the pipeline for them rather than looking around to see what else exists around them, things for me.

Chris Butler: Things for marketplace you mean, right?

Lindsey Barlow: Mm-hmm (affirmative)-

Chris Butler: That’s pretty interesting. I’m kind of curious, now that you’ve been multiple times, and we’ve adopted year after year more and more of what either Saleforce or what the Saleforce marketplace has to offer, going in, what were the things that you were maybe hoping to have validated in terms of our thinking by information that was being presented there? And then, what were the things that you were surprised by? I’m sure there’s something of each, right, where you went in yes we’re on the right track, or yes that’s still real, and then also wow, I hadn’t thought about that at all. That’s coming, better get ready for that.

Mark O’Brien: And actually, before we do that, we should let everyone know who’s who, what the roles are.

Chris Butler: Oh yeah. Yeah. So Holly you’ve on the podcast before, so maybe-

Holly Fong: I’m the Senior Digital Strategist here at Newfangled.

Lindsey Barlow: I’m our Salesforce Administrator. I’m not sure that captures quite the breadth of it, but …

Chris Butler: Well, you have created our Salesforce consulting program. You have offered that to clients as well. You manage our Salesforce instance, and you do a lot of production coordination as well here. You wear many hats.

Lindsey Barlow: I do.

Mark O’Brien: And again, we’re a wall-to-wall organization, and so being the administrator of that is a very big deal.

Chris Butler: It’s quite complicated.

Mark O’Brien: Very complicated. And that was another thing I saw in that very first year was a small company, probably 80 employees at the time demoing their wall-to-wall instance, where every point data points around the company was managed out of Salesforce, and that just blew my mind. And I thought it was a pipe dream, and thanks in large part to the expertise that Lindsey has developed over the years … Again, at that point you were with us, but you didn’t know anything about Salesforce. You’ve championed that whole growth track for Newfangled. Now, a lot of other people inside of Newfangled, primarily Chris, you and Dave Mello have been very, very involved in that, but being the Salesforce Administrator here, just the wealth of experience you’ve developed over the six years of us really to center around the entire company. And all the products on Salesforce is pretty impressive.

Lindsey Barlow: Yeah, well I would say that gets validated every time we go Dreamforce is that decision to do that.

Mark O’Brien: Right. Yeah. Yeah, that’s true. And Holly, as a Senior Control Strategist, it’s really about automation. Right?

Holly Fong: Yeah.

Mark O’Brien: Your crew is helping everybody make the most out of their automation investment.

Holly Fong: Yes. Definitely, and probably what I was surprised by there is that Salesforce in the past, and it still is to an extent, has always been … It works with a lot of different automation tools. And that’s something that has been really great about Salesforce is that you can kind of plug it in with all the different automation tools and build different reports based off of that. And they’re doing a lot more with Pardot specifically now, because those two are going to be so closely integrated.

That’s not to say though, and I think one thing that was really validating, that a lot of their reports are starting to build with Pardot. We’ve already started building with some of our clients, and we’ve already started doing with in Acton. So that was really validating, but it was a little surprising to see how strongly they’ve kind of veered in that direction as well.

Mark O’Brien: Right. Yeah, seeing that. The acquisition … They bought ExactTarget … I think it was five years ago now, and ExactTarget had just recently purchased Pardot prior to them being acquired by Salesforce, and so Pardot just kind of went along for the ride. And there wasn’t really a meaningful integration at the time, but now there really is much more of a meaningful integration, where Salesforce and Pardot are all kind of the same thing.

So, my answer to that question about what was affirmed and what was challenged … I go into each year with as much of a blank bind as possible, as if we were going into business that day. Right? And there’s nothing sacred. And that can be hard to get yourself in that mindset, but I don’t know. Going year after year and having epiphany after epiphany, it’s become easier.

But the other thing I went in assuming was that my mind was gonna get blown, that there was gonna be something that warranted the trip. And just to put it in context, the trip was really expensive. It’s a huge investment, the tickets. The flights are expensive The hotels are expensive. The event’s expensive. Then you’re there for three and a half days with three people. Actually we had five people.

I mean it’s a massive investment, and so I feel a lot of pressure to get value out of this thing each year. And what happens every single year in the first few days, because it is so crazy and hectic, and there’s so many things to choose from, and you don’t really know what the right new thing is, I always go through the point where I feel like oh this was a wasted trip. We shouldn’t have come here. And I almost texted that to you on the first day, Chris, that oh you’re so glad you didn’t come here. This is a bunch of crowds, and you’re waiting in huge lines for anything at all. There’s a massive line for anything. When you get six Chicken McNuggets, you’re gonna wait in a 40 person line, you know what I mean? And it would drive you crazy. And it was driving all of us crazy in various regards.

But then the second day, we found the thing. And then it set the track and gave me the guidance I needed to figure out okay, this is where I need to spend my time on the rest of the time I’m here. And then once you do that, because there are so many tracks, and everything works in your favor, there’s so many opportunities to learn about that thing and speak to so many people, because all the decision makers and the creators are there somewhere. And they can point, and you can go find those people and again, once you know what you’re looking for and so the big revelation for me was what they’re calling Einstein Analytics. Last year our big revelation was Wave, which Salesforce purchased, and then they rebranded it as Einstein, and I had thought, we all had thought that Einstein was just Wave rebranded.

And Wave is basically a data visualization tool that’s really beautiful and powerful, and we’ve completely implemented that, and it’s a big part of what we do now. But Einstein has affirmed our pursuit of machine learning and augmented intelligence, and Salesforce has made a massive investment in this. Now they have a tool that basically takes large volumes of data, and that’s the only input you give it is just the data. And it figures out the meaningful relationships, and the data tells you about them and then tells you what it thinks you should do next as a result of what it learned through that data batch. It actually does that. That’s amazing-

Chris Butler: It’s huge

Mark O’Brien: … and will have a huge impact on we can do for our clients.

Chris Butler: Right. I think it’s great to know that that’s happened, because one thing that I’ve been somewhat surprised by is that two years ago, November 2015, I had come back from a sabbatical, and we started talking about this. And I had a sense of urgency, because I felt like well it’s just a matter of time before other companies that we’re working with will unveil something like this. And so we’re not doing this as a means of getting there first, but it’s just a means of providing this to our clients before they want to jump shit to go get it elsewhere. Because eventually, they’ll be able to.

It’s gonna happen. And I just figured it would happen sooner. We work with numerous larger companies than us that have much more discretionary funds to throw at R & D than we do. And so we’ve been plugging along building our inside engine, and we’ve come pretty far on that, but it’s great to see that validated, especially … I think the Salesforce contacts can be a little bit more interesting for us in terms of what we would want to do with the data. Now you mentioned to me that there’s a limit or a base requirement in terms of data.

Mark O’Brien: Yeah, and we don’t know what that is.

Chris Butler: Right, but it’s a sufficiently large enough data set to make this functional.

Mark O’Brien: Exactly. There’s a minimum requirement for the data set size, and one of the main things I’m looking forward to from Lindsey is picking out pricing and what that is. And that’s the other thing. It’s so mysterious to figure out how they operate with these products. They spent all this money launching this, but then they’re kind of secretive as to how they actually sell it so there are many, many mysteries to be discovered.

And the fact that that exists … So Salesforce as a force in the economy … They’re so big. They move so fast. They do so much R & D. They create so much value, but they’re moving in a thousand directions at once, really more than that at once. And it’s very, very hard to understand well, what do you need and what do you not need? And I know the thing that was affirmed to me was our role as a distiller. We go; we immerse ourselves in this chaos, find the meaningful elements that we know are going to be very relevant to our clients’ businesses, bring it out, figure out how we can best deliver that to them and deliver the minimum viable product to them that’s not too expensive and make these really amazing technologies that are being used by the largest companies in the world available to our clients.

And the fact that we’ve been going down the path of machine learning and A.I., I think, see you and Dave, for the past two years we’re meeting them in step. Right? It’s not as like well, what’s this machine learning thing? No. We know exactly what this thing is, and we’ve been trying to figure out how we’re going to approach it so we can then take everything they’re doing and make that valuable and accessible to our clients. And that’s what Newfangled does.

Chris Butler: Yep. What’s intriguing to me about that direction is that the only purpose to pursuing it is to help us with the complexity. Right? Everything that we’re doing on the marketing side of things, it’s everything we’re doing now, but everything our clients want to do … It’s sufficiently complex enough to leave you stranded with the sense of not knowing what to do next. Not knowing how to interpret the data you have, not knowing what connections to draw, not knowing what to look at, what not to look at etc. And this whole machine learning thing, it sounds fancy. It sounds like oh, it’s a robot somewhere doing something spooky, but really what it’s doing is it’s drawing analytical conclusions that we’re just not able to make fast enough.

Potentially somebody, maybe somebody like Lindsey, would be able to sit down with a spreadsheet of data for a week or two and figure out something interesting just messing it for a week. But this is offering you that time back. That’s basically the point, but what’s really kind of spooky about it is that the more effective it becomes over time, the less sense that you have into how that conclusion was drawn. For instance, there was an article recently that, I don’t know if you saw, but it’s been making the waves, where people are talking about how on Facebook now, if you log in, people are often getting suggested that they friend somebody that they don’t believe Facebook should know that they know, like your therapist, something like that. And people are wondering well why is this happening? And so this is becoming an issue, because there’s privacy concerns. People are wondering well how are those connections being drawn? And the programmers who are responsible for writing or nurturing these algorithms, maybe that’s a better word at this point, they don’t even know.

Mark O’Brien: They don’t care, yeah.

Chris Butler: What they know is what they will produce, but they don’t know how they’re functioning at this point. And that’s what machine learning is really about is that you sort of create a cataclysm within this environment, and the machine itself starts to mix and match data points, add things on, run other queries, and do its own thing to the point where it becomes so complicated that you can’t tease out where that conclusion came from. All that you know is that it works.

Mark O’Brien: Yeah, that’s a little scary.

Chris Butler: Well, but that’s what we’re talking about here. You have a machine learning algorithm for Salesforce with the sufficiently large enough data pool, it should be able to deliver what we’ve been wanting to deliver for a long time, which is what should an organization like me, with problems like mine, with prospects like mine, with a landscape around me like mine, do next in this context? And we even named our webinar about this, that this machine will tell you what to do next. So I’m perfectly happy that Salesforce would have solved this problem before we did.

Mark O’Brien: Sure.

Lindsey Barlow: I mean one nice thing about it … One thing that I thought was interesting watching … I think it was … It was one of the Einstein keynotes. I don’t know if it was Einstein or Einstein Analytics, but part of how they’ve developed it, in listening to kind of the head of that product talk about it, was to make these suggestions and be transparent about them. So not knowing why it would tell you something, they’ve basically built in … It’s saying why it’s giving you the suggestion or why this is significant. And it also gives you the chance as a human to say it’s not significant. I think there was a healthcare example where the revelation was that …

Mark O’Brien: Only women give birth.

Lindsey Barlow: Yeah, only women can get pregnant.

Mark O’Brien: Einstein thought that that was amazing.

Lindsey Barlow: Right because it’s significant, but then a human can look at that and be like I don’t need to see that kind of revelation. So it’s nice that they’ve built that into it, where it’s like we’re doing all these crazy calculations, and then we’re telling you here’s what we found. Is this something that you care about? And so you can tell it, and it can learn from that.

Mark O’Brien: That’s quite cool.

Chris Butler: Yeah, that is very cool.

Mark O’Brien: It might be helpful to actually give an example of how an agency might use this in the real world.

Chris Butler: For sure.

Mark O’Brien: So today, with automation, we can tell them, after a blast, or when you’re doing a monthly meetings we can look at a month of blasts and say okay click through rate is this, open rate is this, SPAM rate is this, unsubscribe rate is this, and okay, that’s good, that’s bad, that’s on par, that’s not on par. With this system, we could go to them and say hey, in every October, when you send emails about branding to directors of marketing in the Pacific Northwest, it’s a good thing. And you can expect these kinds of results. It’s a whole different conversation. And there’s no way, as you mentioned, Lindsey could go away with spreadsheets for weeks, for that one case issue and try to figure that thing out, but you upload this data set in Einstein, and within a few minutes, it’s pulling that and a million other things as well.

Chris Butler: And even then, you don’t know why, but it doesn’t really matter.

Mark O’Brien: It doesn’t matter.

Holly Fong: As you guys are talking about all this, one thing that kinda comes to mind for me is just how important data quality is given everything we’re talking about right here. And so, as machines are learning, and they’re basing decisions off of the data that’s input, it’s super important that the quality of data that clients are inputting into their Salesforce and that’s getting input through your website, through your marketing automation system into Salesforce is accurate. And so one thing I took away from there is for our team to really focus on how can we make sure our clients are getting quality data through their website into the CRM, because that’s super important and then making sure that they’re inputting data in the most quality way possible as well.

Mark O’Brien: Yeah, and that’s actually why it’s so interesting is to think about that, the clients’ main responsibility is just data management and housekeeping, and if that’s good, then let the machines process it and tell us what matters. But we’ve built a business around trying process this stuff and trying to figure it out. But now we can build a business out of making the processor work for us and bring the most relevant revelations to our clients, and that’s cool. I’d rather be doing that work.

Chris Butler: Yeah, absolutely. I’m kind of curious about that. I hadn’t thought about data quality and tending to that. What kinds of issues do you see now that you think this could start to matter to us, stimulate on your team ways of thinking about fixing that? What kind of data quality issues do you observe regularly?

Holly Fong: The biggest thing comes from open text fields on websites and those being a part of forms and the number of things that people will put in there and then that coming through into a CRM. And there’s a few issues with that, right, so one of which is basically impossible to segment by, because there’s so many variations of what people put in there. But it’s also gonna be next to impossible to create any intelligence based off of, because there’s gonna be so many variations, again, so it’s really important that when we’re helping clients figure out what those progressive profile fields are going to be, that they’re using ones that are gonna actually benefit them and that they’re gonna be able to use that information based off of … especially because there’s gonna be a limit to the amount of data they’re actually getting in. Of that data, there can only be so many subsets. If there’s a bunch of subsets, then it’s gonna limit their chances of actually being able to make a decision off of that.

Chris Butler: Well that’s actually a really good point about the size of database that would be required in order for this to be functional, because the larger the data set, the more a machine could extrapolate, extrapolate contextual information that isn’t provided, that would allow it to figure that out. For instance, the example you mentioned, that all people giving birth who are women, right? If the machine doesn’t understand that that sort of how human biology works, that’s a contextual rule, then of course it’s going to think it’s a conclusion it makes based on that. But if it knew that because there’s a wider repository of data to tell it that, then it would ignore that. That’s why I think Facebook is able to make such spooky determinations, because they’ve got such an enormous data set.

Mark O’Brien: Well, and Facebook’s best data set is to actually counter your opinion and throughout their open text fields. Right? And this is the advantage of splitting up at Dreamforce and going to different sessions. I saw a session that was all about open text fields. And a limitation that we’ve had plenty of times is oh it would be great to measure this, this, and this, but oh our past ten years worth of data, we didn’t have that field broken out in Salesforce or on our website, right? It’s a big problem. So what Einstein does is it mines open text fields and figures out what’s actually being written about any open text field and then how that relates to all the other data points and creates indexable fields based on what it thinks is relevant inside of all these open text fields combined.

Chris Butler: Yeah, that’s very cool.

Mark O’Brien: Which is whoa, like that. That’s just one of these paradigm shifting ideas. Everything you just said I would have said a week ago. Right?

Holly Fong: Well, I still think though, that it’s relevant. I mean an open text field for notes versus an open text field for what that person’s title is, is very different. And to your point about the open text field, especially for like notes or description or something like that, another really cool example they gave, which I liked, is that they had a company that was basically helping to make sure first generation students were staying in universities, right? It was mining the comments section of that, and so when these support case workers were working with individuals who were going to those universities, it was looking and seeing oh, that person lost their housing this week. And then it could tell Einstein that that person has a higher drop out rate, because it’s mining those text fields, and I think that that’s something that a lot of people could use when it comes to retention of client bases and potential of opportunities closing and things that they might have said on the call if it’s mining all of that information.

Mark O’Brien: Right, yeah. I think all of the ground rules you stated are necessary ground rules. And the better you are at that, the more you’ll get out of it. But the fact that it knows that we can’t reverse engineer that stuff, and it’s going and finding all of the old things and doing its best is just shocking to me.

Chris Butler: One thing that I expect we’ll see in the next few years if this delivers on the promises that you’re describing is probably a pretty quick immediate or immediate or quick impact on progressive profiling. Because right now, progressive profiling is about identify all of the discrete data points that you think you need to know in order to properly vet a prospect and prescribe the order in which you ask for those things. And then map them in a specific way, but in theory, if this works that way, it could sort of identify its own course of progressive profiling without you even having to identify what those fields are. And that would probably be the ideal way to do it, so it’s almost like segmented, automated progressive profiling, where it knows okay well, I’ve received this thing, and I might have this information. So based on that, this is the direction I’m gonna head in with gathering data, which would actually be much more interesting than what we do now in theory.

Holly Fong: Yeah, I mean that’s something that Acton’s already started kind of working on, which is conditional branches with the form. So, if they say one thing, they say I’m a CEO, then the next question is different than if they said I’m a marketing manager.

Chris Butler: Right, but you decide what that next question is.

Holly Fong: Yeah, it is.

Chris Butler: What I’m envisioning here is that it would do that on its own, and so you might actually discover a lead, I don’t know, hypothetically, and you realize oh that branch is completely different. That was created by the machine.

Holly Fong: Got it. So it’s coming up with unique questions, based on that?

Chris Butler: Yeah, I mean in a way it would obviate the need for progressive profiling the way we do it now, which is … I mean the way we do it now is hypothetical. It makes sense. But you know how much our clients struggle with that. They ask us well, which fields should I put here? And a lot of times they’re really uninteresting, and I think what would be interesting about what a system like this could do is actually start to insert questions that have a higher likelihood of being answered accurately, that have a higher likelihood of contributing more on your actual nurturing path, which you would never think of, because you just don’t have access to enough data to make that determination.

Mark O’Brien: Well, and to even take it a step higher than that in terms of autonomy, say you close three jobs in a month, it’s gonna see which jobs were closed, and based on those jobs and the prospects and contexts associated with those jobs, learn more about the ideal type of prospect, and then just on its own, start looking more for those kinds of people, because it learned through the closing what was more valuable and who in that database represents those kinds of people.

Chris Butler: That’s a great idea. I was actually just thinking of that too. I think what, has it been every year you’ve initiated a new contacts purchase?

Mark O’Brien: No, well, maybe every year or two.

Chris Butler: Maybe every year?

Mark O’Brien: I guess so, yeah.

Chris Butler: It would be interesting if, given a system like this, you had a monthly update where it said here are 10 people for you that I suggest, and you could approve or deny. But that would be a better use of your time. But yeah, I mean, this kind of thing could open so many doors. It would be interesting for us to figure out where the best time spent for consultants such as us in that kind of condition, because for example, if you have a system that’s doing the progressive profiling for you, who know what else it could do? And then what is our job in that scenario to evaluate that data? I’m sure there’s something significant, but right now we spend a good amount of time helping people configure that logic.

Mark O’Brien: And with any of these things, you could choose to look at this as something positive or something negative. And I don’t know. For me, every time it’s both. It’s intimidating when I see the future, which I do every single time I go to this conference, but it’s as inspiring as it is intimidating every single time.

Chris Butler: Yeah, well all this technology requires people like you all and people like us to help negotiate transition. I mean that’s really what … It’s like the definition of leadership and management, dealing with complexity, and we are offloading more and more management to systems, which is good. But leadership is about helping broker change, and I think in this case, that’s where the line starts to blur, right, like a sufficiently complex machine that sort of thinks air quotes for itself can initiate change, but I think you still need a human to help another human get onboard with that change. And that’s an interesting opportunity for us.

Mark O’Brien: There’s plenty of it coming.

Chris Butler: Yeah. Any last thoughts about Dreamforce?

Mark O’Brien: I’m booking hotels as soon as possible to-

Chris Butler: For next year?

Mark O’Brien: … pay as little as possible.

Chris Butler: Yeah, I’m sure there’s some gouging going on.

Mark O’Brien: Well, that’s its present strategy, right? And that’s AOK. But there’s just so much demand. And so … Yeah.

Chris Butler: One last question for you all. So a bunch of years ago, Mark and I were at a HOW conference, and at some point during the event, Mark said that the greatest thing that he had learned there was a new idea to cupcake, because one of our mutual friends who was there introduced … I had never seen it before, but she ripped off the bottom of the cupcake, put it on the top, and you ate it like a sandwich. Genius, because otherwise there’s no elegant way to eat a cupcake. I’m curious. Were there anything like that that you experienced, like good restaurants, something funny that happened, something fun?

Mark O’Brien: We didn’t have any good food. We didn’t do anything fun at all.

Chris Butler: That’s a lie.

Holly Fong: I had the worst bottle of wine in my entire life there.

Mark O’Brien: Holly learned how bad wine can be.

Holly Fong: Yeah, it was super cheap too.

Chris Butler: It was actually bad?

Mark O’Brien: No.

Chris Butler: It was a really expensive good bottle of wine.

Mark O’Brien: It was not cheap, but it wasn’t expensive. It was just amazing. It was really, really good. Yeah, and that’s the thing. That city is such a great food city although we tend to find and get stuck in our habits.

Chris Butler: What restaurants did you go to?

Mark O’Brien: Sushi Ran in Sausalito.

Chris Butler: I’ve been there. Oh no, I have been there but not with you.

Mark O’Brien: Yeah, that’s weird.

Chris Butler: Yeah, I went there.

Holly Fong: Mark bought a jean jacket.

Mark O’Brien: I bought a jean jacket with the Levi’s.

Chris Butler: Yeah, I heard that was coming. You’re not going to do the whole denim and denim. You’re not gonna do denim pants, denim jacket …

Mark O’Brien: We’re calling it Jeanforce. Yeah but they’re different colors.

Holly Fong: It’s a black jacket.

Chris Butler: Oh, that’s cool.

Mark O’Brien: Oh, you’re doing that right now.

Chris Butler: This isn’t denim.

Mark O’Brien: Yes it is.

Chris Butler: No, this is twill.

Mark O’Brien: What’s the difference?

Holly Fong: Yeah, that’s more-

Mark O’Brien: It looks like denim.

Chris Butler: Denim is its own thing.

Holly Fong: Is this thicker?

Chris Butler: This is not denim. Just for the record listeners, Mark is describing a grayish-green work shirt that I’m wearing.

Holly Fong: It’s definitely not denim.

Chris Butler: There’s no resemblance to denim whatsoever.

Mark O’Brien: It’s got a Def Leppard iron on the back.

Chris Butler: It does not. So you guys had a good time. That’s good.

Mark O’Brien: We did, and it was so intense, and it was great to get home. I think we spoke three words to each other the entire trip home. We were just done, but it was wonderful, tons of value, tons of fun, great mind opening experience.

Chris Butler: Well, typically we wrap up by showing some content that we think someone would get to value out of looking at, and I know that none of us prepared to do that, because we were wanting to talk specifically about this, but it occurs to me that in relation to what we’ve been talking about, you should really check out the last two webinars we did, which are called The Firm of the Future. One of them features Lindsey and Mark talking about the wall-to-wall experience that Mark mentioned earlier, and the second involved David, I, and Mark talking about machine learning and the systems that we built around that. So if any of this stuff is of interest, and you haven’t been to Dreamforce and you want to cut your teeth on it, start on our site, and other than that … I think you talked about Dreamforce in some other content in years past as well, so people can catch up on that.

Mark O’Brien: We did a newsletter on it years ago.

Chris Butler: There was.

Mark O’Brien: … as part of pitch for Dreamforce.

Chris Butler: There might have even been a video, like one of our casual Friday videos.

Mark O’Brien: Oh wow.

Chris Butler: Actually I think last year we did a podcast as well.

Mark O’Brien: Wow. Gosh.

Chris Butler: So there’s some material on Dreamforce, but thank you all for talking about this.

Mark O’Brien: Yeah, it was a lot of fun.

Chris Butler: Yeah.

Mark O’Brien: Thanks Chris.

Chris Butler: Alright, see you all next time.

Mark O’Brien: Bye.