How to build a great chat platform for remote teams

+ Add to

Joining me is someone who created a way for us to talk with the people we work with remotely. There are hundred of tools out there that can be used for the same thing so I want to find out what made his solution stand out from all the others.

Rajiv Ayyangar is the founder of Tandem, a chat platform for remote teams.

Rajiv Ayyangar

Rajiv Ayyangar


Rajiv Ayyangar is the founder of Tandem, a chat platform for remote teams.


Full Interview Transcript

Andrew: Hey there, freedom fighters. My name is Andrew Warner. I’m the founder of Mixergy, where I interview entrepreneurs about how they built their businesses for an audience of entrepreneurs. Joining me is someone who created a way for us to talk with the people we work with remotely. And I know, you know what run Rajiv.

I feel like every time I say that, and I’ve talked to people about the fact that you and I are talking today, they say, oh, I’ve seen that obviously. And then they think. Well, there’s zoom and it’s, that’s it. And the other approach just have not worked, but what you’ve done with tandem is you’ve created a sense of presence, right?

Yes. We could do video chat, like we can with other video apps, but in addition to it, we can kind of see what people are working on you. Should you send me a screenshot of what you were doing the other day? And what your teammates were doing. And I could see that there was one of them, I think, in a Google doc, I could have sworn I saw someone in notion, but I’m not sure.

And then right, little icon next to their name. See what they’re doing. You could wave at one of them. If you want to just kind of check in, if it wasn’t critical or you could jump in and start a conversation. Am I describing tandem? Right?

Rajiv: Yeah, there’s the sense of presence and all these little cues that you mentioned, and, you know, a few more like times zone, whether they’re in a meeting who they’re talking to add up to this sense of who’s around in your team and like these little micro stories about who’s around, like somebody late to a meeting, somebody is talking with somebody else, somebody deep in deep focus in code or in design.

That presence is one of the things that unlocks a more conversation like you’re in the same.

Andrew: That sense That you had when you were in an office with someone and you’d kind of look over their shoulder and see what app they were in or see that they’re kind of staring into space and maybe thinking, or could tap them and say, Hey, I’ve got a question. That’s what you’re doing with tandem. I should say, Reggie.

I and Gar is the founder. I invited him here to talk about how he did it and we could do it. Thanks to two phenomenal sponsors. The first we’ll host, uh, the hotel, oh, HostGator was such a long time sponsor. I was about to do one of their ads. No, it is. It’s a company that will help you do marketing automation.

It’s called send in blue, and I’ll tell you later why you should go to send in And if you’re hiring a developer later on, I’ll tell you why you should check out my sponsor lemon at is it true that an investor wanted so badly into your round that they sent you a tandem bicycle so they can invest.

Rajiv: That happened? Um, I think, I think, you know, uh, things get crazy when fundraising, but you know, of course fundraising’s not really a reflection on you or your company. It’s just a moment in time.

Andrew: Did you take their money?

Rajiv: We did not.

Andrew: Oh, wow.

Rajiv: But great investors, you know, a lot of factors at play there

Andrew: is it a small investor or a, or a bigger venture firm that has a personal touch like that?

Rajiv: uh, that was a bigger venture for him. You know, I think the, if, if you know, people have been through accelerators, like Y Combinator, the, yeah. The demo day is a, it’s a forcing function. For fundraising. If fundraising is going to occur before demo day, it’s a backstop where right after demo day, you have the promise of you’re going to talk to a bunch of investors and really open it up.

So it creates a lot of FOMO and time pressure that in normal fundraising is very, very difficult to create. Uh, and so you can get, you can get people trying to do anything to, to cut through the noise and similarly, The founder’s side. You’re trying to get to know these people to build a very long relationship in hours or days.

And so it’s a very, very crazy, crazy environment.

Andrew: Kind of exciting. How much money did you raise?

Rajiv: Uh, we, we raised seven and a half million

Andrew: From

Rajiv: from Andreessen Horowitz.

Andrew: impressive. I do though. Like despite the fact that It didn’t work out with them, I love personal touches like that. When somebody can see you and send something that says I’m, I care enough to put in the effort, I care enough to find something. I like those personal touches. My problem with them is what do I do with it afterwards?

I guess if it’s a tandem bike and your company’s called tandem, you can display it or keep it in the house and you have a good story.

Rajiv: It was surprising and appreciated. And, uh, you know, I think it w w it was just one component of like what we were, what we were seeing from the other investors. But I think the thing that we. Really we’re excited to see. And we’re a little surprised to see is that this promise of remote connection, remote collaboration was, was seen by other people.

And they added to our vision right? In those conversations, even the investors we didn’t bring on board. Um, and that was cool to see, because this was before the pandemic, our, our kind of focus on remote work was a little bit unusual and the remote work. It was not seen as the dominant market. And so this point of view that like remote is the future was extremely controversial in 2019, and not at all controversial, you know,

Andrew: Especially fricking out in Silicon valley, it was, it was more so than anywhere else. Right? If you paid for a San Francisco home or a south bay home and an office there, and people around you had done the same thing. You are buying into the cult of that’s in culture. And so anyone who is outside of that, you felt they didn’t see it.

You felt they didn’t see what it was like to have that comradery. And also you didn’t maybe notice in yourself that you had been so highly invested in it and surrounded by people who are invested in it. So I could see how that was a problem. Meanwhile, though, this wasn’t where you started you, you were a guy who started working at Yahoo, um, and then.

Well, yada yada, yada, we’ll skip over some of your timeline, but you got into crypto. What do you think was missing in the crypto space that this, that that’s what you created as part of this company before tandem?

Rajiv: So I started my career in data science. I was a data scientist on the founding team of a productivity app that, that got acquired by Yahoo. And then I was a data scientist at Yahoo. And so I’d spent a lot of time in analytics and how people understand analytics and not just the doing that, getting of the answers, but communicating those answers to people.

And there was a, uh, there was a real difficult problem, uh, in kind of late 2018. Uh, Okay. You’re investing all of this. You’re you’re trading like people just regular consumers are retail. Traders are trading almost like hedge funds would have traded in regular finance and they do not have the same analytical tools.

And so, uh, what we, what we saw in me and my co founders was the opportunity to create. Uh, consumer portfolio tracker that gives you clarity on your crypto trades, the ROI, uh, the ability to see different reference currencies, the ability to automatically integrate with a lot of exchanges, which was not at all a thing at the time, because it was so hard to integrate with exchanges and give you

Andrew: Coinbase And, Binance didn’t know exactly what I had in either one until I went in and looked in both accounts and then maybe added to my spreadsheet. You said, screw that. We will go in our software. We’ll grab your data put it into one, clean, easy to digest chart and table. That’s what you were doing.

I asked you before when

Rajiv: Yeah.

Andrew: data explained data density, because you’ve actually told me this several times and I didn’t appreciate the significance of it until I started noticing it everywhere.

Rajiv: this, I think the first time I kind of encountered this idea was from Edward Tufty in, uh, visual, uh, the, you know, um, the visual display of quantitative information, uh, where he talks about what’s the data to ink ratio. You like for, you want to maximize the data to increase SHEEO and what we call it in user interface design?

Um, my, my, my co-founder and design. Coined this term pixel power, like what’s the maximum pixel power you can have. And so the first and this, this actually connects to tandem. So the first, the first time we encountered that was when we launched our portfolio tracker. It was called crypto gun. It was one of the top portfolio trackers.

Uh, it was enterprise grade because my co my other co-founder had built hedge. Tools for Palentier and Bridgewater. And he’d been familiar with enterprise grade portfolio tracking. It was, it was enterprise grade, had a lot of data density, but unfolded it in a way that was understandable. So you could dive deeper and deeper into the data and get the answers you wanted.

And similarly with tandem, we give you as much density of signs of life and presence and inform useful information about your team. In a small compact app that lets you sit, you can sit the app to the side of your work. So it’s almost like your peripheral vision has your

Andrew: Oh, you imagined, oh, that’s what tandem does now. I, when you are doing crypto dashboards, you weren’t thinking this would be up and present on people’s screen throughout the day. Right? This.

Rajiv: no, no. The crypto dashboard was immersive.

Andrew: Got it. By the way. I think the first time that I heard and understood the significance of this was when I talked to Alexis Ohanian and he said that Paul Graham told him when he was working on Reddit to squeeze in as much data as possible.

And I could have sworn, he even said, get rid of the L the Reddit logo. Why have that be the top left thing? The first thing people see, and they were really good at cramming, a lot of data and links into their home screen lately. It’s. The opposite of that. I’m on there right now. And all I see is one post that’s taken up my whole screen.

It’s, it’s completely the opposite, but I’m with you on the importance of data density. You and I also talked about how you got data into your dashboards back when this wasn’t just an easy thing to do. You weren’t allowed to, to log in on behalf of users and scrape their data from the different exchanges where they had their crypto currency.

How did you do it?

Rajiv: I, my co-funder would remember the details more, but it was a lot of manual work, a very custom work for every exchange. Parade of the exchanges and started with the top ones and then it would break off. And so we had to either create automated tests or create some cadence of like manual testing, uh, to make sure we fix them as the, you know, either the API or whatever, you know, export interface.

We were using a change.

Andrew: Okay.

Rajiv: work, but valuable in terms of the, it just removed an entire step from the UX and, and give our users way more.

Andrew: All right. I know why you shut it down and pivoted to tandem, but let’s understand what worked about it before we understand what didn’t. How did you get anyone to sign up? How did you get anyone to use this?

Rajiv: well, we, we launched, we just launched it on product hunt. I think we, we had initial validation from just mocks. So the first thing we did was we created a mock in, I think it was in sketch and then we just, uh, posted it in a Facebook. To see if people were interested and we got a ton of responses, like take my money now I need this.

And so we, we didn’t start building. We actually just asked them to fill out a survey and started talking to them and try to understand like what feature set they would want, uh, in a portfolio tracker and then, and then built that. And then. We just, we launched it on product hunt. That was our first experience, launching something on product hunt.

It made, it made a splash. I, it wasn’t, it wasn’t even number one. Uh, but it made a splash. It brought a lot of people who were watching that space and investing in crypto into the platform. And, you know, very quickly we had a, you know, relatively small amount of users, but a lot of. Cryptocurrency tract. So about like a hundred million dollars of cryptocurrency tracked on the platform, which was kind of interesting.

Andrew: Why is that the metric that you cared about back there?

Rajiv: Well, it’s, it’s a, it’s a metric of, so I think there are two metrics we cared about. One is like the user engagement and daily active users. Uh, and it was in the low, low thousands. And that was the metric we, we really watched in the short term, but the metrics. Of total volume kind of points to a couple potential business models where we might say, Hey, if we’re going to provide exchange services on top of this, or if we’re going to provide ads on top of it or any sort of service, it’s probably more valuable to an advertiser, or we can get more exchange fees for people who are.

A lot more and encrypto is in a lot of finance. There are huge whales, right? The people who trade an order of magnitude more than everybody else. And we want it to understand that that Pereda,

Andrew: Where are you thinking of becoming an exchange?

Rajiv: we saw kind of three paths to monetization, and now that’s why we pivoted away because we didn’t feel like we could do any of the three paths. The first path was. Uh, ads for ICO’s and that seemed pretty, pretty difficult to do in a way we felt okay with in 2018, there are a lot of scams. Um, the second, not impossible, but just difficult for us.

And then the second, the second one was tax, uh, tax services. And that’s where we, we didn’t feel like we had the competency to do that. And, uh, one of our competitors coin tracker did amazingly and that’s. Uh, and, and then the third one was becoming an exchange and that’s where we, again, felt like we didn’t have the, um, kind of the, the financial and legal, uh, background to, to really pursue that and make that amazing and do it in a safe way that wouldn’t get us, you know, slapped with a ton of legal fees.

Andrew: All right. And then since you couldn’t figure out a way to make it work financially, you said we’re going to close it down. Did you know the tandem was the next thing before you close it down?

Rajiv: no, we, we started exploring, um, and, and tandem was not the next thing. There were two more things before tandem actually, but we started exploring around and I think that’s a very interesting moment because it was similar to our exploration before we launched crypto con um, we looked at. All sorts of ideas.

We tried to come up with something that we felt high conviction on together, and ultimately we ended up building a personal finance app, so the same market, but solving a bigger problem for them, which is clarity on all of their personal finances.

Andrew: And how did that do

Rajiv: We were not able to get traction. And that was a really interesting, like we had 24. People 25 of our friends use it. And we really believed in this idea. I think we had a lot more conviction on the problem. Like personal finances are so tough, helping people manage them, helping people get out of debt, like prevent them from going into debt.

That’s such an important problem, but finances are stressful and ultimately it is less stressful to just not open the app than to use the app. I think that was our first experience with. This is a cure, not a vitamin, not this is a vitamin, not candy. Like nobody actually wants this, even though people may need it.

And we over a couple of months of iterations, we did really fast iterations. Like we were in coffee shops again, pre pandemic we’re in coffee shops, testing with five people, changing the prototype testing with five more. We just could not get people to use it. And so, although we had a little, really high conviction in the product, we just had to realize we can’t, we don’t have a business model and we can’t, people just don’t want to use this, even though it’s good for them.

Andrew: That’s a really hard thing to accept, especially when you put in that much work and time where you running off of your personal savings at that point,

Rajiv: we had raised a kind of small, you know, friends and, you know, kind of friends and family round, um, before, uh, for the crypto. Uh, for crypto gone. So we had some, we had some runway, we had not hired a team, just, just the three founders. So we’re, we’re staying pretty lean. Uh, but we had some, we had some time, but you know, not, not infinite time.

Andrew: By the way I went to product on, when you said that that’s where crypto gone got its early traction. And I saw that right underneath it. You’d created something called banquet by delectable. Was That one of your apps?

Rajiv: That was so I worked at delectable as a product manager and, uh, I was the product manager that, you know, helped them launch banquet. Uh, so that was a second app for delectable. That was a white, flexible, it was like Instagram for wine. Uh, it lets you see what wines your friends had drinking identifies the labels and then banquet.

Uh, a secondary app for wine stores. So as a wine shop, you can in an hour create a mobile wine store using that computer vision.

Andrew: All right. I guess the thing that I’m trying to understand that. Where do you get the nerve to say, we have something working. We’re going to shut it down instead of, you know what, it’s not working. If we keep at it, maybe something will happen. And the reason I’m asking is I’m the kind of person who just keeps at stuff at everything.

I’m not, I took to heart, the never quit, never give up message. And I admire when people can quit and give up when it’s right. How did you know when.

Rajiv: That’s a really interesting question. And I think so having kind of been through several iterations of this, like we’ve changed directions in a major way, four or five times, and having talked with a lot of friends and other founders, who’ve been through similar pivots from small pivots to huge pivots. I think it’s very dependent on a number of factors.

So one it’s dependent on your timeline. Like, do you have, have you raised a little bit, like, do you have runway, do you have personal runway? Like what are your kind of expectations that, that kind of determines how deep you explore any one idea maze before you decide to like change. But then I think the second I can describe what.

Um, what we did was every week we would track learning rate at a really high level, each of us individually, low, medium high. And then we would try to identify what the biggest risks for the business are. And I think this is the challenge in the early days. The thing you’re most afraid of is likely the thing you need to work on, and probably the thing you don’t want to work on because you’re afraid it’s not going to work.

And so we would try to work on those big, big risks and validate. Uh, which is often not writing code, you know, and, uh, if we’re not learning for like three or four weeks in a row and, or we’ve kind of validated a risk, a major risk, and we don’t see a way around it, or we don’t have to see a way around it with our resources, then we have to decide to change.

Not saying it was easy, but in retrospect, a lot of these decisions made, made a lot of sense. Like there was a clear dead end. I think we could have reached them a little bit earlier.

Andrew: You would, you would have some dashboard that told you how much you were learning every week.

Rajiv: It’s subjective, but in retro every week we would get together on Friday kind of review what we did that week. The metrics, you know, measuring things in the early days with no users is difficult but worthwhile. So for example, with the earliest version of tandem, our metrics. Can we have five conversations with teams and they don’t tell us they’re completely not interested.

That was the first metric,

Andrew: Interesting.

Rajiv: mild curiosity. And then the second metric was interest because we’re changing the idea and how we describe it. And then the third metric was, I want to use this, like, can we get a couple teams to say, I want to use this and what do they want to use? And over this, so we’re tracking very crude metrics, but then we also ask.

For us into each of us individually on the different areas where exploring learning rate low, medium, high, really simple, you know, if your learning rate’s not high, then what we get it to high. And that’s often the most useful part of that question. Um,

Andrew: Okay. I guess so it’s more like, but, but then again, that still doesn’t tell me to move on. You’re learning a lot and it feels like when you’re stupid, you’re learning the most. And when people don’t love you, you’re learning a lot about why they don’t love you. So the rate keeps improving. It feels like then you’d just be stuck doing something that’s not working, just because you’re being told.

It’s not working and you’re learning a lot of reasons why it’s not instead of giving up because it’s not working and moving on.

Rajiv: Well, if it’s not working, then you have things to try generally. And, you know, in the really early days, you know, some of these pivots, we had basically no active users, right? So nobody’s telling you, it’s broken. If people are telling you it’s broken, that’s great.

Andrew: Okay.

Rajiv: Indifference is the thing that causes you not to learn.

Um, and so it, you know, in those early days, uh, just kind of getting no new information or feel it. You know, feeling like this thing is broken, but we have no control over it. And we haven’t made any progress there. And in, in know, a few weeks that’s kind of the signal, but it really depends on how much conviction you have in the idea how many users signal, how much user signal you have, you know,

Andrew: I liked the idea of learning rate. If I’m understanding you, right. If I were to do a learning rate for myself, it might be at the end of the.

week, just saying I’m trying something new. How much did I learn? Let’s just write it out in bullet points. How much smarter did I get? Is that it?

Rajiv: that’s one version of it. Yeah. Yeah. It’s different in different phases. Um, no, I think that’s a version when you have, when you have traction and you’re deeper into a project. In the really early days, I think the value we got out of learning rate was putting three facets together. What’s the biggest, what’s the biggest reason this may not work out.

That’s the thing to learn. Like the second is what is our learning rate on that thing, that big risk, the thing that may kill the company. And then the third is if it’s not high, what would get it? And that kept us reoriented towards working on the biggest risk. Um,

Andrew: And the

Rajiv: but in later phases, yeah,

Andrew: people are not going to use this on a regular basis and we’re not learning enough. That’s telling us how to get them to feel comfortable, opening it up on a regular basis. That means this is not a good learning rate. You’re nodding

Rajiv: it was in that case, it was actually two things. It was, we could not get people to use it on a regular basis. And despite. Interviews and even getting, trying to get our family to use it, friends and family to use it. We could not get people to use it. And the second one was in terms of a business model, we were exploring lots of different business models and we couldn’t find a path to monetization, um, that we felt good about.

It’s difficult to like charge users who are financially.

Andrew: Uh, yeah. Alright, Rajiv. The other thing that I’m taking away from this is alternative metrics. That it used to bother me in the beginning of Mixergy, where someone would say, just look at your metrics, see how many people are coming to your site. What the bounce rate is, it’s meaningless when you’re starting.

And there aren’t that many people who come in. And then I remember talking to Mike McDermott of FreshBooks and I said, how do you, how do you know you’re on the right track? And he said, truthfully, I didn’t know he was running a consulting company. So it’s not like. Like starving, but he still had to put a lot of time and effort into FreshBooks this a invoicing software, which has now expanded into something much bigger.

He said, I just looked to see how many people are saying, thank you, how many people liked it? And that was a metric he would look at. And it seems kind of touchy, feely and silly, and not really as, as quantified. It doesn’t seem as serious as looking at your actual dashboards and seeing what your metrics are.

But I think alternative metrics in the early days are really meaningful for you. It was can we have five conversations with teams who don’t tell us that they won’t need any of this? That’s that’s significant. How many people tell me that they like this is significant, right?

Rajiv: Yeah, it really is. And it, it helps you move forward because like the early days really do feel like you’re lost in a maze and. The, the metrics, breaking that down into more discrete steps rather than this product is completely working and people love it, or it’s not like that. That’s kind of the dichotomy you tend to exist in, in the early days.

But if you can break that down into measurable steps, like I like this idea. I use this idea. People I’ve talked to understand this idea. That’s, that’s an important one. Don’t, don’t skip that. If they understand, if they don’t understand it, then nothing else matters. People understand people understand it and are curious about it.

People understand it and like it, people understand it and want to use it. People understand it, want to use it and know how to bring their friends in. People are using it. Like you can break all these things down into more granular steps and then work your way up that.

Andrew: Alright, I’m going to talk about my first sponsor. It’s called sendin blue. I’ve talked about them. I think. And, you know, in a way that’s too small, I always say that their email marketing, but they do so much more. When you think about the marketing stack, when you, for your company, you think, yes, you need email, but you also need landing pages to come meet.

Collect email addresses. And then you go and you get a Zapier connection that connects the landing pages to the email software. Uh, you also might need one of those chat widgets that goes on your site so that you can communicate with people who are interested in buying and understand why they’re hesitating or answer the questions.

You might also want SMS marketing. I’ve talked to people here on Mixergy who say we don’t get that much from email at anymore, but they’ve said in their businesses. SMS text-based messages do even better. You need tra you need all these things and segmentation. What sendin blue does is they combine it all in one easy to use package.

That is super smart. That’s why I call it marketing automation. It’s intelligent. You get automated tasks and actions based on what people have done and it’s inexpensive and it stays low. Unlike some of these email marketing packages that.

are really cheap in the beginning, and then later on. They just whack you over the head with a price so high that the most expensive piece of software you have is your email marketing CRM software.

Send them blue can do all that inexpensively, easy to use. If you hadn’t heard of them by now, then it means, first of all, you probably haven’t been listening to enough Mixergy because they’ve been buying ads because it works here. People are signing up, but also I think they’re just not playing in our world.

And I think it’s time where, where them, this is a company that raised according to Crunchbase hundred $97.7 million. Credibly, well-funded incredibly good with our customers. And it’s a big package of software that does all the things that you’re going to spend a ton with others, but it integrates it and works really well together.

And if you use my URL, they’ll even give you a lower price than other people do. All you have to do is go to send in that send in. There it is. I love that they even put the mixture to name and a message to Mixergy listeners with their big discount, uh, right on that URL send in

Um, all right. You decide. None of these things are working. We’re going to pivot again. And all of this happens because one of your co-founders, I guess the three of you were working in person. One of them had a baby and couldn’t work in person, had to work from home. My understanding.

Rajiv: Yeah, well, and too two of them are actually both of them had, uh, babies or, you know, kind of around the same time. And so in the early days of a startup, you’re not taking much paternity. You’re just kind of trying to work with the, with the baby, get a coating with baby. In one hand was a very common site.

I got really used to. Baby’s crying on calls and got pretty, pretty comfortable with it. Um, and so we’re working remotely most of the time, and that was a huge shift ’cause before we were working in person, we would lug around a whiteboard set up to each other’s houses to like, uh, to really be able to think together.

And that was a huge shift. We didn’t feel like a team anymore. We weren’t talking in the same way. We weren’t talking as much. And this happened at a phase where we were trying to figure out the next thing, and we felt this huge gap. And so we started asking why don’t, you know, at first we just started to solve our own problem.

Like what tool can we use to solve her problem? And we tried basically everything, uh, with discord, maybe being the most promising, but, uh, we found

Andrew: you actually use

Rajiv: us that we did. Yeah,

Andrew: And what was the problem with this

Rajiv: this, so it was not, it wasn’t. It wasn’t designed it wasn’t designed for work. And there just wasn’t presence that was relevant.

The presence in discord is music. There’s game. This what game you’re playing. It’s amazing for a social, right? Like I want to know what my friends are up to and what they’re doing. It’s great for social does not translate to work at all. It’s basically useless for work, uh,

Andrew: so, so discord will tell you what your friends are listening to, but it can’t tell you what app is present on their screen or what website they’re on.

Rajiv: no, no. It tells you what game they’re playing, but that was it. That was the seed of the idea. Right. And it was also, you know, a little bit difficult to, to do some of the things we wanted to collaborative features like screen share and just working together. They kind of exist, but they’re not, they’re a little bit deeper than they should be for a collaborative tool.

Um, you know, at this time,

Andrew: That’s not really working For teams.

Rajiv: example, I don’t think they had, I don’t think they had remote control a screenshare wasn’t you couldn’t like annotate or have your cursor on somebody else’s screen. There’s just a lot that felt missing the video at this. At this stage, it’s gotten a lot better, but the video was very bad. It was very audio first.

Um, but really the, the thing that we noticed most of all is that we just didn’t tell. That’s that that was the first problem. Like we’re not

Andrew: discord

Rajiv: we schedule a meeting.

Andrew: Yes, you can talk, but discord is much it’s chat first,

Rajiv: yeah, it’s chat first. It’s it’s and the presence doesn’t really give you the signal to talk. And when you try to talk to somebody five times and they pick up zero of those times, you stop trying to talk.

Andrew: ’cause so what’s the difference with discord ear ringing. And if I don’t pick up because I’m doing something right, then you you’re saying, I wouldn’t call you back. Why wouldn’t text in and say, what’s going on?

Rajiv: well there’s, there is, um, you just don’t have the confidence there. There’s a lot of awkwardness to initiation and if you don’t have a lot of information, what’s going on with the other. Then you’re not sure whether you should call them or whether you should kind of chat them and say, Hey, I’m here. And, um, you know, what, what we found was that it just the combination of the presence and kind of what you needed to do to initiate, even though we knew each other very well and we wanted to work together in practice, we just weren’t able to connect.

We weren’t talking about.

Andrew: yeah.

it’s subtle, which is why I’m asking, because the things that keep you from doing something or encourage you to do it are so subtle. And I kind of like the comparison to discord much more so than I like the comparison between tandem and zoom, you know, to say it’s like zoom is almost, I don’t know.

It’s, that’s not really the magic that you’re aiming for. Right? It’s not about putting you and me in a situation where our video. Is lag free and where we can talk to each other. It’s more like in between those zoom calls, how do we still feel connected? You know what it is. I remember Fred Wilson, when he invested in Kik, he said, there’s something magical.

You feel like you’re in the presence of someone when you’re using the Kik chat app. And the magic was something that I messaged and others have copied. When the other person was typing, you can see on your screen, those dots, or some equivalent of it going off that says they are typing. And so you feel their presence even when they’re not typing, that’s what you’re going for.


Rajiv: Yeah. Super important. Yeah, that that’s, that was, and then you can go one further, you know, if you’re typing in slack, we share presence that you’re actually in attendance. So you see somebody typing and you see that they’re actually connected in tandem and you realize I don’t have to slack back. I can just talk in tandem.

Um, but yeah, that, that, that presence it’s subtle. It requires a lot of experimentation to figure out what’s the right thing, the right visual cue, what’s the right kind of a handshake to unlock the behavior. Um, but we, we had this sense that while discord was the closest and really. You know, it, it, it brought kind of, uh, belonging and togetherness in the gaming community.

You know, at that time it was not working for work. Uh, and so we started kind of prototyping ourselves. Like, can we, can we build something that really achieves this for.

Andrew: We talked earlier about how you knew that the last ideas weren’t right. How did you know this idea was.

Rajiv: I’d like to say that we immediately got traction, but we had the least traction with this idea. We just had the highest internal conviction because we were solving a personal problem. So we actually, the first thing we built was utterly useless. Uh, it was a walkie-talkie where you just push and you start.

And it had no presence. We had not really identified presence as a thing. Uh, and so we were just trying to remove friction. At this point, we were just working on friction, one click talk, and it was terrible. Like nobody was free. You didn’t know whether you felt like you were blind. Um, and ultimately, even though everybody was one click away, you felt like you were alone.

Cause nobody ever picked up and nobody, we didn’t like it. Nobody wanted to use it. We did this next iteration was. Let’s go asynchronous let’s build voicemails. So when you press the talk, it’s simultaneously recording, like an answering machine. This was great. In theory, in practice, nobody like none of the few people who tried it liked it because you just don’t know whether you’re leaving a message and then you think you’re leaving a message and somebody interrupts you and it’s really jarring.

And then you get these long rambling messages. Like nobody likes listening to messages. Yeah. And, and so after this. We had no signal. Like we weren’t using it. Nobody else was using it. So, you know, if it had been any of the previous ideas, I think we would’ve stopped, but we just needed to solve this problem because we started the company to work together and we didn’t feel like we were working together.

And I don’t know, I don’t know what it was. We just won. We had a ton of ideas on how to solve it. So this was not the end of our ideas. And two, we just felt so personal and it also felt like this is a problem. Nobody has solved before. Uh, So we did another iteration, a really ambitious one. We, we said, how are we going to get you to pick up the call?

We are going to build a workspace and you are going to work in that workspace. We can’t build code and spreadsheets and design. We can build a document editor. So at least if you’re writing in this document editor, you are available to talk. And so we built a full like Figma for docs, like very real time multiplayer document.

Voice and video attached. So when you’re in a document, people can see that you’re in a document, they can see what document you’re in and they can talk to you.

Andrew: Okay.

Rajiv: And that was interesting. That was the first one that was interesting. Like the documents were really janky. People didn’t want to switch, but when they did switch, there was some really delightful interactions by knowing what document you’re in.

And that was the first clue that like, oh, knowing what you’re working on can be present in a very real way. There was this moment at midnight where I was brains, I was in a brainstorm doc that my co-founder Tim had written and I saw him in it too. And we talked and we had like a little brainstorm at like 12:30 AM.

Andrew: Um,

Rajiv: And we’re like, this would never have happened without the.

Andrew: docs and like Google docs has that chat feature where you can chat with someone who’s in the doc at the same time. Your version of it allowed me to have a dock. And if someone was in it, I could voice chat with them at the.

same time.

Rajiv: Yeah. Yes. And it was very, the key was very real time. You saw their, not just their carrot, their presence, you saw their. You could follow them so you can click on them like Figma and follow them through the doc through multiple docs. And so you really felt like somebody was in the room, like there, the cursors, like their index finger.

Right. You know, what they’re paying attention to. Uh, and then it was just really obvious talk button. One-click talk. Um,

Andrew: Yeah. Google docs does hide that chat. Like if you ever, I I’ve been in a dock with someone and tried to get their attention and it just doesn’t work.

Rajiv: Doesn’t work. Yeah.

Andrew: It’s so poorly done and it’s too bad because it is a helpful feature. If we’re both working on the same thing, just let me talk. We just type in the dock itself then, which is an annoying thing

Rajiv: Hey, let’s chat.

Andrew: Right. Okay. Okay. So for you, you said, look, we built this thing, Google docs or Figma, but essentially with voice, that’s not really the magic. Do we want to create an alternative to these platforms? No, but we want something that gives you the same feeling that we just created. What if we could port that to every app that they’re already using was the next step, right.

Rajiv: Yes. Yes. We, we realized after talking with the teams that use it, that the documents themselves were not the interesting part of what we had built. Uh, and we, we knew we had to focus on either the documents or. The the real-time layer because it’s very difficult to build both. And so we found that people were using it despite the fact that they had to switch to our document editor, not because of it.

Uh, and so we, we, we took what was good, the presence, the quick real-time conversation, and we stripped out what. Not good, which was our own document editor. And we just integrated with other document editors and eventually like every app to build that sense of presence.

Andrew: Okay. I’m with you on that. And I’m guessing these are a lot of your friends you’ve then I’m guessing chatted with them and said, Hey, if we built something like this, for the apps you already using, would you want it? And they gave you the same, shut up, take my money type of reaction.

Rajiv: No, I wish they were the case, but no, this one was a little bit of a leap of faith. We, it was some, you know, some of our friends, a few people in our coworking space, cause it’s like part of the time we were working out of a really cool coworking space with a lot of international companies in San Francisco.

Um, and we, we found when we, we started kind of redesigning the next. People were a little bit hesitant. They didn’t really understand it. That was the stumbling block for the next thing. Um, and so we had a lot of conviction based on the past three iterations that this is the thing like this checks all the boxes, but we couldn’t describe it.

Um, it’s not a document editor with voice and video attached. It’s some sort of voice tool. Uh, and our first description was like a real-time communication for remote teams, which is kind of vague. And we actually went through a bunch. We realized like maybe one of the biggest risks is people don’t understand it and don’t care about it.

So we went through, uh, an exercise of testing, a lot of different promises, even outside of the things we were building and testing competitors to like slack and zoom, to try and figure out how to describe this in a way that people understood.

Andrew: Who would you ask? Who would you talk to to see if they liked it? Do you have enough friends that you could just cycle through them and give, give them enough of these ideas and

see what they respond to?

Rajiv: the promise testing exercises is really interesting and we added, you got to do it with strangers. I think, um, we would walk around to our coworking space. I’ve done this at the rock climbing gym, you know, in cafes when people order. Right after they order coffee, they take out their phone for a few minutes before they get their coffee, you can kind of interrupt them and ask them to give you feedback.

And you show them a couple, a couple of different promises and ask them like, would you click on this?

Andrew: That’s one of the things that I loved about San Francisco. This was in the bay area, right?

Rajiv: Yeah.

Andrew: it’s so good for that. You really could just walk around and talk to someone else who is either in the business or cared enough about it and had the same kind of needs as you. It really was a unique, I say wise because I’m not there anymore.

I’m assuming you still are though, from what I’ve seen. Yeah.

Rajiv: Yeah.

Andrew: I’m in Austin now. And there is a lot of tech here. It’s not the same. It’s not the same experience as walking around and seeing that everybody is on whatever the newest thing is, by the way, I should talk to you about lemon. At some point, Rajeev, you’re going to need a developer, right?

You have a big project. You have an interesting thing that you don’t have enough time for your people to work on. You just want to hire somebody. When you’re ready to hire a remote developer, I want you to go to They will talk to you, understand what you’re looking for, and then match you with a phenomenal developer.

You could start working with them quickly, and if you’re not happy, I love this. They have the best copy on their website. If you’re ever looking for inspiration, or if you ever saying, you know what our site is just too, blah, we’re missing something. Go to lemon. It will shake you and show you what else you could do and how out there you can be and still be respectable.

Anyway, I say here’s how they say it on their site. And if you’re dev turns into a devil will bestow a new Divo T on you. Should your lemon developer, miss deadlines or fail to meet expectations. We’ll match you with a new developer without shifting your project schedule or cost. Admittedly, we’ve never had to do.

But it’s our promise, just in case, what I’m saying is they stand behind the connections that they make. So if you are looking for a developer, if anyone out there is looking for one, I want you to go to not just, but This company has developers from countries all over Europe.

They find you great developers at a low price. And if you use my URL, they’ll knock that price down even further. All you have to do is go to Um, and you know what, they didn’t even ask me to talk about this. They haven’t even brought it up. It looks like now they’re also doing an outsourced CTO and on-demand CTO, which is good for some companies.

This is going to keep growing. It’s fricking guys. Alright. Um, you had conversations. What was the magic that got people to say? Uh, hi. Yes, I get this.

Rajiv: The virtual office for remote teams.

Andrew: That’s what still, I think on the top of your site right now, let me see

Rajiv: Yeah. It became a category. Yeah. And,

Andrew: office for remote and hybrid teams.

Rajiv: Yeah, the virtual office for remote teams was the, the, the, and this, by the way, as we get into premise testing, we have started Y Combinator. So one of the things that we got a lot of push and a lot of push for was just refocusing on this problem, that people didn’t understand what we’re building.

And if we can’t explain. Nothing else matters. We can’t, if you can’t explain the idea, you can’t test the idea. If people don’t understand it, they can’t tell you whether they like it or don’t like it. I think you can’t get signal and they can’t tell their friends. You can’t know whether they told their friends or not, because they didn’t understand it.

Everything breaks. So for the, in Y Combinator, like one of the maybe counterintuitive things that happens is every week, the first question is what are you. Every single week, they just start, they go around the small group, tell me what you’re building. And they, and as much as this is about building this complex startup, you know, with lots of risks and you know, lots of code and lots of different customer segments to explore it all starts with.

Can you describe what you’re building? And so after a bunch of iterations, we realized that the virtual office for remote teams. There were a couple of components of that. Remote, not distributed. People identified with remote. They didn’t identify as much with distributed. It was a little bit imprecise virtual office.

Um, initially we were afraid that people thought it was VR AR, or we were afraid. People were familiar with the previous generation of virtual offices, which is like a telephone number and a post office box, but it turned out nobody remembered the previous virtual offices. And as soon as we described it as not VR, or like, you can see who’s around to talking about.

People quickly got to the right idea. And that was, that was an unlock.

Andrew: No, my challenge, having spent some time preparing for this interview is. I’d need to try it in order to really fully understand it. And I don’t just mean try it on my own and say to our producer, Hey, can you quickly do this? And I’ve got the interview coming up in an hour. I mean, having experience with it, I feel like for slack, one of the big aha moments was when you’d get into some kind of community that you wanted to be a part of and it was on slack and you’d see how slack worked.

And you said. I think my team could use that. And you’re, you’re really super generous with the free version. I feel like a lot of people who are listening to us could just go to, use the free version. It’s up to 50 people, which I think is a mistake, um, for free, if 50 people free is, is basically the number of people who are listening.

There are 50 or under. But I feel like it’s still a big barrier to trying it. What would be really interesting is if there were communities that were using tandem, do you think that do, does this make sense where we could just log into a community of people who are all working on something together, stay in touch with each other and then say, you know what?

My team could use this.

Rajiv: No, I think that makes a lot of sense. We haven’t really been able to build that direction, you know, for focus reasons. Although we do have some communities like student groups using it and we try to support them. One thing we’ve found is that. Yeah, we have a generous, free trial. Like even if you’re above that, it’s a 30 day free trial and it does, uh, you do need to experience it with your team because the magic moments are team-based, it’s often something like seeing, uh, a colleague who’s a developer is in vs code.

There’s an unlock there where you understand presence or realizing. You can just click to talk and then pull in a third person or having a quick conversation after a meeting to debrief and like, Aw, man, that was a crazy meeting. You know, these are the magic moments and they all require you to invite your team on board.

And so that is getting your team on board and, and is one of the things that we’ve, we’ve worked with. Over the past two years, and it’s been a challenge for us, right? But once you get your team on board, you just have these magic moments. You have serendipity the other way. We’ve tried to create magic moments and, and come up with some new moments is in the call itself by focusing on what’s great with your team.

The first thing people often notice is like, It’s a celebration, you know, you can just, high-five your teammates. It’s kind of fun. It’s, it’s, it’s, uh, it’s kind of awkward in a, in a funny way, and it lets you celebrate with your teammates in a way that’s a very immediate and quick. The second thing people usually notice is that screen share.

You see other people’s cursors. You don’t have to dig for it. Remote control is just one click away. It’s really like turning your computer to another person. Or seeing their attention on your screen. And it’s, it’s just so quick and easy and those two kind of things. Yeah.

Andrew: with zoom, when I could do people, most people don’t understand that in zoom, if someone says sharing their screen, you could start to annotate or move a little cursor and point things out. And I know that people don’t know that because when I do it, they’re always. Wow. How did you do that?

It’s kind of a pain to get to, but it’s there. And if you could get to it, you really can say, this is the part I think you should change. Here’s what we should focus on. So I get that. Let’s let’s then turn our attention to growth. How did you get people to sign up? What’s been working for you.

Rajiv: That’s that’s where I don’t think I have a great answer because it’s been, it’s been organic and that, that there’s a pro and a con the pro is that, you know, teams seek out tentative and find it. And the con is that we haven’t until somewhat recently kind of learned how to affect it or how to control it or how to predict it.

Um, I think that there’s, you know, in general, we focused after, on, on activation. Uh, you know, kind of after the team gets a few people on board, uh, but a couple things that we’ve been working on and we, we brought on a head of marketing about six months ago, is that, that content piece, like building that soft landing for the website.

And so people can understand not just what is tandem now, what does tandem do now, but where are we headed? Like, what is, what, what is. You know, bridge between remote employees bridge between the office and hybrid employees that we’re trying to create. And what are the other facets of that? Like where do we sit in the industry?

Um, and that’s, that’s led to, you know, a fair, a fair bit of, you know, top of funnel as well.

Andrew: Overall, how many people or companies are using it?

Rajiv: We have about 800 companies using it and, you know, I think it’s, it’s still, still relatively early, but the thing that we are. Really really proud of is the engagement is super, super intense. Um, and one, one kind of metric that we, uh, we have tracked is how many serendipitous conversations is tandem and unlock.

And over half of the calls on tandem are under 15 minutes, which is, which is really, really cool. You see these patterns that are well outside, the one hour standard meeting slot.

Andrew: Yeah, those shorter conversations. I could see why that would be a good metric to follow like a longer conversations, a waste of time and could be done on other platforms, but a quick, Hey, what about this? What do we do here? Is it’s something that’s just not handled yet. Alright, it’s You’re an entrepreneur through and through where you want as a kid.

No, you’re more of a data guy. Feels like

Rajiv: I was data guy, but you know, my dad did startups. He’s still doing startups. I grew up around startups. I didn’t really know that much. So he, uh, he started his first startup was token technology. It was token land technology, a competitor to ethernet, uh, sold it to IBM. And then he’s in a bunch of stuff. He started a live streaming service, kind of like clubhouse 3 million active users, uh, could not figure out how to monetize kind of wound that down.

He’s done stuff, kind of like clubhouse. Yeah. In like the nineties,

Andrew: Yeah. Wow.

Rajiv: you know, and then now he’s working on a machine learning for he’s got machine learning for international. So using machine learning to help you figure out duties and taxes and make international shipping like a Shopify plugin that makes international shipping really easy.

So he’s always just done startups as a career. And I think if there’s one thing I took from it is that you, aren’t your idea. You aren’t your company. Like if you’re a founder,

Andrew: Uh, huh.

Rajiv: you are a FA you’re, you can be a founder over a long period of time.

Andrew: And don’t be so connected to this one thing, that if you have 3 million users and people know you for doing it, but you can’t make it work financially that you stick with it because you think that this is you and part of your identity. I could see that working through you. All right. The website is, and I still would love it.

If maybe there’s someone in my audience, who’s going to create a tandem chat community or something. So we can just try it out together. I’d love it. And if you will have. If you want to, if you could, I’m going to ask you guys for help, actually, please. If you’re listening to me, I am out in the backyard.

Look at how beautiful his space is. Dude. I did not

Rajiv: co-work with Andrew.

Andrew: Uh, no, I would love coworking here and people do that a lot. Um, So, yes. If anyone’s wants to spend the day in co-work it’s great. My challenge with coworking is we often don’t get enough coworking done when it’s, when you really working together.

It’s amazing. But when you just kind of distracting each other, I don’t like it. Um, no, but I want to keep recording out here. If you’ve heard all the way to the end. I’d love to know. Did you hear any random sounds here? There are all these random sounds going on in the background. Everything from some kind of, I guess they called cicadas that are making that, that sound to the wind right now.

I want to know. Did you hear any of it? Was it a distraction? Did the voice come in in an awkward way? I keep wanting to battle, test this system and see if I could keep improving it. Um, eventually I’ll record indoors, but I’m just enjoying the outdoors here. So Frick and much. So here’s my email address.

It’s Let me know what you thought or Steve, thanks so much for being on here,

Rajiv: Thanks so much, Andrew.

Who should we feature on Mixergy? Let us know who you think would make a great interviewee.