Jan-Erik Asplund
Russ Heddleston started and grew DocSend to 17,000 business customers before selling it to Dropbox (NASDAQ: DBX) in 2021 for $165M.
To learn more, we teamed up with Sandhill Markets to host a conversation with Russ to talk about DocSend's early near-death experience, the pivot to self-serve that got them growing 100% year-over-year, and the inside story of their $165M Dropbox acquisition.
Transcript below.
Adam:
Cool. Just as the first to kick the dumb questions off, self-serve versus product-led growth, is self-serve software a part of that movement or the PLG?
Russ:
That's a great point. I have not thought a whole lot about the branding around it, because yeah, what does product-led growth mean? I think as a founder, I'll get at this in the slides, it comes down to running a good business and what's your acquisition cost and then what's your LTV. Investors get very into these things. But even just practically running a company, you want to lower your acquisition cost if possible and then have a good LTV to it. So, just running a good business.
So, when I say self-serve, what I mean for DocSend was the relatively small accounts that would come in and just put their credit card into the product. So, we were happy to talk to them and you could talk to sport sales, but they were distinct and different from the big accounts where we'd be selling 100, 500 seats. It was like outbound or enterprising thing, and those people wouldn't come to us or they'd come to us relatively rarely. So, the self-serve part of the company was the part that we got that engine going and that was great.
Yeah, so for those of you who don't know, just a little history on DocSend. We started the company in 2013. We were first headquartered in San Francisco. We in the life of the company raised 14.7 million in venture funding. We sold it in March of 2021 for 165 million, which was 11X ARR at the time. We were profitable and sold to Dropbox. They were very pleased with the acquisition and we hadn't raised that much money. So, it was a good exit for everybody involved. A lot of that had to do with the go-to-market motion for DocSend is pretty similar to that Dropbox.
So, we will go into the first part of this, which is basically just a story in cash and I didn't put any numbers on here, because I just want to give a sense for the arc of stuff rather than go into specific numbers. So, I'll just fly through this, Adam, but pause me at any time. I think for anyone else who's been a founder, your cash bar is your stress level as the CEO of the company. So, it says a lot about my mental state, how low that green bar is.
Yeah. So, we were going to be this self-serve document analytics tool. We launched it at Disrupt in New York in 2014. So, it took us about a year to build out our first version of DocSend. Then shortly after that, actually, Q1 at 2015 is about a year later. We decided, "Okay, we got to add in some revenue here." So we started with a $10 a month plan just to start charging anybody something for the thing. We got a preemptive series A from August Capital, now Lottie Capital. They're also fantastic. Highly recommend them. When we weren't quite ready for it yet, we didn't have the numbers, but they had a lot of faith in us.
Adam:
Explain what you mean by preemptive there where you weren't really ready to fundraise but they caught you like, "Oh, they're going to go fundraise, let's go early."
Russ:
Yeah, so preemptive will happen if as a founder you've got a relationship with a VC and they've been following you for a while. In the case of August, now Lottie, I had pitched them for the seed round, but we were too early. They were at the time investing out of a $500 million fund, but we just kept in touch and then they just watched us execute on the product. They just started getting DocSend links from people. So, they were like, "They're going to figure it out." So they came and said, "We would like to lead the series A for you ahead of you going out to raise it." So they gave us decent terms and I shopped around. Everyone else said, "Yeah, you're not ready for your series A yet." So I said, "Okay, we'll take this from August Capital." It was great.
I'm very happy that we took that round for them. Could we have gotten a better price later? Maybe. Could we have also died? Probably. It depends on how much pushing you want for how long it's going to take you to figure things out. So, one of the things we did, because our self-serve at first wasn't growing that fast, it was growing a little bit but not quite fast enough. So, we had sold a couple 50K sales deals. So, one of our investors was like, "Do that, sell up market." I was like, "Okay." So we hired up a whole sales team. We said we're going to sell up market just into sales teams. That was a big education curve and it is working but not very well. So, the red bar here is the enterprise era, and I just distinguished this, it's not really enterprise, more like mid-market.
These are like 20, 40, 50K deals, the occasional 100K deal, but they're not a million dollar deals obviously. So, we're closing them. We're doing outbound. We're doing cold outbound. We got SDRs, AEs. We're not growing fast enough and our investors were losing patience. Then fast forward and self-serve, the blue bar is ticking along and then enterprise is also ticking along, but it finally overtakes the self-serve ARR. But at that point in time, we spent a lot of that series A just on this outbound sales and the payback period is long and it's getting more competitive. Our competition, the biggest one was Highspot in sales enablement is just beating us because they were aimed at the economic buyer for an organization like sales enablement.
DocSend was originally built for the end user. So, we tacked all the stuff to be relevant to the economic buyer, but the bigger the company, the more different it becomes selling to the person who's in charge of the budget versus the end user. The buyer often doesn't care about the end user experience. So, we've run into situations where the people at the company who are using DocSend were like, "We love it. It's the best tool, it's great." Then the person in charge would be like, "No, I'm buying this other tool for it because it has more of what I want for admin dashboards or whatever it happens to be." So anyway, at this point in time, it's looking not great for DocSend. Investors kindly tell me that it's probably not going to work, said, "Good job, you tried hard, it's not your fault. Return money. We'll back on the next one. You did what you could."
We had four million ARR at that point in time and I was not ready to throw in the towel here. So, I went off and interviewed hundreds of our users, because I was like, "There's something here. We just need to figure out what it is." So based on those interviews, we changed up our pricing, packaging, and positioning. So, a rare instance of an MBA being helpful for the business stuff.
Our VP of Strategy, Brian Gaertner on this, we called it Project Couch Change, because we were trying to just make 50K additional ARR and we got tens of millions out of this little change. So, I think a lot about message-market fit and how important that is, especially when you're in product-led growth. So, after we make these changes, and by the way, we had that $10 a month plan. We then put in $150 a month plan. Instead of selling enterprise, we just made it like you can get all these features just with a credit card. You don't have to talk to sales. So, we put that in as our standard plan. It was fascinating to see how many people were just paying for the expensive security plan.
We don't even know what was supposed to be there at first, but we saw it's working a little bit at this point. So, I'm like, "Yeah." So then based on that, we raised five million from a friend of mine, Kyle Lewis, who's now at Cloud Capital, highly recommend them. He's at DCM at the time. We had a term sheet actually for 20 million and I turned it down for 5 million because I was more aligned with Kyle on what we wanted to do with our company. The 20 million would've come with the expectations that we go try to sell market again. I just don't think that's the right thing for us right now. So, it was confusing for our team because we raised 8 million before, but this was an up round. It was a great up round, 5 million, but it wasn't that dilutive.
We gave Kyle a board seat, which was great to have him involved. Then based on that, we actually fired less than half the company that was working on enterprise, not because we needed to from a cash perspective, but because it was just too confusing to be building the features for the enterprise customers and then following this thread around, "Oh, my God. We're having a lot of people sign in, ones and twos and threes and fours who are just loving a product and swiping their credit card." So we should just do that.
So, that's what I told our company that we could do either, but we don't focus and so we're just going to do the self-service stuff. One of our engineers was really nice about it, because I was very nervous announcing this to the company. He's like, "So are we just doing the thing that we're good at?" I was like, "Yes, that's what we're doing. We're just doing the thing that we're good at."
Wait, wait, wait, so we're going to do the thing we're good at, right? Yeah, yeah, let's do that.
Yeah, exactly. It's a really good way of just making it very simple. It's not personal. We could do enterprise if that was all that was available to us, but hey. It's working. You just show the numbers and we're like, "Eh!" So you can see the blue bar then just take away, we didn't fire the enterprise customers, which is worth noting. We didn't hire back account managers where the account managers were okay knowing that we weren't going to build the features to keep their accounts. So, it was much more of a situation of just keep these customers as long as you can, but we're not going to build their features, but just be nice. Just be a nice helpful human so they stay with us longer than they probably should. That worked out great. That revenue was stable.
It actually started to expand over time and then the self-serve just started to take off primarily from the advanced plan that we had and more security. So, we just kept burned low. We never touched that much of the five million. Then you fast forward to Q1 of 2021, we got an inbound offer from a public company. I didn't turn to Dropbox and kept in touch with them. I asked them. They made us an offer and asked our board and they were like, "You almost died a while ago. This seems great. You should take it." We were eight years into the company and we took the offer. It was a reasonable deal, 11X ARR. Dropbox was happy on our side. It's very hard. How many people do you know have 165 million? We've been at it for eight years.
So, we weren't sure what would happen. Interestingly enough, we ran this experiment selling it to Dropbox, the standalone, just looking at the standalone revenue, it just continued to be super strong. It's not demand-gen oriented. The number one channel's word of mouth, which is fascinating because people like it. You could say, "Oh, my god. DocSend is very niche." It is, but it's also people send a lot of documents. If they're important documents and you want control, Adobe is pretty clunky and Microsoft's not great, so what do you do? Go to a really expensive data room provider? That's not great either.
So, we just became synonymous with this, sending secure documents and maintaining control and fundraising was a great use case for that obviously. A bunch of people tried to displace us from that, but DocSend works pretty well for it. It's hard to 10X better. So, yeah, it just was wild to see the growth continue.
Adam:
When you look back on this, so that decision to take 5, instead of 20, feels like a big moment. If you had taken the 20, would that have made the exit less palatable when you look back on that moment where things are starting to work?
Russ:
I don't think it would've closed doors to raise more money. It would've been at a higher valuation than the five. So, we took a haircut of valuation to raise less money. Would that investor have been okay selling? Maybe not. So, it is possible it would've closed the door. It depends a lot on the investor that you're adding to your board. It depends on how preferred they're getting, preferred rights. I think for me, the five million made more sense just because I had alignment with the investor on what our plan was going to be. So, in our situation, it wouldn't have felt very good to have taken the larger amount of money but then not have wanted to do what they thought we should do.
So, with that in mind, it was definitely the right decision for us. You can also see we didn't burn that much of it. So, if we had raised 20, it wouldn't have changed my thinking around what to focus on. I do think that's a common problem with companies if you raise too much money. It had nothing to do with the money that we got rid of the enterprise side of our company because it's just confusing from a decision-making standpoint and it just takes overhead to have multiple go-to-market motions. So, that was just better for us, but we got to accidentally profitable in 2021, because we were just growing our engineering team. That was where we were spending the incredible money. What I told our team is we're growing faster than we can hire good people and onboard them.
That's the interesting thing about this go-to-market motion is that it doesn't require a lot of demand-gen spend. It doesn't require any outcome sales. So, there's not a lot of go-to-market costs. A lot of your go-to-market costs are relatively fixed. We can get into what some of those strategies were on marketing and go-to-market, but we just primarily needed to keep making our product better, which was engineering hiring. My co-founders and I have all been on quite a few engineering teams over the years and it just gets really confusing if you hire too many people too fast.
Yeah, we had a cool conversation about hyperscaling and people who had been involved with teams that had gone from 50 to 400 people. It just seems just so crazy to stay on top of everything. The other question I had just about this graph, because I think it's the right time to get into this type of thing, is on the market sizing perspective and the pressure to push up market, from the VC conversation, was it a market sizing driven thing where it was like you should go up market because the contracts are bigger and we can actually size that?
Whereas you look at the long tail where ultimately you've had a lot of success more in that long tail of the business. It sounds way more flimsy to be like tons of people are going to spend 50 bucks a month or 150 bucks a month compared to the enterprise version of we're going to close these 10 and that's going to add a million ARR. Was market sizing a part of the conversation or was it something different?
Not so much market sizing as it is just like the other investor wanted to run the playbook of get big contracts and stack on top of each other.
Adam:
So more of a playbook thing than a...
Russ:
Correct, yeah. Although in 2018 there wasn't as much belief in product-led growth. When DocSend started in 2013, there was nothing around it. Now you've seen a bunch more examples pop up.
People have been really into it and it feels like it's died again as a fad. There's no more freemium and we actually killed the freemium in DocSend as well.
From a timeline perspective, it is funny, because now in the current context, product-led growth has definitely had this big moment and it's waned for sure, but this makes it feel like it's all right now. But yeah, looking back to 2014, saying product-led growth in a VC meeting in 2014 wouldn't have been like, "Oh yeah, cool. That's the thing right now." Towards 2019 to 2021, yeah. Got it.
To your point about TAM though, we did some rough math around the fundraising market and came out to about 65 million a year is what we thought the total TAM was just if we got everyone who was fundraising to use DocSend. I'm looking at the mix of we'd be paying for, which plans and their sizes. So, 65 million a year is very small, a total TAM, but with DocSend, it was only ever at most a third fundraising revenue because we got a lot of sales revenue. Even when we stopped trying to get more market sales revenue, we still got a lot of small sales teams into DocSend.
There were a lot of internal use cases, a lot of CFOs used it, investor relations. There's just a long tail of banking financial services, a lot of people want to send password protected PDFs, stuff like that. So, there was a long tail of applications for DocSend where people were willing to pay a premium. So, the personal plan is still 10 bucks a month for DocSend, but the advanced plan at 150 bucks a month is most of the revenue and that's where most people went. We also did smart things be on reverse trial. So, you start with advanced and then it's done. You have to go back to personal. That's painful. You feel that.
Yeah. It's a good way to price discriminate because you really only go to personal if you just really can't afford advanced. But for a lot of the people, advances, that's what they want, that's what they need. They want the dynamic watermarking, they need the forward tracking. They want it to be secure and that was just a good price point to have. So, the TAM for that was just bigger I think than anyone thought. Even so, DocSend in this graph, we're not big enough to be an independent public company. So, TBD. What we would've done, we would've needed to go take on the full data room market, for example, a couple billion a year in spend, but a lot of that is in big banks.
Yeah, it's very intense. Yeah, it's not like a software. I don't know. There's cultural aspects of what... We talked about that too and we will continue into the presentation, but we talked about it a lot at OnePager where there's also... You guys did a great job with this in the fundraising in the VC space specifically of sending a DocSend was allowed. Using certain tools, because there's a certain amount of signaling that goes on when you're sharing whether it's IR stuff or its fundraising materials or whatever, that it's not necessarily culturally...
I don't know if cool is the right word, but it's not like the thing to do and that breaks into the software just being better also. A lot of the data room tools are ugly and clunky and all this stuff, but it's what you use. It's what everyone has been using and to bring in a new tool can be more of a trust loss than the efficiency gain warrants or whatever.
We'll get into one couple of the strategies that we had with marketing for DocSend, but it was definitely a social movement type of thing where when we were first starting DocSend, any VC I talked to hated getting DocSend links. They were like, "Just send me the PDF because I can store it on my computer or forward it to people I don't want you to know about me forwarding to." They were so annoying that DocSend was around. I remember going to a holiday party and it's talking to these teams. What do you do? Oh, I started this company DocSend. He's like, "DocSend, I hate DocSend." Someone is like, "Do you say DocSend? I also hate DocSend."
So I would joke that the reason I started it was just to annoy as many VCs as possible in parallel, but I think over time what we wanted was a social norm around I'm sending you important information. I have the right to understand. Did you read it or not? It's really nice for version control, it has all these other positive benefits to it. So, it eventually did come around to being in an accepted social norm because the workflow of fundraising is really hard for founders anyway. DocSend is a small saving grace of the whole thing. So VCs have just learned to put up with it.
We always joke that we'd be an overnight success after a decade, and so the hockey stick is about a decade and we sold after eight years. So, it takes a while. As it relates to inbound or product-led growth or small customer type of acquisition stuff, if you're going outbound and up market, you can create an Excel model that says, "I will have these salespeople, I'll pay in this much. They will do this many outbound touches. They will get to this many qualified leads and then this will funnel down to these contracts."
So you can say, "Oh, we'll hire more salespeople. We will get more revenue. We're going to ramp up our go-to market." So you can plan how much you're growing based on how much you're investing. Product-led growth is frustratingly, you just don't know what's going to happen, because if your growth is coming from word of mouth, you're like, "Okay, cool. People love the product. It's spreading," but how do we do that?
The funny thing in talking to other investors at the time when we were fundraising, they were like, "How do you know where it's going to be in a month or two or how it's going to keep continuing?" I don't. I don't know if it's going to keep continuing. I was like, "I can't tell you where it's going to be a year from now." That was very unsatisfying. The way we did our growth cycle though, and this is probably the right answer to this, is just identify areas in your existing user base of high need and then build them.
I was at Meta before Chris Cox was still the head of product. He was talking the early days saying that they would just watch where users in Meta would go and where they would be frustrated. They would just go check every single profile for new information. So, it's like, "Yeah, we should just launch a new speech. You don't have to go check every single profile."
Follow these people every single time. Yeah.
So similarly in DocSend, we would observe user behavior and people would do these weird monkey things to send one link to a collection of links. We're like, "Okay, I see what they're trying to do here. They're trying to send a folder. They're trying to send a landing page. They're trying to send a deal space." Call it what you will. So, we identify things like that and then we'd go suss out what it means and then try to figure out how you would use it and how common is this need in the user base and we build it. So, then the gross cycle is really just attracting new type of funnel.
So, we call it our spaces product, but it's a data room, it's a deal room, it's an investor relations portal, it's all of these things. We would have all the landing pages of marketing site to tell you that this thing is for you and then it would convert and then we would retain. So, you just make the product sticky. You expand what you can do. You make sure you're really good at what you can do and solve those needs, and then you just keep iterating from there. So, it was really fun to just follow that thread, although it is hard to predict exactly how much growth you're going to get out of any individual feature that we launch.
Fun when it works, right, because when it doesn't work, I'm sure, and maybe you have some stories of... We have this experience at OnePager or plenty of times where you try to suss out what you think people need or what they want or whatever. You spend the time identifying, you spend the time building it. Maybe it sticks, maybe it doesn't. Were there times where you had false starts? Because that's, I think, the part where people get scared about the product-led growth is because there's that leap.
You identify, you spend the money to build it, and then you throw it out into the market and then if it works, people pick it up. It's great. All right, that one's working. Let's go do a new one. But if it doesn't work, then unlike a sales hire where if that sales hire isn't performing, get a new sales hire in there. They should be pretty mathematical about the approach. Were there times when there were false starts in the cycle or was it generally a pretty high hit rate on the identified to build to usage?
It was a high hit rate for the small users, the one, two, three, the end user. It is not a great hit rate on the economic buyer. So, when we were building out deeper integrations or spent so much time on our Salesforce integration, people wanted AI driven content recommendations in the system, which just didn't make any sense to me. I'm like, "You don't even know what you're asking for." We built it and then it was just weird. I'm like, "This doesn't make any sense to me." So there were features we built that were like, "Oh, my God." Then they failed to convert the sales leads that we were hoping for. So, it's possible we weren't as good at sales as other companies. Maybe we could have been better at that. It's still not a guarantee even when you're in outbound sales, if you build a feature that it'll help convert, because sometimes people I think say... It just isn't nice to be like, "Oh, if you had this thing, then I would go." It's no guarantee.
Adam:
It's the classic YC or just general startup mantra of talk to your customers, but I think that there's a difference between understanding your customers, which I think in the product-led, when you actually see what they're doing, they're not lying about where they're clicking or about how the actions they're taking. Whereas interviews or asking an economic buyer, someone who's got a lot of power about like, "Hey, do you want this thing if you had this thing?" That's so much more variable, but it's an interesting balance of we dealt with that so much and it's more of just the challenge of startups in general. If all you had to do is ask someone what they needed and build it, startups would be really easy.
Russ:
Well, I get asked a lot about what you're tracking as well. So, this is a vanity metric of links per quarter where every time you send a DocSend deck, it's a unique link. So, if you're creating new links, then that's a unit of value. So, you can just see this go up over time, which is cool. That's engagements going up with the product. This is how we're tracking data rooms. We had a bit of a false start, took a while, but you can start to see this incremental engagement with the product. Then when we added e-signatures, that also did really, really well. Then we just made it easier to get documents into the system. That was just a pain point of uploading documents to DocSend.
This is very exciting to see these things where it took a while to get adoption of these new products even when we built them in there with the right thing to build. So, there's also an element of patience in this that is hard to stick with. So, yeah, in terms of where the users came from, the direct was always our number one biggest channel. People just type in docsend.com and they pay us. It's pretty wild. Some of the anecdotal ones would be the CISO at a public company is complaining to a friend in VC about sending out specs for things that are sensitive and they're worried about them linking.
You must deal with sensitive documents. What do you do? The person's like, "Are you kidding? Are you not using DocSend? Have you been linking to a rock?" So we get this inbound request for a big deal at a public company and it's just funny that people do talk about stuff. If you do something well, then generally people will talk to others about it.
Adam:
When you were pitching, way back when pitching DocSend, were you aware of and harping on the inherent virality of a sharing tool? It is like a sharing tool, right? It's built to be shared with other people. Was that a part the pitch or was it a happy coincidence? How aware were you of that virality?
Russ:
We leaned on it heavily. But what was interesting is when we first launched it, it was very linear growth. It was not viral. People who received DocSend links often don't need to send DocSend links. So, think about a sales context. If you're buying software for your company and some salesperson is sending you collateral, you're not necessarily sending out a bunch of your own collateral. Similarly, VCs don't send out that many debts with the exception of they do their own fundraising, but it's like every four years. So, it's just different. So, we did get a number of signups from viewing a document in DocSend, but it just wasn't the number one. So, it happened, but it was also very delayed. It was not immediate.
That's interesting, because a really tight and exciting viral loop would be the social media version of you get sent a link. You need to sign up in order to view the link and you go from never having heard about the thing ever to hearing about it to signing up to creating your own link and sending it out relatively quickly would be the great version of it or it'll be like the social media version of it. Whereas in the business use case, it was more like you see a link, you're aware of it, and then maybe the problem arises months later. You're like, "Oh, I need to do this also." So it's more of word of mouth, the better being, the better word than virality maybe.
Yeah, it was definitely word of mouth is the number one source. The other thing about DocSend is we tried to make it pretty low key document sending. We tried to not be overt. It's white label, so you can just put your own brand in there. We just tried to be low key, so we weren't trying to hit the viewer over the head. Our thought was like, "We need to be really useful for the sender and we need to be easy for the viewer to get in and look at the material." It's much more about the sender getting their job done, whatever that is, than it is about DocSend, putting ourselves in the middle and being like, "Look at me, sign up for this product."
We did run experiments on that, which I'll show, but it didn't work that well. Yeah, word of mouth is always the biggest source of new traffic. Then team invites was always a big thing too, just spreading within existing teams, the instances of us going from a few users to over a thousand users at a company. It's not quite like Slack, but it still has some land and expand properties.
Adam:
One last question on this graph, referral code, we spun our wheels on referral stuff. We did so much. We wanted the referral to work really bad. We tried stuff. Did you spend a lot of time there or was that ever something you put a decent amount of time into or was it just something you had because it's a way to complete the channel strategy or what was your experience in that yellow sliver?
We tried so many things and almost everything didn't work. The revenue graph, it looks like this perfect hockey stick. Then the joke was like, yeah, when people look from the outside, they're going to be like, "Oh, they just got lucky." On the inside, we're perfectly spinning our wheels, trying all this different stuff, trying to go a million miles a minute. With the referral codes, we did try a lot of flavors of that. The flavor that worked the best was when we paired research reports on fundraising along with discount codes that would be given to a VC with a special branded portal for that VC.
So, their portfolio companies could get in and get a discount through their VC and they would also be able to read the research reports that we had on fundraising. So, that finally worked well, but it took a lot of effort to get there.
It's freaking referral. I think Dropbox is really the famous example, right? Hotmail is the other startup lore example where the referral, it's just this magic. You do this one thing and it turns into this hockey stick because the referral program worked so well and we tried so much. It had the same exact experience it sounds like where even the referral version that worked for us was pretty manual. It was a targeted partner based approach where they get a code. Since they have the code, they're more excited to share it and all that stuff. Compared to what the Dropbox one was if you get to a certain amount, if you want more storage, refer a friend and you'll get more storage thrown at you. It's like this famous case study of like, "Oh, it works super well."
Yeah, I interned in Dropbox in 2010. I was in business school and they were less than 20 people at the company. It was going great that summer and they created this currency, your space, so more and more space. So, you'd refer and they got the mechanics of that. They just nailed it. It was amazing how fast that worked for them. Hotmail was more of the classic viral loop. A recipient's email is also a sender email that's all this bidirectional. So, it's rare that you can get these things to work perfectly. One of the things though about the self-serve or targeting smaller ACV customers is that the evergreen channels, whatever you can find, you got to go find them.
Sometimes there's a one hit wonder, but with DocSend, we never had one cool trick. It was a bunch of things that worked over time. You can see that finally compounded along with a great product that people generally liked. No go-to-market strategy is going to compensate for a mediocre product, but you can have a great product that fails to have a go-to-market strategy. So, no one knows about it.
Adam:
It can be a trap, right? The answer and the things you learned is that you're going to stack a bunch of strategies on top of each other that grow your inbound rather than hoping and swinging for some one thing that works super well is a tough way to go.
Russ:
Yeah, it varies for every company. It's so interesting. The companies I've worked with are invested in, sometimes they find one big trick. Other times, it's a bit slow and incremental and built over time. One concept that worked out very well for DocSend, though we pivoted back to the self-serve stuff, when we changed the pricing, packaging, positioning was that we went from being sales enablement because when we said sales enablement on our website, people would ask support. Are you the services, blah, blah, blah? Then we'd be like, "Yeah," and then they just like certain angles. So, we were like, "Actually, we are a horizontal product that we market vertically," because that's just what DocSend was and still is. It's generally pretty useful in a lot of situations.
So, we on our marketing site actually talked about the utility of the product for different use cases and that caused our conversion to go up a lot. It helped with SEO too. So, there's a big game we got by just talking about the product and the context of the people who are buying it to give them more faith in the tool, doing what they want the tools do. So, it's definitely a strategy I think more startups can employ.
The downside of it though is if you're selling into sales enablement for example and the salesperson is going to buy $1 million of your software and they go to your website and you say you're for all these things, they're going to be like, "Well, this can't be for me. You're good at all these other things. You can't possibly be great for me in sales enablement." So there's just an inherent handicap you have with this strategy versus the vertical SaaS tool that says we only sell to you and that's our specialty. So, there are pros and cons to this, but it worked great for DocSend.
Adam:
Do you feel like that's a part of the benefit of the price point? Because you're describing the big enterprise contract, you probably have to be verticalized to really take $1 million contract all the way through. Whereas if you're hanging out in that 150 bucks a month range, you can build enough trust in each of them to get to that price point. Is that the lesson more broadly?
Russ:
I mean for DocSend, we built a lot of the enterprise features. So, before I left, there's at least one customer with over a thousand seats on DocSend, which is a lot of seats on the product. So, it scaled up pretty well. I think when we think about the trade-offs, it works well in that context when it's land and expand, because if someone comes in and says, "Oh, I like this other tool more," the inertia from the internal teams already using it, they can't get away from the "Oh, it just does these things really well." So then the other thing needs to displace you. So, we would usually win in land and expand situations.
The enterprise revenue hung out and it actually started to grow, but we would not do well if we were at a cold start coming in from the top versus another tool that is a specialist.
Same with data rooms though. For a bank, if it's DocSend versus Intralinks, Intralinks just does this all day every day. That's all they do. So, whereas Interlinks is definitely an older, slower company in some regards. So, DocSend had started to take away some of the bottom end of their market share, but yeah, this is both good and bad.
I called it a transactional price point.
150 bucks a month, $1,800 a year. If it's some other data provider that's $15,000 a year, it looks like a steal. So, people were like, "Oh, my God. This is so much cheaper." The way we thought about that was we're not spending money on sales and marketing for this, so let's just make it pretty transactional as a price point and have people try it out. So, that worked well in a lot of situations.
So yeah, a lot of customers like $1,000 ACV, so you couldn't afford to spend money, do $1,000 ACV, right? That's a pretty low price point. It's not consumer. If you got an acquisition cost of $20,000 from an enterprise account, that's just not going to work. So, you were asking about some of the things that we tried that did not work, and this is a very non-exhaustive list of a few of the things. So, we already talked about selling at market too early, just having multiple go-to-market motions, really hard. Having found a market fit, selling, I was like, "I'm not in sales historically. My co-founders weren't in sales historically."
We tried to get smart on that. I talked to VC whose advice was like it takes about two years to learn enterprise sales if you're a technical founder. That's about the amount of time I spent at it, and by the time I got good enough at it, we were not doing that great from a money perspective. I was like, "This other thing's better. We're going to do the other thing." So that was tough, but I do feel like I got this merit badge for being a VP of sales, just learning a lot about it, which was very fun in retrospect. So, building for the end user versus the economic buyer, big distinction as well in terms of where you're getting better and where you're putting your chips and it's hard to do both.
So, I think that that was a mistake. Maybe just attempted to do both for a while. Then yeah, if you're selling at market and your competition raises a lot of money, you can just get priced out of the market. They can just beat you with putting more effort behind sales and just blanketing the market with information, conferences. So, that was tough for us. We'd only raised $10 million and our competitors have raised hundreds of millions of dollars. So, that's also bad because they can say, "Hey, look, this other startup, they haven't raised that much. If you go with them, they might go out of business. We're going to be around, we've raised more money."
No one gets fired for buying IBM. So, that's just another thing you have to overcome when you're trying to sell into bigger deals.
So things like outbound sales, sponsoring events. We talked about virality in the document viewer, we thought it'd be viral. It was, but it didn't matter what we did in the document viewer. We tried all of these things.
Especially if you're trying to be light about it, if you're trying to be respectful about it, there's not a ton that you can do.
But even when we were more aggressive, it didn't work. We were more aggressive. We tried it then. We didn't do anything and it made people mad. So, we're like, "Okay." We tried to do partnerships or integrations and none of that moved the needle for us at all. Demand gen didn't work at all for us. Any short form content didn't work at all for us. So, like listicles, BuzzFeed things, anything for us. So, we tried them all though. So, the things that went right seem obvious in retrospect, but we made it really easy for one person to sign up and get through the whole funnel and understand what DocSend did for them and how it worked. So, the single player mode, great.
It was easy to come in and be like, "I get it. I send a document. I see what happens to it. I can control what happens to it. This makes sense to me." We found really good fit with the niche of fundraising, which you were trying to do as well, but that was an underserved market. Nothing general worked for that very specific thing. So, it was easier for us to be just outstanding above the other options that were there. The word of mouth spread within founders and investors was great for the product, but just even between use cases, just building a great product has these knock on effects.
Where if you build for the end user and the end user really likes you, the people generally talk about software they really like. So, that worked out well for us. Having a flexible product was great for not limiting our TAM over time. So, we were only at 15 million ARR and we sold. So, there's still no fear of that, but I'm glad that we built it in a general purpose way that certainly made Dropbox a lot more interested in the product, because the potential for it is much, much larger. We did get really good results from research reports and content marketing around fundraising research.
So yeah, those are great. Those are still great. They're super informative. So, what we would tell our team is we're not going to verticalize the product, but we will verticalize our go-to market. So, vertical marketing. So, we had a team that just did these research reports for fundraising and they were great. People loved it, but that didn't matter for sales or the other use cases. That was fine. The research reports were super informative. They gave us all this press and earned media. Even if you didn't use DocSend, they were so helpful for understanding what you should do in this really wonky fundraising process.
Then incubators and VCs, we already talked about that as a good channel for us, and that was very specific to us. But I think for any company figuring it out, which even very small channel you have a hope of getting into is anything evergreen is going to be huge. Anything where you put in the effort once and it spits off awareness of your product over time is huge. I won't go through too many of these things. 90% off to the startups, that's a little bit more normal now. Help center, I really think if you're doing product led or anything with a small ACV, building out your help center and investing in support is huge.
Our support team always sold more product than our sales team just because people had a few questions here and there. They'd talk, they'd convert. They'd set the meeting with sales. By the time when sales showed up, they're like, "Yup, I already bought it. Done." So that support was great. Increase your pricing, so we increased our pricing a lot. I feel like that's a little over talked, but for us, it worked great too. Focus your scope, just have high urgency in what you're doing. It was very smart for us to expand into data rooms and e-signature.
Then the last one is just the message-market fit. Even if you've got a great product, play around with how you talk about your product because that's going to matter a lot to your end customer. They're not going to bother giving you a try if the language isn't something that speaks to them and it's compelling for them to actually put in the effort to go try.
Adam:
Could we stay there for a second? It's super cool to see. Maybe I've heard about this before, but I don't think I've seen it articulated that well or as well as you put it, of the support channel being a sales channel. I think I'm understanding that correctly, where investing in really, really high quality support not only keeps your customers happy and retention and all that stuff, but actually can be a net revenue driver from people signing up or upgrading and/or the word of mouth cycle.
Russ:
100%. It's one of those things where it's very clearly high ROI for us, but measuring it specifically is tricky. But our support team would have live chat anywhere on the site and anyone while trialing, we don't know how much you might pay us, so we don't know should we invest in our timing you or not. But invariably, it was just always worth it to invest the time, answer your questions, make sure you understand the product, because that helped us have learnings on how to make the product easier to use. We also had learnings for us and how to talk about the product back to the marketing site. Then yeah, people got their questions answered and they just bought the product. So, the sales team, when we looked at the support touching something to closing, the support team just had amazing stats around.
The store wasn't trying to sell. They were more like, "Do I need to say yes or no?" Just trying to help them.
Adam:
I want to ask you about on the pricing side, because I feel like we went through this exercise again with the OnePager stuff and how much science on it when you were putting this all together and any false starts? Because in our experience, again with this, I don't know. It can be hard. You don't get unlimited tries at pricing. You can really screw it up and you don't have a ton of data. You don't want to leave money on the table. You don't want to piss people off by going too high. What was the process on deciding on 45, not 49? How did that go down?
Russ:
It's so anecdotal and I thought about hiring a fancy firm and then I was like, "I don't think that's a good idea." Also, we don't have enough time and they don't have much money. So, me and Brian, yeah, it turns out if you ask users how much they pay for you, they lie.
I read some good blog posts about how you can structure your questions to try to get in the right ballpark with people, but it's easier if you're selling to enterprise than it is if you're selling to the end user, because there's a giant difference in willingness to pay for the user base that DocSend had. So, the $45 a month price point was just because that was the average price our sales team was getting negotiated down to. We'd start off at $90 or $100 and get negotiated down to $45. So, I was like, "Fine, just make the sale." That was what we're selling sales, just make that $45 the average that we sell when we get negotiated down.
So, maybe that's the market price for that. Also lined up well is cheap relative to other sales specific tools, but expensive relative to a box or box type of product. Then the personal $10 a month, we had no idea or I don't know. I think the internet should have a pretty cheap send and track or PDF function. That just seems like a cheap thing the internet should provide. So, we're like great, personal send track of PDF and then we have all these calls with people that we couldn't quite place with investor relations or financial services or investment banking and they were super price insensitive. It looked wonky to have it just be 10 and then 45 and enterprise.
So, we decided to just put up what we called at first a finance plan for 150. The thinking was like, "Well, it'll make 45 seem cheap." So that's why we have 45 highlighted because that's what we thought we were going to sell. Turns out the finance, which we then renamed advanced, because that fit better. It was all over the place in terms of who was buying it. That actually started selling really well. People wanted to pay more. So, that's where we focused our time was just building out the advanced feature sets. It's dynamic watermarking. It's the data room functionality. It's the one click NDA or it's the NDA management. It's a whole bunch of these interesting workflows that people really want and didn't live anywhere else. So, we put those all together in advanced and that worked super well.
Adam:
Pricing, man, it's so big too. It's such a huge part of what the model looks like of when you're actually doing how many people are you selling to, how much are they buying it for?
Russ:
Yeah, yeah. Exactly. Then in practice, who knows? We did a whole survey thing, whatever. You don't have unlimited goodwill with people too. You can't just throw it around or experimenting is a great place to have people be super pissed at you. Oh, one person got an email about this price, other people got email about this price. They're like, "What are you doing to me? Why is my price higher than..." You shot and it worked.
We tried to do variable pricing too, which totally didn't work. We tried to have it be volume based, based on how many visits you were creating, or based on how many documents you're putting in. None of that works. One wonky thing to notice here is that the advanced plan has three users included, so it's almost the same price as the standard.
The reason we did that was it was a lot of single seats in high security use cases and they were sharing their login with other people on their team and that was just ruining the whole point of security because you're like, "Okay, someone's linked out. Who shared it? I don't know. We're all sharing a login." So then we would get in trouble for that. It's not our fault because you're sharing a login. So, we just had to be three-seat minimum so that wouldn't be as much of a problem. So, people would expand a lot beyond that, but it was a lot of money actually at that price point where people only used one or two seats, but it was still worth it for them to pay that amount of money and they're very happy to do that.
Adam:
Are you happy to just have it figured out and do you think about pricing at all anymore? Usage is really coming up. Do you think about pricing still or is it just like this works, let's not touch it?
Russ:
Well, with DocSend, when it started to work, we were like, "Okay, let's leave it for a while. We'll come back to it later." We ended up changing the messaging a lot around it. We didn't end up changing the price points. When we did this one, we also got rid of the freemium or the free tier. So, we killed the free tier and it is still there. It's still hidden. If you just don't pay for DocSend after the trial or if your credit card fails, we don't just delete everything in your account. That would be pretty horrific. So, you go back to this hidden free tier and then it's handicapped in ways. So, then you have to put in your payment information to get back and get the rest of the functionality.
So, we didn't revisit this. We thought we would in the future, but it was also performing so well that there were just other things we wanted to work on. It was just adding more functionality to the advanced plan. So, it was also transactional price point. So, could we have charged more? Then maybe our churn would've gone up. So, it's really hard to say.
I did hundreds of interviews with users and really did a lot of research. It's still a shot in the dark, but at least it's an informed shot.
Find evergreen awareness generators, even if they're small, just things that spit off awareness for you that keep making people aware of your product over long periods of time and then be the best at something narrow, especially when you're trying to do anything that's self-serve or product-led or inbound or whatever you want to call it. You're not going to have word of mouth unless people love your product. So, be really, really great at something narrow and then just expand from there.
So anything I'm working on with Tim Su on Distill is... I think you did a great job at beginning, Adam, talking about it, which is the problem of hey, ahead of a call, I need to go research a person. It's the situation. The president's got the person following them around, telling them about who you're meeting with.
Adam:
Yeah, I'm thinking of the Veep. I don't know if you've ever watched that show, but there's the bag guy.
Russ:
I haven't, but yeah, I know what you're talking about. So, yeah, that's it. We're just trying to create these great profiles on people. So, it's just in beta right now, so you can sign up for the beta to distill.fyi. It's really fun just to play around with LLMs, play around with the different models, try to do entity disambiguation. Is this the right Adam? Is this the wrong Adam? Pulling everything out of these sources and then taking this fuzzy information and trying to make a story out of it. It's just something that couldn't be done before. So, it's really fun. We just started in February, so it's very early days for us, but we're also hiring mostly engineers.
Adam:
Yeah, so you're full time on this now?
Russ:
Yup. We wrapped things up with Dropbox, parted in great terms, and took some time off. It's just fun to build at the early stage. So, with all the LLM stuff that's new, this is a great time to go try something different. This is a little bit more similar to what I used to do when I was at Meta actually around just identity online. I worked on the pages team for any business identity on the internet. So, this is a little similar to that, which has been fun to get back to.
Adam:
Yeah, very cool. I'm excited to get into it and to try it out. Really in any space, it's similar, you're going down a similar path where knowing who you're going to meet with, knowing who you're going to talk to, researching people. Just off the top of my head, I can think of so many different use cases.
Russ:
It's very general. Yeah. This is very general where we want to have it be a horizontal product and market vertically, because you could use it for fundraising, you could use it for a number of things. So, getting the messaging right there, but also the differentiation of the product will be different by use case. But underneath it all, as you pointed out, is a very similar workflow of I need to learn what there is to know about this person.
What are their interests? What are other GitHub projects? It depends on what you find. Some people are very private, there's not a lot to know, but at least this still will tell you that they're very private and there isn't that much to know. So, don't stress about it. In other instances, there's a lot to know and you can find some really interesting nuggets either for icebreakers or to just show up prepared and have the other person feel very flattered that you know so much about them.
