Welcome to Episode 213! Tom Tunguz is General Partner @ Redpoint Ventures, the venture fund with a portfolio including the likes of Stripe, Netflix, Zuora, Hashicorp and Juniper Networks just to name a few. As for Tom, he joined Redpoint in 2008 and has since led investments in Kustomer, Looker, Expensify, and Gremlin all prior guests on the show, I hasten to add. He is also the co-author of Winning with Data: Exploring the Cultural Changes Big Data Brings to Business. Tom has also been named on the Forbes Midas Brink list. Before joining Redpoint, Tomasz was the product manager for Google’s AdSense social-media products and AdSense internationalization.
In Today’s Episode We Discuss:
* How Tom made his way from creating software with his father in Brazil to being GP and forefront figure in the SaaS investment community as a GP at Redpoint today?
* Annual contracts: To what extent do annual contracts dominate today? How does this differ when comparing enterprise to SMB? Why does Tom think in the early days one should be wary of signing too many multi-year contracts? What are the dangers there? How does Tom think about calculating churn when it comes to multi-year contracts?
* What were the findings on what good looks like when it comes to logo retention? How does this differ when comparing SMB to enterprise? What were the commonalities of leading indicators of churn? Is it fair to always surmise that when serving SMB one will always have a higher rate of churn? What is the right way to conduct a churn analysis?
* Assisted vs unassisted: What does Tom believe are the leading benchmarks for both? How does this differ when comparing SMB to enterprise? How does the impact of a salesperson change the conversion rate? What time frame from SAL to closed lead suggests product market fit? What one question must all founders be asking in the sales process?
* How does Tom think about constructing comp plans the right way today? How should comp plans differ when comparing AEs to customer success? Where should the responsibility for upsell lie, customer success or sales? Should sales commission be paid on renewals?
Tom’s 60 Second SaaStr:
* What does Tom know now that he wishes he had known at the beginning?
* What is Tom’s favorite book and why?
* What is Tom’s most recent investment and why did he say yes?
You can also take your SaaStr to go:
Listen on iTunes.
Listen on Google Play Music.
Listen on Spotify.
If you would like to find out more about the show and the guests presented, you can follow us on Twitter here:
Jason Lemkin
Harry Stebbings
SaaStr
Tom Tunguz
Transcript
Harry Stebbings: This is the official SaaStr podcast with me, Harry Stebbings @hstebbings1996 with two Bs on Instagram. A combination of mojitos and workouts will be on the show there. And to our episode today. And some of our shows are just a lot more fun to record than others, to be quite honest. And that’s the case with today’s episode. If you do not read his newsletter, quite frankly, I would almost go as far as questioning your commitment to Saas. It’s incredible and always a must read for me.
Harry Stebbings: Can you guess? Well, I’m thrilled to welcome back Tom Tunguz, General Partner at Redpoint Ventures. The venture fund with a portfolio including the likes of Stripe, Netflix, Zuora, Hashi Corp, and Juniper Networks, just to name a few. As for Tom, he joined Redpoint in 2008 and has since led investments in Kustomer, Looker, Expensify, and Gremlin, all prior guests on the show, I hasten to add.
Harry Stebbings: He’s also the co-author of a brilliant book, Winning with Data: Exploring the Cultural Changes Big Data Brings to Business. And Tom’s also been named on the Forbes Midas Brink list. And before joining Redpoint, Tom was the Product Manager for Google’s AdSense social media products and AdSense internationalization. I must also implore you to check out his blog, tomtunguz.com. As I said, really is one of the very best and always a must read for me. However, without further adieu I’m now very, very excited to welcome back Tom Tunguz, General Partner at Redpoint Ventures.
Harry Stebbings: Tom, what can I say, my friend? It is so awesome to have you back. I’ve been looking forward to this for a very long time. So a huge thank you for putting up with these dulcet British tones once again.
Tom Tunguz: Thank you so much for having me back on the show. I really appreciate it.
Harry Stebbings: Well, not at all, but I want to kick off today with a little bit about you for those that maybe missed our first discussion. So how did you make your way into the world of VC and come to be a GP today at Redpoint?
Tom Tunguz: I had a curious journey. When I was about 17 I got my first taste of startups. I started a little legal software company with my Dad in South America. And then I went to work after college and grad school. I went to work at a startup in Washington, D.C. Called Appian that went public two or three years ago, and we were building business process management software. And then I went to work at Google for about three years as a product manager in ads. And then I joined Redpoint almost 11 years ago. And it’s been an amazing journey. It’s such a privilege to be investing in all these wonderful companies.
Harry Stebbings: Well, from legal software in Brazil to Redpoint, that is a unique journey. So I love that description. But today, Tom, I want to focus on an incredible survey that you did with regards to free trials, and their subsequent effects on how many different parts of both the kind of the financial growth and condition of a SaaS company happened. Does that work well for you?
Tom Tunguz: Absolutely. I’d be thrilled to talk about it. Thank you.
Harry Stebbings: So let’s get the ball rolling, say with number one, annual contracts dominate. So what were the learnings with regards to the relationship and prominence of annual contracts in relation to free trials?
Tom Tunguz: Well, yeah, absolutely. Let me take a step back. So we ran a free trial survey because we found that in most of our startups, many of them at all different price points, even starting around 1K is obvious, but going up to 150K ACV, they were running free trials as part of their marketing campaign. And a lot of the disciplines are the best practices around. Those free trials aren’t really well known. So we sent out a survey and we got about 600 responses back. And so these results … we’ll be talking about the results from that survey.
Tom Tunguz: So we published a presentation about the 10 learnings of the free trials. And the first one as you said, Harry, is around annual contracts. One of the questions that came up was if I’m at a small price point or a big price point, should I tie my free trial to an annual contract, a month to month contract, or a multiyear contract?
Tom Tunguz: And what we found is for most companies about 80% in the 15 to 50K ACV range use annual contracts and that tends to be the standard. And so what would you do find is in the SMB, some of the companies tie the free trials to month to month. And then in the enterprise, some of them tie them to multi-year deals, but there’s actually no impact on conversion rate from the free trial to a paid customer as a function of the contract value.
Tom Tunguz: Oh, if you’re an SMB, the data suggests that transitioning your customers from a month to month to an annual shouldn’t have any impact on your conversion rate. And we all know that an annual contract paid upfront is much better for your business because you’re basically borrowing money from your customer and you can reinvest that cash to grow.
Harry Stebbings: Absolutely. And my word, I can tell him I’m going to enjoy this episode. But you said that about multi year. I had David Skok on the show before and he said that unless multiyear contracts are paid up front, you’re merely shifting the burden from customer success to accounting. How do you think about this? And is that a fair summarization?
Tom Tunguz: The way that we think about multiyear contracts is at the beginning you don’t necessarily want to sign lots of multiyear contracts because on one hand they’re really compelling events and they show a lot of commitment from your customers. But on the other hand you want to go through a couple of renewal cycles really early on in the business to determine whether or not they’re are meaningful churn risks in the company. And so going to a multiyear contract actually kinda deprives you of that opportunity and that learning experience.
Harry Stebbings: I totally agree with you there. You mentioned the element of churn and really kind of seeing that repeatability through those cycles. Dave Kellogg on the show said in the case of multiyear contracts, calculating churn is much more complicated. How do you approach this and what do you advise founders?
Tom Tunguz: Well, it’s tough because you can construct a multiyear contract in a bunch of different ways. You can, let’s say we work for a startup, and we’d sign a multiyear deal that’s worth 3 million and we can structure it in a way that the customer has a 90 day out or has an out after the end of each year. The way that you’re going to calculate churn on that is going to be very different because if you’re 90 days out, you’re only going to book the first three months of revenue. And then the remaining 33 months, does that count as churn, does it not? Whereas if you have a three year deal with one year outs, you could look at each contract separately.
Tom Tunguz: So I think the simplest way to do it is really to look at each year as a separate contract and try to structure the contract in that way so that you can rationalize churn based upon ACV. And then you can compare apples to apples. Or if you’re comparing a three year TCV deal, to three one year deals or a collection of one year deals, then your churn rates are all going to look the same and you can make judgments. Because what you’re really trying to do with that churn number is look across the entire customer base or look across the entire TCV, multiyear deal customer base and make a determination about whether or not your product is a good fit there. And so it’s really important to normalize.
Harry Stebbings: Absolutely. And we’re trying to minimize churn and retain as many customers as possible. And that moves us to the second, which is now we have the client signed up. And we need to keep them, retention. Number two is aim for 90% logo retention. Towards me, Tom, what was the findings on logo retention from the survey?
Tom Tunguz: Yeah. We found about a third of respondents retain 90% or more of their logos, which was much higher than we thought. So the companies targeting enterprise and mid market customers retained much better their logos then the SMB. So about 45% of the respondents in the enterprise, retained 90% or better, and about 37% in the mid market. But if you look at the SMB, you only have about 23% in that upper quintile, which was surprising to me. About 20% of companies in the SMB churning more than 40% of their logos, which is a very difficult place to be.
Harry Stebbings: Absolutely it is. Now this is a very unfair question, but you’ve worked with some of the very best SaaS companies and founders. What do you think the best do so well in terms of retention?
Tom Tunguz: I think it’s different in different segments. So in the SMB, the SMB tends to be predominantly a product driven sale and so the SMB is really focused on ensuring that the onboarding is really quick and smooth and it’s almost entirely product driven. The other thing that’s really critical there is getting to longer term contracts in order to show value. And then the last thing is really around showing ROI. In the enterprise and mid market you can afford to spend much more on customer success, particularly at the beginning.
Tom Tunguz: So at the Series A, what we often advise companies to do is to–on enterprise, you can actually use professional services in order to smooth over the divots in the product in order to increase retention. So if your product is missing a particular feature you can take somebody in professional services or a presales engineer, and they can actually build a prototype and solve some of the feature lack in the product. And so we often encourage companies to do that even if they operate those professional services teams at negative gross margin.
Harry Stebbings: Yeah, no, absolutely. I do have to ask, obviously I speak to a lot of SaaS investors. Today also with the fund [inaudible 00:10:20] wise, and often when kind of kind of discussing an SMB opportunity, they have this inherent thesis that you’re always going to have higher churn rates due to SMB’s going out of business far more frequently. Do you think that kind of inherent bias is really fair? And how do you analyze it?
Tom Tunguz: I think it’s true that SMBs do tend to go out of business more quickly, but I think you got to look at the SMB the way that you look at consumer, which is the payback periods have to be much shorter. So when we look at it, Redpoint invests in both enterprise and consumer. When we’re looking at a consumer business like a direct to consumer business, we we’re looking for payback periods in the three to four month range. If you’re looking at a SaaS company that’s selling to the enterprise, you’re really looking at payback periods that are somewhere in the 12 to 24 month range and so in the SMB, they’ve gotta be sub a year, just because of the cost of customer acquisition, the payback dynamics do look much more like consumer.
Tom Tunguz: That’s really important. I think the other thing is when you’re selling into the SMB, you really want to be part of the core function that doesn’t go away. And so there are a lot of companies that sell into the SMB who sell marketing products and those are products kind of ebb and flow with different cycles. It’s sometimes can be difficult to show ROI. And so there tends to be a lot of churn in those kinds of products. The contrast would be if you’re selling back office software to an SMB like a Xero, like an Expensify, the churn rates are going to be substantially lower because without you, the business is gonna have a much harder time operating. From our perspective, you’d rather be back of house at the SMB rather than front of house.
Harry Stebbings: Yeah, no, I completely agree in that. It goes to that VC orgasm statement, which is a system of records, so to speak. [crosstalk 00:11:46]
Tom Tunguz: [crosstalk 00:11:46]
Harry Stebbings: I would love to kind of switch tacks on retention from logo retention to net dollar retention. What were the discoveries in terms of net dollar retention from this study, Tom?
Tom Tunguz: Yeah, this was surprising. So we look at net dollar retention, 110% net dollar retention means the cohort of customers that we signed up last year, spending 1.1 times as much this year. And what we found is that the top quartile of respondents expand accounts by 120% or more. You’re definitely seeing that a lot of the public companies, if you look at like a New Relic or a ServiceNow or a Box, those accounts, particularly the larger ones, in Box’s case, 5K and above, you’re seeing 120 to 130% net dollar expansion.
Tom Tunguz: And so that’s where the data suggests that startups should really be focusing. The surprising bit for me was if you break that down by looking enterprise, mid market, and SMB, you find that the enterprise accounts are the ones that actually have much better NDR then in the mid market or the SMB. I would have thought in a lot of the SMB companies that we see, they tend to have really terrific NDR.
Tom Tunguz: We’ve got one company that’s doing something like 160 to 170% NDR in the SMB and that’s predominantly because they’re targeting companies that, smaller companies that tend to be growing really quickly. And when you think about a SaaS company targeting enterprise, a 170% account expansion on annual basis seems to be a pretty tall order. But what we do find is that it’s easier to expand for most companies in the enterprise because you can go from department to department and then eventually enterprise license agreement where you’re selling across the enterprise.
Tom Tunguz: So just to kind of summarize all that, what we found is in the enterprise it’s easier to retain logos and it’s also much more common to expand them in the 120 to 140% range.
Tom Tunguz: Whereas in the SMB you’re going to see substantially higher churn rates and the logo basis and then also about 40% see 80 to 100% NDR. That’s clearly influenced by the fact that they’re losing lots of customers.
Harry Stebbings: Now, again, another unfair question, but when we speak about NDR and 170, in some cases, as you said there, in terms of the holy grail question for customer success, how do you think about responsibility with regards to upsell? Does it lie with customer success or does it lie with sales? What’s your thinking on that?
Tom Tunguz: Yeah, the thinking? Well, kind of what we’ve seen in history has changed a lot. And I think where we’re ending up is particularly in the mid market and above, it’s the AE’s responsibility to sell the account. It’s the customer success manager’s responsibility to keep the account happy, creating an account plan, improve the NPS, potentially set them up to be a reference customer. And it’s now falling again back on the AE to do the expansion.
Tom Tunguz: And one part of it has to do with the fact that there tend to be different personalities that gravitate to each job. CSMs tend to have different personalities and different desires than an account executive. I was talking to a CEO recently who said–who had tried to take his CSM and put him into an AE role and what they found is the CSM continued to one, have ongoing conversations and keep building the relationship, where the AE felt more pressure for time and wanted to drive the conversation to a decision about whether or not to expand. And so that may be broadly true.
Tom Tunguz: The second reason that we’re starting to see this kind of demarcation between CSM and AE, particularly for the upsell has to do with compensation. At the beginning, maybe two or three years ago, there were lots of different ways of compensating a CSM. You could do it purely on … So some of the softer metrics like customer satisfaction, that kind of stuff. Or you could do it based on an account expansion NDR multiplier. And what we found is … Or, what the industry’s gravitated to is we’d rather compensate CSMs on a much more salary basis, maybe 15% of the compensation’s variable and goal based, whereas the AEs who are already kind of 50/50 on a full OT plan, 50% is the salary and then 50% is the commissions, they tend to lend themselves much better to the account expansion.
Tom Tunguz: I would love to see and I haven’t seen, but I would love to see a study side by side of people running experiments between CSMs doing upsell and AEs doing upsell, and I think the key variables there would be the time to close the upsell. And then the second would be the amount of the upsell. There’s one downside to setting it up this way, which is the risk that the AE oversells the account before the transition to the CSM. That’s something that every company is gotta manage separately.
Harry Stebbings: They absolutely do. One thing that I go back and forward on is kind of AEs and their relationship with renewals in terms of should they get commissioned for renewals? What are your thoughts on that? ‘Cause it’s often one that founders ask.
Tom Tunguz: Yeah, there’s a really good book. Mark Roberge’s book, talked about the four different commission structures that AEs have for renewals and they went through a bunch of different ones. There’s two questions, I think. The first is do you retire quota at the same rate for renewal as you do for new business? In other words, if I’m an AE and I book a $100,000 deal, $100,000 new customer versus a $100,000 expansion on existing customer, does that retire my quota at an equivalent rate?
Tom Tunguz: The thing there, is in the beginning, what we end up seeing is you see a lot of companies actually valued new ACV much more, say 2X more, 3X more, 5X more, 10X more than renewal. Because at the beginning what you really want to do is start up in order to have a successful growth. And fundraise is try to maximize the number of logos that you have.
Tom Tunguz: And then the other thing that’s really kind of a key part of this is that it’s way easier to an expand a customer than it is to find a new customer. And so in order to show that repeatability to investors, it’s really critical to be able to expand. What you do end up having in the longer term is you kind of section the sales teams. So you’re going to have a sales team that’s going to go after named accounts or the very largest customers. Because let’s say you land a departmental deal with someone like an Apple or a Google or a Salesforce or an Oracle. You really want an account manager who’s going to understand or an AE who’s really going to understand the dynamics of that account and how to expand and grow within that account. Because obviously it’s going to be a strategic channel for your business forever.
Harry Stebbings: No, I totally agree with you there in terms of kind of that segmentation and really honing in on that customer and their needs. On the mechanics of the trials, so to speak there, I often have many founders ask me about kind of optimal structures and incentives behind the trials themselves. Can I ask, Tom, what are the four ways to limit trials? And what did the results suggest when considering which limit performs best for conversion?
Tom Tunguz: So we looked at four different limits. The first is a usage space trials, number of API calls like a Twilio, the second is a time based trial, 14-day trial, that’s the most common, the third is a seat-based trial. Two free seats, if you want the third one you have to upgrade to be paid. And then the fourth is features. Most famous one is Slack. So unlimited users unless you want to search of a particular capacity. What we found is that time is by far, even across all price points, is by far the most common. Roughly about 60% of the respondents are actually using a time-based trial.
Tom Tunguz: And there’s a reason for that. It turns out that the median conversion rate of time-based trials is higher. It turns out that time- and usage-based trials have a statistically equivalent conversion rate of something in the 10 to 12% range. Whereas feature- and seat-based trials have a conversion rate that’s half of that. Those numbers are actually independent of price points. They’re also independent of buyer. What we found in the survey is that if you’re selling to marketing or legal or HR or sales, they all actually buy and behave in the same way irrespective of price point and department. So our conclusion was if you’re going to do a free trial, you really should be doing either a time-bound free trial or a usage-based fee child.
Harry Stebbings: So we should be doing a time-bound free trial? What does that actually look like then in terms of what effect does trial length have on conversion rate? How should we think about that?
Tom Tunguz: It’s totally independent. So it turns out the conversion rate of trials, we looked at seven-day, 14-Day, 21-day, 30-day plus. Turns out if you have a salesperson calling the people who are doing the free trial, they all converted about 15%. And so if you’re doing a time-bound free trial, the best thing that you can do is shorten it. Because your conversion rate is independent of lengths. And so our hypothesis, which we weren’t able to prove, is when somebody signs up for the free trial, that’s the point of maximum intent. That’s the point in their buyer journey when they’re most likely to buy. And so whatever you can do to help them in a shorter period of time buy in a comfortable way, that’s only going to benefit you,
Harry Stebbings: So that’s in terms of kind of independence of the actual length of the trial. What role then does the sales person to have in terms of this impact on increasing conversion? Or is it again independent?
Tom Tunguz: Well, we found that this sale people typically 3X the conversion rates. We looked at unassisted conversion, which is no salesperson touches the lead, and then assisted conversion. And you look at the difference and it’s on the 50th percentile, it’s about a 4% unassisted conversion rate and then a salesperson or an assistance conversion is about 15.5%. So there’s basically a 3X difference. And if you can paint this picture in your mind, what we found is that the conversion rates of unassisted follow a power law and then the conversion rates of assisted, if you look at distribution of different respondents as much more even, kind of a uniform distribution.
Tom Tunguz: And what that means is that if you an unassisted product, there are some companies that have exceptional product market fit that have really honed their funnel and so they’re getting really terrific like 10, 12, 15% on assisted conversion, entirely product driven, but they’re a very small number of companies who can achieve that. Whereas if you’re a company that has, say less strong or less perfect product market fit, salespeople actually help you smooth your performance and actually allow you to get much better conversion rates. And so one of the conclusions from the survey was that if your price point allows for it, which is typically something like a 10 or a 12K ACV, you really should be hiring it AEs to call your leads.
Harry Stebbings: Yeah, absolutely. Can I ask, just in terms of adding those first AEs here, as many are first time founders and kind of scaling founders are, how do you think about when the right time to add those first AEs is?
Tom Tunguz: Yeah, I think this founder’s going to do the first to say 10 to 20 sales and then after that I think it’s typically time to hire an AE. And you have a decision point about what kind of an AE you want to hire. And the way that we think about it is there are people who like to solve problems and there are people who like to execute game plans. And so what you really want to find in the very beginning is people who want to solve problems because the founder is going to have led the first 10 or 20 sales. They’re going to have discovered some of the techniques, but the playbook and the game plan for mechanizing the sales process really hasn’t been established. And so your first AE, or your first two AEs are going to be responsible for doing that.
Tom Tunguz: They’re also going to need to do things like build their own field marketing materials, build the case, understand the accounts. And even the battle cards, the competitive dynamics between one product and another as opposed to somebody who’s come out of a really great sales force, has been given a playbook and can execute that really well and really quickly, you really want the first.
Harry Stebbings: In terms of kind of the pricing that these AEs are selling, I’m always [inaudible 00:21:18]. Honestly, quite stuck actually, Tom, ’cause I always think that you don’t want to discourage users from using your products. But if usage discouraged product use because obviously you have kind of limits on their usage itself. Seats can be shared pretty easily within organizations. How do you think about maybe the right variable pricing strategy given that really you could shoot a hole through pretty much all of them?
Tom Tunguz: Yeah, you can. You know you had Brad Birnbaum, who is a founder who I work with, on the show. He is the CEO of Kustomer, which is a really fast growing customer support platform based in New York. And he was talking about how in his business he was trying to change the industry in two different ways. The first thing he was trying to do is innovate on the product, and we can talk about all the differentiation that they are building there that’s meaningful and important. And then the second, he was going to market trying to change pricing.
Tom Tunguz: And what he found in trying to change the pricing is that customers didn’t have a pain point around the pricing. They had a budget, they had a purchasing process, they discussed it with procurement, so they wanted product differentiation, they didn’t want the pricing differentiation. And so in most cases where you were selling into an existing category with an existing budget line item, innovating on pricing is probably going to slow the sales process down because you’ve got to not only educate the buyer on the differentiation in the product, but you also have to educate buyer why it’s beneficial for them to change the pricing model.
Tom Tunguz: There’s some small subset of use cases. I’m thinking about Splunk or other places where you’ve got other areas in the market where you’ve got one very large incumbent. Mixpanel is example of this. One very large incumbent that everybody in the market perceives to be really expensive, and so people want to buy differently because you’ve got a [inaudible 00:22:45] there. And so differentiating on pricing is actually something that’s going to resonate with a buyer.
Tom Tunguz: But it’s just like the product. Before you want to jump in and try to differentiate on the pricing, it’s really important to have conversation with lots of different customers. And I think the questions or the feedback that you’re going to get, they’re going to suggest that it’s time for an innovation in pricing is, it’s really expensive as my business grows or the incentives of the vendor are not aligned with mine, that’s a pretty clear signal that people are ready to think about buying on a different metric.
Harry Stebbings: Absolutely, there. Almost no stronger signal. I do, and I loved that episode with Brad. He was fantastic to have on. You did mention though, assisted and unassisted, two passion points of mine, which is one of the many reasons I’m probably still single. Yeah. I would love to discuss that. And this is again, insanely unfair of me, but considering it’s round two, I think I can. So what were the discoveries on optimal targets for unassisted conversions? And then this is the unfair part. As an investor today, what benchmark makes Tom very excited?
Tom Tunguz: Well, what we found the 50th percentile for unassisted conversion is 4%. and like I said, it follows a power law. So 75th percentile is 12%, but it’s quite rare to see that. So I don’t think that’s the right metric that people should be aiming for. I think in the four to 5% unassisted range is pretty incredible. And then the other thing they really consider is the sample size. So you can have really great conversion rates on a very small number of people, but that’s not nearly as meaningful.
Tom Tunguz: If you saw a company with a six to 7% unassisted conversion rate that was growing something like 15% or more a month on a super capital efficient base, I may be falling over myself to mess in that business.
Harry Stebbings: Well that was a great answer. Now, I’m thrilled that I asked that one. [inaudible 00:24:17] I can apply it to assisted now on the flip side. [inaudible 00:24:22], here, and again, what would get you falling over yourself to invest in my SaaS startup?
Tom Tunguz: Yeah, so here a 50th percentile assisted conversion is 15.5%. 75th is about 30%. There’s some nuance here because some people measure conversion from different points in the cycle. There’s stage one and stage two. SQL versus the MQL. But as you see a company closing something like 30 to 40% of their leads from the MQL level or even the SQL level, that’s really attractive, particularly for combining that with a shorter sales cycle.
Tom Tunguz: So anything under a 90 day sales cycle with a 30% conversion rate from a sales accepted lead, just means you’ve got a really great product market fit. And the other thing that’s really critical, and it’s a question that I ask in almost every pitch meeting is how do you create urgency in the sales process? A lot of products that don’t really lend themselves to a sales pitch that says, “Hey, you need to buy this now.” But if you can create that urgency, then you’re going to benefit from both of those two things. Really short, short sales cycles and the second is a much higher conversion rate from a sales accepted lead.
Harry Stebbings: Can I ask, what was the best answer that you’ve heard? And you don’t have to say the name, if confidential, but to that, how do you create urgency in the sales cycle?
Tom Tunguz: Well, recently we’ve been seeing lots of different companies that are trying to solve regulatory requirements like the GDPR and all that kind of stuff. And so a lot of them say we create. And there’s this May 15th date, that was a really great answer for a while until May 15th had passed and that regulatory dis-continuity kind of went away.
Tom Tunguz: A lot of the times founders answer at, there’s a 10 X ROI story, that’s really important. Those are probably the two most common answers. I mean outside the regulatory one, it’s probably much more around ROI. Or the other one that’s tends to be really important is around visibility. What we’re starting to see as you’re talking about before, Harry, is these kind of systems of record now almost every team within the company has got a system of record and so whoever the C level or the VP level executive who presides over that department, they want to have the metrics to report to the management team, and if they’re blind then you can go to them and say, “Hey, I’ve got these metrics to help.” And so that’s good catalyst.
Harry Stebbings: No, I love that on the visibility. The final element that I do want to discuss before my favorite, the quick fire, you mentioned getting the sample sizes there, and maybe how that impacts, so to speak, in terms of both unassisted and unassisted. In terms of kind of scoring, now we have all these leads. It’s conventional wisdom in terms of efficiency that it can be gained from scoring these leads effectively. How do you advise founders on the right way to score leads today?
Tom Tunguz: So this was really interesting. So I have a machine learning background. We did lots of lead scoring at Google. We had this internal tool called Glenn Gary that did a lot leads going on the advertiser side. And we always thought it to be super impactful, but what we found is using the data that we had, we found that lead scoring particularly in the enterprise price points, so 15K and above, or mid market and enterprise, is actually a negative signal.
Tom Tunguz: So if you’re selling to a small account, lead scoring provides no marginal information, supposedly, given aggregate data that we have, provides no marginal information over how salespeople are filtering through the leads today. At the enterprise, it’s actually a negative signal. You’re actually filtering out leads that could be really good accounts that you’re not calling because of the scoring. And in particular we’re focused on activity scoring, which means using engagement metrics within the product during the free trial to determine and assess whether or not you should be calling each lead.
Tom Tunguz: And the hypothesis that we have here is that mid market enterprise accounts, people are spending 15K or more, are not going to use the product in the same way that an SMB would. They’re not going to use the product the way that a consumer would. And so they’re going to use a product for five or 10 minutes, then go talk and build consensus internally, talk to a procurement team, try to get budget. And so they may not be so active in the product during the free trial. But if you disqualify them out or don’t emphasize them, then you may be missing on a material revenue opportunity.
Tom Tunguz: And what we found is if you’re looking at 50 to 150K price points, the conversion rate when companies were not using activity to score leads, is about 4X higher than when they are using. And so it’s something to look out on a per company basis. But I think it’s something pretty important to look at.
Harry Stebbings: I want to move into my favorite there, as I said, the quick fire round. So, Tom, you know the drill here. I say a short statement, you give me your immediate thoughts. Are you ready to rock and roll?
Tom Tunguz: Let’s do it.
Harry Stebbings: So your favorite book and why?
Tom Tunguz: I just read Kai-Fu Lee’s AI Superpowers. I think it’s one of the most insightful books on how the Chinese startup economy has evolved and how it’s different from the US. And also what a future world could look like between US chips and US AI versus Chinese chips and Chinese AI.
Harry Stebbings: Yeah, absolutely. I yeah, sorry I just had a guest on the show that said 50 Shades of Grey and–
Tom Tunguz: I remember that one.
Harry Stebbings: Sorry. It was very different to your suggestion. What’s your biggest weakness? And what are you doing to improve it, Tom?
Tom Tunguz: I think my biggest weakness is prioritization. I like to do lots and lots of different things. And one of the things I talk about on the blog is focus, so 2019 for me is a year of focus and the Pareto rule, 80/20.
Harry Stebbings: Why are AI agencies a powerful new go to market for startups?
Tom Tunguz: So in the recruiting world, 5% of spend is spent on tools, 95% is spent on recruiting. In the recruiting world, that 95% is actually commanded by agencies. But a lot of those agencies don’t want to buy the tools in order to kind of become the next generation. And so what we’re starting to see is a bunch of startups in every category of agencies that are masquerading as classic agencies, the power of being powered by AI. And so the trade that they’re making is they’re depressing the gross margin a little, but they’re actually increasing their TAM by a factor of 10 or 20x. And so that’s a super compelling place for us to invest.
Harry Stebbings: I love that. Tell me, Tom, what keeps you up at night?
Tom Tunguz: We have five small children. So keeping them–What keeps me up at night? I think figuring out how to get in front of the next great founder.
Harry Stebbings: No, I’m totally with you. I absolutely, uh, permanently on my mind. So tell me, Tom, what’s your superpower?
Tom Tunguz: My partners tell me my superpower is taking in lots of information and being able to summarize it really quickly.
Harry Stebbings: I’m very envious of you. What are your player coach fraught with risk?
Tom Tunguz: Okay, so lots of startups at the early stage want to do, there’s kind of this classic mistake which is, hey listen, we’ve got a really great performing account executive or somebody who’s in CS, and we want to promote them to lead the team. I’ve seen this time and time again and I’ve talked to many different VPs of sales in some of the most successful startups and none of them have ever seen it work. Not to say that it can’t work, but it’s really challenging and the main reason is it’s a very big jump to go from an individual contributor to becoming a manager in any function, and in startup world, when you’re doing that, what it means is that the business is starting to grow really well and most of the time the startup doesn’t have the time for that person to understand the role, grow into the role and be successful in the role, all in the span of a quarter or two until you’re just asking way too much of that person. So it’s way better to hire somebody or find somebody who’s done it before to train those people and then as the company scales, train those people and help them grow rather than do it when it’s kind of in the critical path that business.
Harry Stebbings: Now the final one, what’s the most recent publicly announced investment, Tom, and why did you get so excited?
Tom Tunguz: So we just announced an investment in Mattermost. It’s an open source Slack product, a communication platform that we’ve been tracking–I’ve been tracking for about 15 months. I’m really excited about this company for a couple of different reasons. The first is cost of customer acquisition in software has been increasing monotonically over the last 10 years since we were looking for new forms of customer acquisition. I think open source in the applications tier could be one. The second thing is we’re noticing a difference in architecture when SaaS companies go to market. So if you think about the first wave of software you’d client server architecture. The second wave was Salesforce.
Tom Tunguz: Putting the application and the database in the cloud so that the customer didn’t have to run it. Now these advances have all these regulations. What we’re starting to see is companies who want the SaaS vendor to operate the application, but they want to operate the database, whether that’s in a VPC or on prem and so Mattermost is architected to go after the market this way. And then the third most compelling part of this business is just the growth rate and the capital efficiency. Know the company is, you know, has more revenue than most series Bs that we see but had raised about 3 million. And so the kind of the growth dynamics within the messaging market are really terrific. And then the fourth is the team. Just loving and everybody’s working in Mattermost.
Harry Stebbings: My word, that must’ve been a long investment mandate, but hate investors, man.
Tom Tunguz: I was just reading this book called In Defense of Troublemakers. Have you read this book, Harry?
Harry Stebbings: No, [crosstalk 00:00:32:05].
Tom Tunguz: They talk about this hedge fund that whenever they make an investment, you have to write an investment memo for and then an investment memo against. And so you’d probably not like working there.
Harry Stebbings: I am not [inaudible 00:32:19] that far. Please, if you find an AI for investment memory creation, I will personally double down on it on. But listen, Tom, this is going to be fun. I didn’t quite expect it would be this much fun. So thank you so much for joining me today.
Tom Tunguz: Thank you so much, Harry. This was a ton of fun. I don’t think I’ve spoken so much or so fast in a long time.
Harry Stebbings: And I’m not sure I’ve had quite as much fun on an episode for quite a long time. So huge thank you to Tom for that. I really do so appreciate it. And if you’d like to see more from Tom, you can find him on Twitter @ttunguz. Likewise, you must check out his blog at tomtunguz.com. It would also be fantastic to see you behind the scenes here at SaaStr. You can see us on Instagram at @hstebbings1996 with two Bs. It’d be great to see you there.
Harry Stebbings: And as always, I so appreciate your support tuning in today. It really does mean so much, and I can’t wait to bring you another fantastic episode next week.
