David Skok, five time serial entrepreneur turned VC, came to talk to the Lean Startup Circle Boston crowd last week. If you’re not familiar with David, then I encourage you to check out his blog, ForEntrepreneurs.com, where he regularly shares his expertise and lessons learned from successful companies such as SilverStream Software, JBoss, Tabblo, and Hubspot.
From optimizing sales & marketing funnels, to unclogging blocked funnels, David explained different types of sales models, and how to take advantage of the web to drive down the cost of customer acquisition and move from high-touch to touch-less sales models. Slides from the presentation are below. After his presentation, David stuck around for an additional 30 minutes of Q&A with the audience. After flipping through the slides, scroll down to find the audio recording and text-based transcript of the Q&A, which contains a ton of great, very specific, advice.
Slides from How To Build a Sales and Marketing Machine
Play the Audio of the Q&A with David Skok
Read the Transcript of the Q&A with David Skok
John: So as is typical here, we love audience participation. And David’s agreed, like Eric Ries before him, to do a little Q&A. So to folks to have questions, just raise your hand. Matt has a mike, I’ve got a mike up front, and we’ll do Q&A until you run out of questions, or David needs to go home. So, go ahead.
Audience: Thanks for the presentation. What kind of relationship do you like to see between your CAC and LTV?
Skok: So, I have a rule that I’ve come up with, and it’s a rule to be broken and tested and played around with. But I think that if you have a LTV three times greater than your cost per customer acquisition, you’re in great shape. If you go less than that, if you’re a very big company you’re probably OK if you start reducing the margin closer than that. But if you’re a smaller company you’ve got to still pay for your engineering group, you’ve got to pay for your finance and engineering group, and you’ve got to pay for some other expenses there, particularly the growth part, because you don’t really have to invest money ahead of the cash coming in. So that’s my rule of thumb. Definitely open to being broken, but it’s a starting point for you. And companies that are really good out there have CAC LTV ratios of somewhere between five, and sometimes we see them even as high as eight, but not many times higher than that.
Audience: So going back to when you had that book that you were reviewing all that software, if you started giving away that book what were you guys actually selling then?
Skok: So we sold…At the time, I can take you back to when this was. This is 1985, 1986, a typical customer you walked into had maybe 10 or 15 PCs at the time. They knew they liked Lotus 1, 2, 3, but they didn’t know which project management or 3270 emulation package. We sold the entire suite of software. We had arrangements with Lotus. We were Lotus’ largest reseller in Europe, Microsoft’s biggest reseller in Europe, and we just made margin on the sale of the software. I have to say that I learnt at the end of that, that wasn’t a great business model. I wouldn’t do that again. I like to own my own IPs, so I don’t like to resell other people’s products. But nonetheless it was still a fun business, and it went public in the 1988 time frame. Does that make sense?
Audience: So on your CACs, you factor in marketing cost as well as sales, yes?
Audience: Can you talk a little bit about how you allocate marketing cost, because sometimes that’s a little more challenging than cost of sales. Also because often times companies that have lower cost of sales actually spend more on marketing.
Skok: Yeah. Yeah. So the answer to that is, the way I like to calculate CAC is take the whole of your marketing headcount, sales headcount, marketing spent, lump all of that together, and ask how many customers did you get during that quarter, and divide the two. Now if you’ve got some executives in there that will scale as you add more sales people, more marketing stuff, maybe take them out or allocate just a small part of that, because you know that they won’t need to be increased to give you a more accurate number there. So, it’s not too much. One doesn’t want to get so hung up on the math here that you get wrapped around the axle, but it’s great to be thinking about it in broad terms rather than just going down to every last penny I think. Does that answer it?
Audience: Hi. When do you suggest companies begin this process relative to a stage in product development. Do they have to wait until a product is ready, can they do it sooner than that? What do you suggest?
Skok: I think you want to be thinking about this at the time that your product is starting to get to beta, and you’re beginning to think about, OK, what are we learning from these first customers that we’re doing beta tests with, and what are they telling us. So when you’re collecting feedback from them, you’re collecting both product feedback and also go-to-market feedback. Because they are the interviews that you want to be doing there. So I think that’s the right type of time frame. And you’re going to just guess at this stuff, and that’s fine. But if you don’t put something down on paper and, particularly, start addressing these concerns and what are they like, those will give you these brainstorming insights that’ll allow you to come up with some of the cool breakthroughs, like the free website creative-type products.
Audience: What’s your take on live chat on the website as a way to increase [xx] due to recent claims [xx]?
Skok: I’ve never worked with it, so I can’t answer that question. Does anybody else in this room have any experience with live chat, that can answer that? A couple of people, one in the staircase.
Audience: Yeah, it does increase [xx] and it’s a really great tool for [xx]. Because people give you insights on live chat, but they won’t be bothered to drop an email [xx].
Skok: Great. That’s very useful to know. Which product were you using, by the way?
Skok: Yeah. Excellent. I know there are several others of them around that have been…
Audience: What’s the cost of having black jack support?
Audience: We pay [xx] per month [xx] per month. I spend time [xx] spend time [xx].
Skok: Do you remember the percentage conversion increase, by any chance, from the A to the B?
Audience: I mean we turned it on one day [xx] settings [xx]. But there wasn’t [xx]. I can’t remember.
Skok: You can’t remember, that’s fine. Don’t worry. That’s a really interesting, great question.
Audience: I recently interviewed a company, and they were big fans of the high velocity sales model. And so, and they apparently had automated their entire process and instrumented it. And then I got a call, someone I know recently registered a number of their white papers. And I got a call a week later from one of their sales people trying to sell to me. And I was thinking, how did I become an NQL? And that has always been a kind of [xx] contention in the organizations that I’ve worked out, is the definition of an NQL, or sales accepted later. In your experience, how are people getting to that definition: is there a consensus built between sales and marketing? How is it defined?
Skok: Yeah, it’s really interesting, because those were slides that I pulled out of this deck to try to make it short. So, the common argument is the sales guys, if you ask them, “”Are you getting enough leads?”" they always say no. And if you ask the marketing guys what’s going on, they’ll say, “”That’s crazy, we gave them a ton of leads!”" So the issue is exactly what you described, which is there isn’t a common definition. So I think the answer is exactly what you suggested. You want to get the two of them together and define the criteria, the set of questions that you want to ask that lead before sales feels comfortable that it should be handed over to them. So you know when you’re going to hand it over to them, they’re not going to have it sitting around because they don’t trust marketing to have given them something worthwhile. And then, sales accepted lead. Generally there’s only a very small drop-off between NQL and sales accepted lead, but that’s when the sales person gets ahold of them, and discovers that maybe there’s something completely wrong in the way they answer those questions. So there’s generally like a five percent drop-off, where they answer the questions in some way that was misleading. By the way, I meant to talk about one key function in the middle of the funnel, the Mofu automation software. This was key at JBoss. Remember we turned on 10,000 leads? If you were a sales guy and I gave you 10,000 leads, you wouldn’t know what to do with them. There’s no possible way you can handle that volume. So the next thing that we did was we actually installed [xx] really, really early in [xx] life time, before they figured this out as a technique. And we used this concept where they dropped a cookie into the browser at the time we gave them the email address, and we then tracked that person around the website, and whether they opened emails. And we assigned a score to them, and this leads scoring allows you to figure out pretty nearly automatically when people have started to reach a level of qualification. Now they weren’t so qualified enough for us to hand them to sales, so we then did a further thing. Which was we either ran emails out to them or phone calls out to them, to ask them some questions. And how many of you have heard of the term “”BANT,”" B-A-N-T? So you know what I’m talking about. BANT stands for budget, authority, need and timing, and that’s a commonly used qualification criteria that your telemarketing folks will do before they really call something an opportunity. That would be the next stage of qualification. So we went from raw lead to marketing qualified lead to sales accepted lead, to opportunity, and then closed deal. And you may have other stages in your sales cycle. That gives you some sense of the flow there. Does that make sense to you and does that answer your question?
Audience: Could you talk a little more about what separates light touching insight sales model to a high touch insight sales model?
Skok: It’s simply the amount of calls and the length of the sales cycle. So some products are much more complicated and require several demos before somebody gets it. Sometimes they may actually call their boss in to see the thing. So that starts to become a high touch. And it’s a gray-scale. There’s no, again, hard and fast rule about it. But you will notice that if you start computing the cost of customer acquisition, that there’s a big difference between those two, because the amount of human touch really has this huge like 10X factor increase in the cost of customer acquisition. So you can look for ways to simplify the product or cut the sales stage down, sell them something early and then upgrade them or upsell them afterwards to something else. Great.
Audience: A quick question. A great presentation, thank you. If you’re a [xx] and you’re building things up from ground zero [xx], what would you suggest on these things would be the primary focus of what you talked about today?
Skok: So the top of the process for me would be, unfortunately you can’t take one of these steps here and just do that one step. But you do need to think of how do you close deals? So you’re going to have to design some kind of a process. And I think the top message that I’d want you to leave with here, is design it from your customer’s point of view, not from your own standpoint. So, I hope that answers your question because I think you were asking me to give you like one single thing. But…
Audience: No, I think it was more involved around two or three critical things, or what are the top most things that you focus on as part of that [xx]. [xx] putting that in that system, that approach [xx].
Skok: Yeah. Yeah. I don’t know that I know how to answer that. I wish I did. You can skip the automation step in the early days, there’s not enough volume to automate it. You don’t have to spend money on the automation piece. But you are definitely going to want to figure out what do you need to do to get somebody from being a suspect to engaging with you, to closing a deal with you. And there’s no easy way to cut out many of the other steps that I talked about in there. I’m sorry I’m not answering your question, I’m trying to.
Audience: No [xx]. There’s no easy way to [xx].
Skok: Yeah. Yeah.
Audience: I was just wondering, considering the low cost sales model, and particularly no touch premium, have you found there’s an upper limit on the purchase price per unit of the products and services?
Skok: Yes. So far for the no touch, the highest price that I’ve seen somebody doing around that is in the kind of $500 a month range, but that’s pretty high. Many of them have actually, they kick off with a much lower sale than that. It’s sort of more in the $60 to $100 range. I don’t want to create a hard and fast rule there because you could find certain types of customers where they’re selling much more expensive. So this shouldn’t be a rule, but it’s just kind of like there’s going to be like a curve. Most of them are under that $100 range, a few start to get higher than that.
Audience: Hi. My question is about as you scale, start to think about different markets segments, whether they’re different geographies or customer sizes, how do you