David Skok on the Keys to a Balanced, Profitable SaaS Sales Model
One of the biggest mistakes entrepreneur and VC veteran David Skok sees software as a service (SaaS) startups make is trying to “buy their way into a repeatable, scalable model.” Sure, investing in hiring more sales reps will naturally result in more customer acquisitions. However, more sales reps is not synonymous with profitability.
In this week’s Labcast, Skok, author of the popular For Entrepreneurs blog, explains why you need to make sure you have a sustainable go-to-market sales model in place before expanding you team. Listen in to learn the metrics to use for measuring the profitability of your model and the goals you should be setting for your sales funnel.
This Week’s Guest
“Ask yourself the following question: Is the lifetime value of your customer more than 3x greater than the cost to acquire that customer? If you’re in that greater-than-3x range, you have a viable business model and one that you should feel confident hitting the accelerator pedal on and expanding.”
— David Skok, Matrix Partners
Startups can be broken up into three phases:
- Search for product market fit. [1:10]
- Search for a repeatable, scalable, and profitable economic model. [1:25]
- Expanding your business. [2:40]
There are two important metrics for measuring the profitability of your go-to-market sales model:
- Lifetime value to cost of customer acquisition. [3:30]
- Number of months it takes to recover cost of customer acquisition. [10:35]
When examining the sales funnel, measure the conversion rate from visitors to leads, to opportunities, to closed deals. Over time, also measure your number of visitors. [14:35]
Conversion rates vary from company to company. Generally speaking, if you’re implementing free trials, aim for a 10-20% conversion rate. On the flip side, if you have a heavily-invested sales proccess, aim for a 80-85% conversion rate. [12:55]
Labcast 118: David Skok Helps You Determine if Your Go-to-Market Economic Model is Viable
Subscribe to Labcast
Kevin: Hello, and welcome to this edition of Labcast. I’m your host, Kevin Cain, and today I’m joined by David Skok to talk about the viability of go-to-market economic models. For those of you who don’t know David, he’s a general partner at Matrix Partners, a venture-capital firm in Massachusetts that has invested in such companies as Apple Computer, Veritas, Sycamore Networks and Guild Group. David has a wealth of experience running companies, having started his first in 1977 at the age of 22. Since then, he’s founded a total of four separate companies, three of which have gone public. Hey, David, welcome to Labcast. How’s it going today?
David: It’s great. And yourself?
Kevin: I’m doing very well, thank you. As I mentioned in my introduction, we’re going to be talking today about economic models and go-to-market economic models and whether or not they’re viable. My first question before we even get to that is, if we think about start-ups and the different stages of start-ups, what exactly are you going to see as those stages and how does that break down?
David: I think of a start-up in three phases. The first one would be the search for product market fit, which is really well talked about with this whole lean start-up philosophy. The second and third, I think, are slightly less well understood and less well talked about, so that’s where we’ll spend a bit of time today.
The second one I think of as being the search for a repeatable, scalable, and profitable sales model. That would start typically towards the end of your work on creating a clear product that actually has product market fit. Typically in that phase, one of the mistakes that I see happening which I would recommend avoiding against is venture investors can sometimes think they can buy their way into a repeatable, scalable model. So they spend money on adding sales people before it’s clearly proven what your process is and how to repeat that process and before it’s known that what the process is can actually scale.
My advice is to avoid that very strongly and to keep your spending quite low, and particularly to be using the founders of the company on as many of the sales calls as possible because they’re the ones who have the ability to change the messaging, change the product, and react to what they’re discovering in failed sales calls and working sales calls to repeat what what’s working. Perhaps use a small number of additional sales people, maybe one or two, to help create the appointments and do the follow-up work that the founders are engaging in.
At the end of that, what you hope to achieve is a process where everything is both repeatable, so you know if you add a new sales person that they’re able to do what the first sales person did; that it scales, so you know that if you want to hit the accelerator pedal and add six more sales people or seven more sales people that what you were doing with the one sales person will actually continue to work as you add more people. And that it’s profitable as well. I think we’ll come back to what we mean by this word profitable with a later question of yours to discuss this whole viability of the sales model.
Kevin: That’s a really great setup of the various stages. Now, does the viability of a go-to-market economic model vary depending on what stage you’re in, and how do you determine whether or not it’s going to be viable?
David: I think the best way to look at the profitability of the sales model is to introduce this notion of unit economics. What I mean by that is to look at the average cost that it takes you to acquire a customer, and then how much profit an average customer brings in. Let’s take a look at both of those things. To compute the cost, what you do is you take all of your sales and marketing expense that you spent in maybe the second quarter, and divide that by the number of customers that you’ve signed up.
So let’s say you spent $100,000 in total in all the salaries and marketing programs in sales and marketing in Q2, and you signed up ten customers. You would the $100,000 and divide it by the ten customers, and that would tell you that you spent approximately $10,000 for each of those customers to acquire them.
Then what you’d want to do is look at the lifetime value of those customers. In many of today’s new models we’re seeing subscription-based models, which makes it a little bit harder than with a perpetual license where you sell a one-time deal where you can immediately look at the value of that deal and say it was greater than the $10,000 you spent to acquire. If it’s a subscription, what you want to do is look at the monthly payments or the annual payments and work out how long you expect the lifetime of that customer to be.
Let’s say they’re paying you $20,000 a year, and you’re expecting the lifetime to be four years. Then you’ve got an expected lifetime value of $80,000. To work out the profit on that, you have to look at the gross margin associated with your relationship. Hopefully, if it’s a software or service thing, you have a high gross margin, maybe in the 80 percent range or 75 percent range, so you would multiply that $80,000 of revenue that you’d make over the four years by your gross margin to get the profit that you’d make.
You’re wanting to make sure that when you sign up a customer, that your sales process is not running at such a high cost that it’s not losing you money. It’s amazingly obvious, it seems, that you’d care about this, but again it’s very surprising how many start-ups I walk into where they really aren’t in balance and they’re spending too much money on customer acquisition and they just don’t make a big enough amount of money out of each customer to make it viable as a business in the long run.
An interesting study that I did in that area was to look at a bunch of different methods for customer acquisition. If it was a start-up, say at the left-hand end here which is the cheapest way of acquiring customers, and work over to what you think of as the right-hand end where you’ve got the most expensive methods of acquiring a customer. On the cheapest end of the model, you’d have models like Dropbox where they’re acquiring customers using a free-mium technique and virtually no touch at all. What I mean by touch is human touch in the sales process.
If you’re a Dropbox user, you probably hear about it from a friend or you hear about it on the web. You go and try it out as a free-mium product, and then at a certain point in time your storage gets beyond the level that you’re getting for free and you sign up to start paying them. But you never talk to a human in that process, which makes it very, very inexpensive. It’s probably around about the $20 to $70 customer acquisition range for a company that’s doing customer acquisition like that.
Then if you step up one level, and you go to something like Constant Contact where they have a mostly low-touch model. But they do use what they call customer success people to make sure that free trials are successful. They don’t do very much selling in that, but they really are working to ensure that the trial goes well and that the customer is successful. There you have a cost-to-customer acquisition that’s more like the $400 range, a 10x increase over the Dropbox thing.
If you go to something further up the scale, let’s take HubSpot, where you’ve got a much more extensive inside-sales function doing a lot of selling work, you’re going to find that the cost-to-customer acquisition there is more like $5,000. So you’ve got a 10x jump again because you’re spending more human time and human touch on the sales process.
And right at the far-right end of the spectrum, we take one of our portfolio companies which was sold to IBM called Netezza. Netezza was selling million-dollar deals for a data-warehousing appliance. It would be typically a very-expensive field sales person out in the regional areas where the customers were plus a sales engineer doing extensive on-site proof of concepts. The cost-to-customer acquisition, in that case, is up in the $100,000 per customer range.
You’ve seen something really interesting with this which is, as you add more human touch into your customer acquisition technique, you’re getting a 10x increase for every jump that you make from one level of human touch to the next level of human touch which really kind of fascinating. It tells you, if you’re trying to design a go-to-market model, that one of the things to pay a ton of attention to here is how can you remove human touch, i.e. selling touch, and move that to marketing techniques or other types of techniques where that’s reduced.
The other thing I should state here is that it’s not a problem to have a high cost of customer acquisition like Netezza had, in the $100,000 range, as long as you’re selling a product to the customer where you’re going to make more money than you’re spending on customer acquisition. The rough way that I’ve come up with to check as a rule of thumb whether you’re in the right range or not is to ask yourself the following question: Is your lifetime value of your customer more than 3x greater than the cost to acquire that customer? I think, if you’re in that greater than 3x range, you have a viable business model and one that you should feel confident hitting the accelerator pedal on and expanding.
Kevin: That’s a really helpful overview. Just to clarify, it sounds like lifetime value and customer acquisition costs are the main pieces of data that you’re looking at. Are you looking at that historically over a certain period of time or is it also forward-looking based on what expectations are?
David: It’s both. I think in the early stages of a start-up you won’t have much data, so you’re going to have to make some projections based on what you think things will be rather than what you’ve got hard data on. But as you become a more established start-up, then your historical data becomes very valuable.
If I can just give you an example of why this metric can be so valuable, there are two metrics that I strongly recommend using. One of them I just talked about is the lifetime value to cost of customer acquisition, the ratio of those two numbers. The second one is, what is the number of months that it takes you to recover your cost of customer acquisition. My recommendation is that you try to create a sales process where you can recover the cost of customer acquisition in less than a 12-month time period. It’s frequently not easy to do that. I’ve known quite a lot of customers that are successful start-ups that are successful with more like a 15-to-18 month period of time.
But the reason why this matters so much is cash flow. If you’re going to invest a lot of money to acquire customers, and you can’t recover that investment for quite a long time period, you’re going to consume tons of capital. That may be fine if you can gain access to low-cost capital, but it’s not going to be good if you’re in a constrained environment where capital is hard to come by.
So, going a step further here, the thing that’s interesting with this is let’s say you sell to big customers, medium-sized customers and small customers. What I think can be valuable is to look at those two metrics, the lifetime value to cost of customer acquisition ration and the months to recover cost of customer acquisition, (CAC), let’s call it CAC to make life simpler. If you look at those for each of those three segments, the large customers, the medium-sized customers and the small customers, I think you often discover that one or another of those segments is much more profitable for you than the others. Therefore it gives you some good guidelines as to know, okay, let’s not focus on this particular vertical because we’re not making much money there.
Or let’s put more focus on bigger customers, because we’re actually really doing very well there. Or maybe you discover here that big customers are costing a fortune to acquire, and they’re really not worth it in terms of the return that we’re getting on investment. I find that these metrics are great ways to help you judge where to spend your money. And also to help you figure out we aren’t currently with a very good, viable go-to-market model and we need to stop and go back and figure out how we can cut the costs out of this or maybe get more money out of the customer to cover our high costs of selling to them.
Kevin: So I can imagine the metrics are just one piece of the pie in terms of assessing the viability of your model. What are some of the others? What are some of the drivers of a viable model?
David: I think the real key drivers are what’s your funnel process look like. I like to break the funnel down into three phases here. One of them would be the top of the funnel, which is how you drive new people into your funnel, i.e. get visitors to your website or start engaging with brand-new connections through things like trade shows or whatever it is that you decide to use.
The second would be what does your middle of the funnel process look like, which is where you take somebody that’s become engaged with you by giving you their name and email address on your website and turning them into an actual opportunity where there’s a budget and there’s a need and there’s a time frame and they have authority.
Then the bottom of the funnel is the sales portion, which is where you actually take that opportunity and try and close it. What I think is very valuable for start-ups to do is to break their funnel out into a series of stages. Now there may be more stages than the three that I’ve given you there, where you start with a visitor and you turn them into a lead, which is where you’ve got their name and email address. Then a lead gets converted into an opportunity, and an opportunity gets closed into a deal.
You may have more stages than that, but what I recommend start-ups do is that they measure two numbers for each stage. One of them is, what is the conversion rate between visitors to leads and leads to opportunities and then opportunities to closed deals. The second thing is you want to measure the number of visitors that you’ve got coming in, and track it over time. I would like to see, in a perfect world, two graphs for each stage. The first graph would be a time series showing me the number of visitors or the number of leads over time, and I’d like to see that increasing. The second graph would be what’s my conversion rate doing.
I’d like to again see the conversion rate between visitors and leads increasing over time, or my close rate of opportunities to closed deals. I’d like to see that increasing over time. Those are the key funnel metrics that help you understand what’s going on in your funnel and help you make the decisions you need to make about what to change and where things are broken and where you need to put effort in to make fixes.
Kevin: Now, are there particular conversion rates that you’re looking for to assume that a model is viable, or is just a positive conversion good enough?
David: You know, I often get asked that question, and there are actually some. For example, one of the things I look at if somebody is using a free trial in a software or service model, if their trial conversion rate is less than 10 percent that’s a bad sign to me. That’s an indication that only a tenth of the people that decided to put their time into doing a trial are finding that the product is good enough to meet their needs. That’s not a good sign. That’s a product market fit problem there.
I’d love to see a number that’s more in the 20 percent range, although I have to say I don’t see it that often, so maybe somewhere between 10 and 20 percent for free-trial conversion to a closed deal. Another area where you might have a conversion rate where I could tell you I have an expectation is if somebody is doing a heavier sales process, and they’re doing a real proof of concept for a customer, I would expect to see something in the 80 percent to 85 percent close rate on those proof of concepts.
If they’re not getting that, it would tell me that they’ve either got a product market fit problem or they might be not qualifying who they’re putting into these proof of concepts well enough and they’re putting in people where they should have asked some better questions up front to discover that they could never have bought this thing because they didn’t have the budget to allocate or other problem existed there.
The high-level question you asked me is are there standardized metrics for that, and the answer is no. They do very greatly from company to company. Really the most important thing is can you get a reasonable enough overall close rate, particularly for the dollars that you spent, so that you end up with a very cost-effective cost to customer acquisition relative to the profit that you make from that customer.
Kevin: At the end of the day, what all these start-up companies are trying to do is achieve a certain amount of revenue. Can you speak at all to the equation that needs to underpin a viable economic model to allow you to achieve a desired revenue goal?
David: Yeah, that’s a great question actually, because it flows right out of those metrics that we just discussed. Let’s say, in your forecast, you’re projecting that you need to do 10 million this year. Let’s take your average deal size, and let’s say your average deal size is $100,000. That tells me that I have to take my 10 million and look at how many deals have to be closed at $100,000 each to get to my 10 million number. That ends up being 100 deals that have to be done during the year.
So now I know, if I need to do 100 deals, and my close rate from opportunities to closed deals is 50 percent, that tells me I better have 200 opportunities that I’m working on because that’s the number based on that conversion rate to get to 100 closed deals. The same thing would be true if I now then look at how many leads do I have to generate, and therefore how many visitors I need to drive to my website. I can just back the numbers up using those different conversion rates.
It’s a great way of checking to see okay, do we have enough leads coming in to get to this $10 million number? Normally the answer is no, people don’t have enough leads coming in. So they’ve got to address that with more marketing, and often marketing isn’t scalable. They’ve got to figure out how are we really going to be sure that we can generate that number of leads here, and what’s the process that we’re going to do.
Another thing it will help you with is how many sales people do you need. You will have known by now that an average sales rep, let’s say they’re able to close for argument’s sake $1,000,000. That tells you that to do 10 million, you need ten sales people that are fully productive at the beginning of the year. It’s no good hiring the tenth sales person in the last month of the year, because they’re not going to be productive to help you out.
That’s how you can work backwards from your sales target and understand we need this much happening in marketing to drive this much lead flow, and we need this much productivity out of sales to do the closing between these opportunities that are coming in, and this is the number of people that we need in sales to be able to actually produce that much productivity.
Kevin: Broadly speaking, are there any guidelines that you offer companies in terms of what their sales and marketing spend should be as a percentage?
David: It does vary quite a lot. Certainly in the early start-up phases, it’s hard to be too prescriptive about this because, frankly, you’re not looking for profitability so much as you’re looking for the maximum growth rate. I’m more worried, actually, about these two metrics I described to you earlier on, which is are they still in balance with cost of customer acquisition and lifetime value? And are they still maintaining a very good months to recover cost of customer acquisition?
Because if they’re doing those numbers correctly, I know they’ll automatically end up in the right range of sales and marketing as a percentage of the revenue, i.e. a viable business model in the long run. That’s much easier to say than I want to see sales and marketing at 25 percent of revenue, because the problem for most start-ups is that they’re not trying to get to profitability that early in their life and therefore that 25 percent target is kind of meaningless. It’s too difficult for them to actually use.
Kevin: David, I want to thank you for taking the time to join me today. This has been really helpful in talking about economic models and their viability. I really appreciate your time. Just so our listeners know, how can they get in touch with you if they want to contact you?
David: Maybe one of the first things to point out is that I write a blog on this topic and so have lots of articles that describe cost to customer acquisition and lifetime value. The blog is probably easiest found by typing “Skok blog” into Google, or the actual address is http://www.forentrepreneurs.com/. On that they will find my email address, which is email@example.com, and my Twitter handle, which is @BostonVC. Hopefully that will help people get a hold of me. I’m very open to interacting with entrepreneurs to try and help them as much as possible.
Kevin: Again, I thank you for your time today, and really appreciate your talking with me.
David: It’s a pleasure. It was great talking to you.
Photo by: Colin Harris ADE
It’s not an easy decision, but most important ones in life aren’t.
We hope this framework makes planning during this uncertain time feel less like summoning a crystal ball and more like navigating with a map.
In this environment, it doesn’t matter if you’re the CEO of a startup or a well-established company—you’re going to have to make some difficult choices that will probably keep you up at night.