In a couple of weeks, Marketing Sherpa will be holding its 8th annual Selling Online Subscriptions conference. In line with the MarketingSherpa approach, speakers at the S0S conference share case studies of how they have improved subscription and retention of online customers.
Among the presenters at this year's conference will be Jesse Lipson, President and CEO of secure file transfer provider ShareFile. Jesse agreed to provide Content Matters with a snapshot of his case study in advance of the conference. The ShareFile case study shows how they were able to build the business from zero to more than 150,000 users in about two years, leveraging quantitative modeling in order to project subscription revenues.
Here's our conversation (conducted via email):
Content Matters: Tell me a little bit about ShareFile.
Jesse Lipson: ShareFile allows companies to exchange files with their clients easily, securely, and professionally. With ShareFile, companies can set up a password-protected area for clients where they can securely exchange files that are too large or confidential to be sent via email. Examples of such files include high resolution images, video, QuickBooks and other accounting documents, medical or legal documents, CAD drawings, and business plans.
CM: What are the primary markets that you serve?
JL: We primarily focus on the small business market to companies with 1 – 50 employees. We have a long tail of industries that we service (well over 50), but our top industries include consulting, advertising/marketing, accounting, and construction.
CM: As a startup, what has been your go-to-market strategy?
JL: We’ve focused primarily on online advertising with a special focus on paid search advertising to date. As we have continued to grow, though, we are beginning to do some industry-specific for the verticals that seem to get the most value out of our product.
CM: You’ve taken a more quantitative approach to projecting subscription revenues. What’s the background on the development of this model?
JL: I have always been very quantitative in my approach to business plans. Whenever I look at a business (even if it’s not mine), I tend to naturally try to break the business into its core components to understand the variables that drive it. So when I started ShareFile about two and a half years ago, my first instinct was to try to create a simple model in Excel that could help me understand the business. Over the past couple of years, as I continued to look at other subscription businesses and attempt to model them, I realized that the key variables in the model I had created pretty well applied to all subscription businesses from magazines to gym memberships.
CM: Can you provide an example of how you’ve leveraged the model in driving business descriptions? What were the results?
JL: One of the things I like best about using models is the ability to run “What If…” scenarios and quickly see what the impact of a potential decision would be on subscribers, revenue, and profit. About six months ago we decided to try to increase or free trial conversion rate by making a telephone call to each free trial to follow up and see how things were going. We ran a test for one month by calling 50% of our free trials and refraining from calling the control group of the remaining 50% of free trials. The test group (trials that we called) had a 61% conversion rate while the control group had a 57% conversion rate. I plugged this new assumption into our model along with the fact that calling each free trial effectively increased the cost of each free trial by a couple of dollars in employee time.
The result was that the change was a net positive for us, and this new technique will generate about $500,000 in additional profit for the company over the course of the next few years.
CM: One of the key values of modeling is to avoid making mistakes. Have you had any cases where your gut feel was to do something but the model helped you avoid making a mistake?
JL: Yes. We combine modeling with testing to help validate our theories. Modeling alone can help you understand your business better and can help you run “What If..” scenarios (which often help you figure out the most important things to test), but ultimately running a real-world experiment is necessary to validate the assumptions of the model. Likewise, just doing testing alone won’t give you a clear picture of the impact of a change, even if you can observe the “winner” of the test.
One example of a case where my gut seems to have been wrong is the price point of our product. I had been reluctant to raise our prices since our current price point had been working well and I was afraid that an increase in price would hurt our subscriber numbers. However, plugging some scenarios into our modeling tool helped me feel confident that even with a fairly large drop in new free trials, we could make more revenue with less cost by increasing the price of our product. This type of decision is difficult to make by gut feel alone since there are many variables involved including revenue per customer, free trial conversion rate, viral growth rate, and churn rate.
CM: For companies who have always relied upon qualitative decision making, can you suggest some small steps they might take to introducing a more quantitative approach?
JL: The biggest objection I hear toward modeling is that there are many complexities to the business and so it can’t be captured in a model. My response is that a model is intended to be a simplification of real life and not to expect it to be a fully accurate way to represent the outside world.
My suggestion for qualitative companies is to start simple with just a couple of variables and the acknowledgment that the model is not fully accurate, and then drill down on those first couple variables and blow out the model over time as they become more comfortable and see the value of modeling.
For example, let’s say the company in question is a law firm. To begin at a basic level, the law firm makes revenues and has expenses. Revenues in law firms are typically a function of time billed. So, you could drill down on revenue and break that into the number of hours billed per month times the average hourly rate of your lawyers. The number of hours billed is a function of the number of attorneys in the firm multiplied by the number of hours that each attorney works per month multiplied by the percentage of their time that is billable.
So, just like in the above example, companies can start with a simple model in Excel and just continue to expand the model as they feel more comfortable with it. I think that even qualitatively focused companies will be able to look at models like the above and start to think of questions like “Why are our lawyers on 40% billable?”, “What can we do to increase their billable time?”, “If we could make 50% of their time billable, how much would that increase our company’s profits?”.
CM: Thanks for sharing your insights. I look forward to seeing your full presentation at SOS.
Jesse Lipson will be speaking at Marketing Sherpa's Selling Online Subscriptions conference May 12-13 in New York. I'll be attending the conference; if you plan to be there, be sure to say hello.
