Scott Maxwell has another excellent post on his blog. This time it is on a company’s need to measure and monitor their business. In an early stage business, I typically see two types of companies. There are those companies that do not measure and monitor much and instead drive their business by a "seat of the pants" decision making process. Other companies manage and monitor everything. The key is not to get stuck in the weeds and get paralyzed by analyzing too much data. For all of my portfolio companies, I like to know what the 4-5 key drivers of the business are. I like to know what leading indicators are most likely to show an increase or decrease in sales 1-2 quarters ahead and what the company is doing to improve those measures.
For example, when I was reviewing a financial model with a portfolio company which generated revenue through advertising it was clear that while traffic was a huge ingredient in revenue generation, RPM was an even stronger metric as each change in RPM significantly drove sales. Over the next month, we experimented with a number of changes (of course we measured and monitored each change against the benchmark) to determine how to finetune RPM. Why spend more dollars on getting more traffic to our site when we were not effectively monetizing in the first place? Finetuning our RPM is an ongoing process but now for each dollar we spend on generating traffic, I am confident that it will return much more than a dollar in revenue to the company. Metrics matter and understanding what inputs have significant leverage on your operating model is the key.
Yes, I know it is hard for early stage companies to accurately predict their revenue. But from my perspective, the financial model is not as important for accurate revenue prediction as it is for understanding how the economics of your business works. Does each new customer make money or lose money for you? In any financial model there are usually a handful of inputs that drive the overall sales and cost equation. Make sure you know what they are, how sensitive your revenue and costs are to that input (in the earlier example, traffic was a key input but RPM had a more direct impact on sales), and measure, monitor, and finetune your company based on these important pieces of data.
I was at a board meeting the other day and while we were pleased with the results for the quarter, we were struggling to understand why we were not getting more customers if we were winning a majority of our proof of concepts (POCs). As we dug through the data we discovered that while we did convert a majority of our POCs, 50% of POCs ended up in no decision. In other words, we wasted half of our sales engineering resources on sales that would never happen. The key was to go after the low hanging fruit first – only do POCs that can convert into a sale. So what have we done to correct this? We now require a more detailed checklist before a sales rep can request a POC for a customer. Even if we are able to reduce the no decision rate from 50% to 35% this means more sales. Clearly this does not mean we need to hire more sales engineering bodies as we can better utilize who we already have.
It doesn’t matter if you are an enterprise, web 2.0, or old economy company because everyone must understand the key drivers of their business, measure them, and finetune their operations to run as efficiently as possible!
I agree completely with the idea that determining what the real drivers of any business are is essential. However, I think the POC example you gave has the potential to be flawed. In a sales cycle there are many reasons why a deal can go south even after a successful POC. I trust that you made the right decision in this specific case, but as a general rule, it would also be worth considering price, contract language, variance between sales reps, etc. as reasons the deal didn’t close. Furthermore, it is definitely worth considering the timeframe and external events involved in the purchasing decision. For example, in security sales cycles, POCs that don’t generate immediate sales now, often lead sales after the next security incident. A 6 month window might not catch this, whereas a 12 month window probably would. Thus in some circumstances a tighter POC qualification process could ultimately hurt sales.
excellent points. yes, these are definite factors to look at. in my specific example, which I should have further clarified, the POCs were all competitive bakeoffs in which there was no winner but a “no decision”. Having 50% of our resources on these was not a good use of our time. Of course, we are tracking the “no decisions” to see what happens a few quarters down the line
Sounds like you made the right choice. A no decision in a bake off are usually related to budget or (and much worse) overall value proposition of the product. Budget is something that can be filtered out through a tighter POC qualification, but unclear value proposition is not. I would pay close attention to whether or not the tighter qualification has the desired effect because, if not, it may be an indication of a deeper problem with the product. In this situation paying attention to demographics is essential and it may turn out that a custom tweaked solution (very slippery slope) for a vertical or SMB is the answer.