Do Metrics Matter?

September 15, 2015
Ashu Garg

It’s become increasingly fashionable for entrepreneurs to start every presentation with slide after slide of metrics. In fact, as investors, it’s normal to assume that if you don’t see metrics in the first few slides, then the business either isn’t doing very well or is still pre-launch. And every year, what metrics matter seem to expand. For every metric, there are multiple variations of the definition; so much so that A16Z recently published a blog post on the 16 start-up metrics that matter and how you should calculate them. Seriously, 16 metrics! And that’s not even counting the variations that they clubbed into 1 metric.

Metrics certainly matter… after all “you can only manage what you measure”. However, there is also too much of a good thing. While we typically start by defining a set of goals and then translating them into metrics, often metrics become the goal and at times incent the wrong behavior. As Eliyahu Goldratt, an Israeli physicist- management guru said, “Tell me how you measure me and I will tell you how I behave.”

Every business is unique and different metrics matter at each stage. Different metrics pull you in different directions, and so while no metric is wrong, an excessive focus on metrics can lead you to make decisions that feel good, but destroy value in the longer term.

For example, I was discussing 2016 priorities with one of my CEOs who leads a fast growing SaaS company that is on a path to do $35-40M in GAAP revenues in 2016. While the CEO tracks 20+ metrics, he has been historically focused on exit MRR. However, of recent, he had been getting advice from prospective investors, mostly growth stage folks, to focus much more on the unit economics and specifically the Magic Number and CAC/LTV. While both numbers are useful and the company had been tracking them for several years, I suddenly noticed a change in behavior. Hiring decisions seemed to be influenced by how they might impact the Magic Number and there was talk about adding either CAC or the Magic Number as a goal for the management bonus payout. At first, this seemed like a good thing. The company has some scale and eventually the unit economics determine whether its business has enduring value or not.

However, here is why that instinct may be mis-leading:

1. In an enterprise focused SaaS business, whenever you hire sales reps, it takes several quarters to ramp up and so if you slow down sales hiring, both CAC and the Magic Number will improve right away. However, is that the right thing to do for the business? Should the company sacrifice growth to artificially improve its Magic Number from 0.8 to 1.2?

2. It’s usual to include all marketing related costs, including headcount costs, in the CAC calculations. That includes product marketing costs which one could argue should not be included in CAC. And if you exclude those costs, the CAC actually looks pretty darn good. So, should the company reduce investment in product marketing so as to improve CAC? Or should it just exclude those costs and end up with a custom definition of CAC?

3. The company has a “land and expand” sales strategy with a strong track record of increasing ACV by approximately 50% in the 2nd year and exceptionally low churn. If you factor that increase into the LTV calculations, then the CAC/LTV looks pretty darn good. However, now you are back to a custom definition of LTV

And so in the case of this particular company, while prospective investors seemed to be focusing on the Magic Number and CAC/LTV, setting some arbitrary target for either metric would cause it to sacrifice growth in a market that is currently a land grab. IMHO, that would be the wrong answer for the company and would sub-optimize the eventual outcome.

So what should an entrepreneur do? Stop measuring the 16+ metrics that matter, according to A16Z? Tell their board members to go to hell when they come up with a host of other “custom metrics” to track? Maybe…

On a more serious note, its really important to be clear about what ‘True North’ for the business is at any given stage and focus on communicating the priorities/goals broadly. Then pick a few (ideally 2-3) metrics that will help you measure progress against those goals and explain the connection to people. And then where possible, track the trend line on those metrics vs. setting some arbitrary target. Remember, metrics drive behavior and the wrong metrics/targets can drive the wrong behavior.

***

PS: If you have stories of how metrics have driven the wrong behavior, do share them with me and I will publish them in a follow up post.