I still remember the mess that performance metrics brought to a struggling e-commerce business. Our enlightened board challenged the CEO to provide performance metrics, which then consumed hours of statistical drivel over "the anal-retentive guide to performance metrics." These endless discussions proved a massive distraction from the core business issues, and about only thing worse than the statistics themselves turned out to be the company's performance, and its future. Don't let it happen to you!

How many metrics are too many? I think you should have no more than four, along with a few obvious, notable exceptions. Everyone watches sales, profits, and traffic: this month vs. last, year ago, next, and so on. Energetic conversations almost always follow.

But how do you measure traction, arguably the most important company metric after solvency? Some VC's make simple, bold statements like "at least 10 or 15 percent revenue growth per month" or "grow 33 or 50 percent year-over-year."

Gross measures are important, for sure, but here are four others every CEO should know--even if awakened at 3 a.m. for a quiz. They're subtler than the obvious metrics, but often more instructive when assessing "traction."

1. Customer Engagement Improvement

Are customers increasingly curious about the product, the company, and the details behind its offering? This is easily measured in one of several ways, or perhaps a combination of:

  • Are people spending an increasing amount of time or page views on the website?
  • Is the website driving a steadily increasing number of contact requests?
  • Is the number of referrals of web content increasing at an increasing month-over-month rate?

Focus on the metric most relevant to you. Track it weekly (not daily, which will drive you nuts) on a graph, footnoting extraneous influences like holidays, marketing spends, or competitive activity.

2. Customer Acquisition Cost Improvement

This critical key to sustainable growth and profit is seldom analyzed in its own right, as a standalone performance metric, yet perhaps more than anything it tells the team if it's found the "magic formula" for sustainable, profitable growth.

Companies need to focus at getting more efficient at customer acquisition over time, as they test multiple acquisition alternatives and learn from failed and successful experiments, learn more about the customer, and ultimately get better at finding and converting prospects. Here, the metric's value is not the cost-per-customer per se, but the company's ability to improve that cost-per, steadily and consistently, as it gets smarter and more experienced.

How much improvement is enough? Hard to say, but work with the team to reduce the acquisition cost by perhaps one or two percent per quarter, which if successful yields a four to eight percent annual improvement in the bottom line!

To do this, first total all the acquisition-related costs, ranging from the obvious Adwords or other media buys, fees from vendors, and marketing(but not sales) salaries. Compute the "all in" costs of driving prospects into the sales team's hands. Then compute the ratio of spending to new revenues from new customers. (Don't confuse upsell and next-sell revenues with new revenues.)

Example: Total monthly spend was $20,000 last month to deliver $60,000 in new revenues, yielding a 33 percent acquisition cost. Next month, the same spend yields $63,000 in new revenue, or 31.75 percent cost! The exciting metric is the improvement of 1.25 percent month-over-month, and that's the number to watch(times twelve, it's huge). If it isn't improving steadily, get to work!

3. Purchase Frequency Gap

Especially helpful for software and subscription or consumable products is whether people are buying or visiting your site more frequently. Its a key health indicator (unless you're selling caskets).

Dig into the purchase data on a customer-by-customer basis to measure either the days between visits or purchases, or the number of either in a specific period like a quarter or a year. Or analyze the market basket to see if the spend-per-visit is increasing, or customers are buying more stock keeping units or fewer on each trip.

For the advanced statisticians reading this, segment the customer list into "vintages" to see, for example, if newer customers are buying more and more frequently, while old-timers seem to be shopping less, or perhaps shopping elsewhere. (At the granular, individual customer level, this can produce some powerful sales outreach activities.)

4. Complaints and Support Decline Rate

Again here, focus on the ratio and the improvement in that ratio: the number of complaints or support incidents, compared first to the total number of customers. Second, is that ratio decreasing over time, as in "three percent fewer support tickets per 1000 users" or some other very simple, inarguable measure demonstrating happier customers at a consistently improving rate.

Whatever you measure, be sure the calculation itself is clearly defined and that you're not measuring more than four or five of these "traction metrics." When they're not improving, it leads to wonderful conversations about how to fix the indicator that's sliding, thereby improving business traction.