Like many startup businesses, Shipa Freight has a north star metric. Ours is bookings. It’s a simple growth metric and very much correlated with our financial performance.
However we’ve found this metric alone doesn’t give us a true sense of the day to day health or the long-term sustainability of our business. And if our only business goal in a given month is a certain number of bookings, we can hack our way to that outcome without adding long-term business benefit; we can increase ad spend, offer discounts, promise things we can’t deliver, etc.
I find true indicators of health fall into one of three buckets:
- Engagement – A customer’s use of our platform correlates well with the value it is bringing them. If they do other things than just book a shipment, such as upload documents, access invoices, and track their shipment, they are more likely to book again in the future. We find MAU as a good indicator of engagement.
- Satisfaction – Customers have several ways to tell us about their experience. The most basic is NPS, which we rely heavily on.
- Loyalty – Our platform is working well if customers are coming back. Bookings per customer, churn rate, and number of cancellations all help us understand if the platform is delivering the value advertised.
I’m a minimalist when it comes to metrics. These are all I need to understand the health of our business at a high level. And when we mix in all sorts of other metrics, particularly vanity metrics like page views or price quotes generated, we get a distorted view of what’s really going on.
But there are lots of metrics we use to understand the ins and outs of our business so we can invest our resources in things that matter. With sales and marketing, we look at ROMI/ROS to understand the basic economics of acquiring customers. We seek to optimize each stage of our conversion funnel, determining where and for whom we use automation (chat bots, email marketing, etc.) versus sales people. Through A/B testing and analyzing step and aggregate conversion, we gain a reasonable understanding of the impact of changes in our sales support model. We do the same for operations, determining which processes are best to automate and where the talents of our ops people can best be applied.
On the pure product side, we collect data on usage of particular features and aggregate impact of a set of new features on the key metrics like MAU and retention (bookings/customer). We compare actuals with our original projections to determine if the features are delivering the value we expected.
As with most products, getting our pricing right is critical. In a relatively low-margin business, we seek to understand how price sensitive our customers are. What makes this challenging is it varies significantly by user persona, trade lane, time of year, and state of the industry. There are few industry pricing benchmarks, and pricing in the logistics industry is very fluid. This is where data science has helped us. By analyzing large data sets of price quotes and conversion data, and segmenting by different criteria, we can coorelate price and conversion rate. We have used this data to make pricing adjustments on key trade lanes, knowing modest improvements to price would boost sales. We don’t have a scalable process for this yet, but early results tell us this is something we need.
What we do today is not the end of the story. We are constantly improving how we use data to make business decisions. We have a lot of the technical wiring in place to give us the data we need. It is mostly a matter of deciding what to look at and determining what it tells us. And most importantly, using it to take some action to improve our business.