Limitations of Key Metrics in Sprk™

Reading Time:

Learn about the potential limitations in the key metrics in Sprk™, as well as how to make good decisions despite these limitations.

All data has flaws and limitations. Most analytics tools ignore these flaws or try to hide them.

But Sprk™ is a different kind of product from a different kind of company. We believe in full transparency to help ensure you never make a bad decision because of flawed data.

Great efforts have been made so that every metric is accurate, complete and provided in full context. If you ever find an example where this may not be true, please contact us immediately at support@getasprk.com and we’ll make it our top priority to research and resolve the issue.

Here are three metrics that have some limitations or nuances:

Tipping Rate

Since Sprk™ does not know what cash tips (if any) are collected by your business, Tipping Rate is calculated by dividing non-cash tips (i.e., those tracked by your POS system) by the non-cash sales (i.e., orders placed using credit cards, gift cards, Apple Pay, etc., where the customer has the option to offer a non-cash tip).

This apples-to-apples comparison should be directionally correct given that cash tips and credit card tips tend to be correlated, but not if your customers tend to be much more (or much less) generous with cash tips than credit card tips.

This metric is most useful when looking at trends over time. For example, a noticeable decline in the Tipping Rate could indicate that your customer service is slipping. In this example, it’s likely a decline in credit card tips is accompanied by a similar decline in cash tips.

Repeat Customers / Frequent Customers

A similar problem exists in tracking repeat customers. POS systems are able to uniquely identify some customers based on the credit card used in the transactions, as well as additional information they may collect. The limitation is that not all customers are tracked, such as those paying with cash. In addition, a customer that uses a different credit card than a previous purchase could be counted as two distinct customers.

To minimize the impact of this, the Repeat Customers metric is calculated by dividing the number of tracked customers that placed more than one order by the total number of tracked customers. As such, customers using cash are completely excluded from this metric. The same calculation is done for the Frequent Customers metric, except that it uses customers that placed 10 or more orders (rather than 2+ orders for the Repeat Customers metric).

As with the Tipping Rate metric, these customer loyalty metrics are valid as long as your loyal customers tend to use cash and credit cards with the same frequency as the non-loyal customers. For example, if you observe that your most loyal customers always use cash and never use credit cards, this metric should be discounted or ignored. This is typically not the case, but every business is different.

Next Steps

The underlying principle of performance analytics is to always focus on making the best-informed decisions. If you have any questions or disagreements with any of these approaches, please reach out at support@getasprk.com.