Departments
Take 5
5 ways to get the most out of metrics
CTO Nick Ibrahim explores lessons learned at
Ruby Tuesday.
1. Invest in the right infrastructure.
To measure metrics in your business, you first must be able to get to the data. You need three foundational items to create the infrastructure necessary to create any metrics. The first is a logical data mart to get your data in a form that is easily accessed and optimized by users. Next, you need a database engine that is optimized for business intelligence [BI] and data processing. Finally, you need a BI tool with the expansion capabilities you require now and in the future.
2. Assemble a team of human BI resources.
You have tons of data, and you have to know what to go after for metrics. Therefore, you need people who understand both the tools and the business. Look for a hybrid group of operators who work in the field and people with IT knowledge. Bringing these two types of employees together will enable you to use their combined knowledge to develop the best metrics scenarios.
3. Select which metrics
to examine.
When you mine data, the tool is going to give you probably 20 different paths that you should measure. In our case, the algorithm may pick up on the fact that almost everybody who orders fries also orders water. But when we look at that result we know immediately they aren’t really connected. It takes a combination of tools and people resources to determine which metrics to examine.
4. Evaluate your results.
A metric is only as good as the results it brings. You have to go through the full cycle of the metric to see if it is really telling you what you need to know to make the most of your business. For example, is it bringing you closer to the real result of boosting sales, preventing loss or reducing labor costs? If a metric is taking you away from the real result, it should be adjusted.
5. Refine the process.
Metrics are tools to help enhance your business. But if one metric is wrong, it will feed the next equation wrong information, too. You must continuously fine-tune your metrics to make sure you minimize the margin of error.
Photography by Bard Wrisley