Welcome to the next article in a series on data analysis doubts. I’m tackling some of the biggest concerns leaders have about using analytics in their organizations. In this article, I want to address the question, “Will analytics slow down our decision-making?” I’ll also give you a structure for your analytics to get the right information at the right time.
Three Processes for Analytic Decision-Making
When it comes to analytics for decision-making, you need the right information at the right time. To get that information, you need a plan for addressing several specific processes.
First, you need a process to calculate and track key performance indicators (KPIs). Your KPIs are concrete metrics necessary to track your progress on strategic efforts.
Second, you need flexibility and capacity to handle ad hoc analyses. Your ad hoc analyses are required periodically to address short-term tactical analyses.
Third, you need data management to transform your raw data into something useful for analysis. Your data management processes ensure you have the most up-to-date data available as quickly as possible.
We ignore the data collection process in this article. If you aren’t already collecting the data, your decision-making will already be slowed considerably.
Key Performance Indicators for Strategic Decision-Making
If you Google the difference between KPIs and metrics, you’ll find some confusing explanations. So let me help clarify…
Metrics are measurements of a specific concept, typically using quantitative, numeric data. There are numerous metrics that can be captured depending on your business goals and processes.
Metrics can be used for several different purposes, including as KPIs. That’s right, a KPI is a particular way of using a metric.
Specifically, a KPI is a metric used to represent an element of performance strongly tied to your strategic goals. If you use a metric as a KPI, you should be able to identify a target level for the metric (e.g., increase sales revenue by 10%).
Across each of your different business processes, you should identify a few strategic KPIs. These metrics should be tied strongly to success according to your business goals.
Your analytics program should generate measurements of each KPI on a regular basis (i.e., daily, weekly, monthly, etc.). By measuring each KPI over time, you can track progress or deficiencies on each KPI.
Ad Hoc Analyses for Tactical Decision-Making
In contrast to tracking strategic performance, there are day-to-day decisions that will benefit from ad hoc, or one-off, analyses.
Because ad hoc analyses focus on tactical decisions, you will often need to assess metrics not captured in your KPIs.
For example, you might compare conversion rates on a specific platform to determine which social media ad to continue using. You may also want to know how the ad’s conversion rate varied across different demographic groups.
If you strongly depend on social media advertising, you might specify your global conversion rate as a KPI. The conversion rate for any one ad, however, provides tactical information to help you progress toward your target conversion rate.
In contrast to the conversion rate example, ad hoc analyses may be completely different from your strategic KPIs. For example, you might analyze customer data to identify the customers most likely to purchase a new product.
Your mix of purchases by customer demographics or behavior patterns may not be strategically interesting to your organization. However, customer purchase data is tactically relevant to developing the marketing campaign for the product.
Ad hoc analyses frequently rise unexpectedly and are not easy to plan for. Therefore, your analytics program should have some regular capacity, or slack, to allow quick responses to ad hoc analyses. Without this capacity, ad hoc decisions will be more difficult to make in a data-driven manner.
Data Management to Set the Stage
Whether your analysts are focused on generating KPIs or ad hoc analyses, they need good data to work with. In the ideal scenario, analysts would never need to clean and prepare data for analysis.
Unfortunately, this is unavoidable, so the analyst’s goal becomes developing a way to minimize the cleaning and prepping required for an analysis. Professional analytics programs often build systems to transform raw data into prepared data for analytics.
The analysts begin by identifying the most common cleaning steps and flexible data structures for analyses. Then they create automated processes and programs that update their analytic data sets periodically with fresh clean data.
In some instances, analysts will prepare data in multiple ways to make ad hoc analyses easier and faster.
Raw data is often messy and requires considerable work to prepare for analysis. With a solid data management plan, however, your team will be positioned to move quickly when needed.
Conclusion
By this point, I hope two things are clear about whether analytics will slow down your decision-making.
First, you should understand that a well-planned, resourced, and implemented analytics program will increase your ability to make timely decisions. You will optimize your program with proper planning, whether you have one analyst or one hundred. You can read more about this in my article on planning organization-wide analytics.
Second, you should also understand that standing up an analytics program doesn’t happen overnight. Your team will not hit the ground running on day one. Plan your analytics to build your program, even if you must wait to build capacity for some analyses. Plan logically and build strategically.
The benefits of a good analytic program will far outweigh the costs and time invested. In fact, over time the ROI increases as you build capacity, infrastructure, and a code base to work from. This is a case where the longer you use it, the better it gets.
Finally, if you want assistance setting up or expanding analytics on your team, email me at [email protected].
