Data Demystified, Part 2: Setting your data strategy

Part 2: The (non-technical) business leader’s guide to setting a data analytics and BI strategy

Data Demystified, Part 2: Setting your data strategy

Whether you’re just getting started, inheriting data responsibilities, or looking to improve an existing program, the first step is to plan out your broader organizational data strategy.

The underlying foundation of your data strategy is critical, because it’s the first step toward producing clean, useful, and impactful data that people throughout the organization can trust to make sound decisions. If you don’t lay the groundwork it’s impossible to successfully operate a data science function. 

The first order of business is to evaluate where you’re at by exploring questions like:

  • What are the most important questions we need to answer to understand how the business is performing? What data assets do we need to answer those questions? 
  • What’s been built so far, if anything? Are you being asked to pick up where someone left off or will you need to build out the infrastructure completely from scratch?
  • What’s the current state of data in your company? Are business stakeholders (e.g., Product, Operations, Marketing, Sales, Finance, Execs) who need data to run their day to day operations satisfied with the quality of the data and analytics they have access to?
  • How do business teams currently pull data and get metrics? Do they use a self-service platform, ask data engineers or analysts for one-off requests, or some combination of the two?
  • Are your stakeholders proficient with data? If not, are they ready to level-up and learn how to leverage data more efficiently in their roles? 
  • Or, will the data team need to be full-service (ie- building datasets and teaching people how to use them)?
  • Do you plan to “build or buy”? This question refers to the initial human capital that’s needed to get the data and BI function up and running. With the “build” option, you’d need an internal team with advanced technical skills. With the “buy” option, you’d purchase professional support services alongside BI tools.
  • What’s your budget - both for technology and headcount/services? 

Don’t hesitate to collaborate with other leaders in the organization during this process – their input will be invaluable. Approach mid-level and senior-level colleagues and ask:

  • What specific metrics do you look at?
  • What tools do you use to get those metrics?
  • How long does it take to get those metrics?
  • Who on your team is working with data?
  • What data would you like to have that you don't have today and what value would having these create in your organization?
  • How much trust do you put in the accuracy of the metrics you are accessing?

The answers to these questions will be eye-opening and you’ll quickly get a sense of the current state of data in your organization.

Determining initial use cases

Once you’ve set the stage for your data strategy, choose one or two data use cases that if resolved, would have a high ROI impact on the business. Think about the business questions you’d like to answer and the metrics you need to answer them. 

Think about what type of datasets will provide valuable business insights, not just in the immediate future, but also years down the line: 

  • What metrics are most important right now? If you’re trying to raise funding, for example, data on customer acquisition, retention, and expansion will be crucial.
  • What business functions do you need to measure the performance of? 
  • What types of questions are people within the business most likely to ask?
  • How detailed are those questions?
  • What factors will you need to slice and aggregate data by? For example, location, product, customer type, and any other important factors?
  • What teams or leaders are currently unable to make timely or accurate decisions because of poor quality or slow data?

Answering questions like these will help you choose the initial data opportunities and data sets that will drive the initial prioritization. Keep in mind that there’s no one-size-fits-all package of data use cases that will be valuable to every organization, but here are some options to consider:

  • Quarterly / Annual Board reporting
  • Weekly and monthly business reviews, both for the executive level and departmental/functional level
  • Team or functional business reporting (e.g., customer support metrics or user engagement metrics)
  • Analyzing and optimizing different stages of the customer journey (e.g., acquisition, activation, retention, referral)  
  • Objectives and key results (OKRs aren’t something that every company sets, however, if your company uses or plans to use OKRs, they can be a helpful initial use case to focus on because they require extensive reporting work)

To start, we recommend you choose something high-level that can be easily verified (or at least sense-checked) by other teams and systems. This approach will help build confidence from the beginning and set the stage for future successes.

Think about the low hanging fruit – an area where you can get data from a self-contained and well-organized system to build out an analytics use case. Zendesk customer operations data is one good example because you can easily test whether data is flowing accurately by comparing a native Zendesk report to the report created through your analytics stack. Something like marketing and sales attribution can be trickier, because the modern buying journey is increasingly complex and it’s not always easy to pinpoint a single conversion event, not to mention the data volumes are typically significantly larger.

This is the second article in our Data Demystified series, where you’ll get a clear roadmap to use when setting up a data analytics and BI program. If you missed the first post on Getting Started With Running Business Intelligence and Data Analytics, be sure to check it out. 

In the next post, we’ll explain How to Build a Modern Data Stack

Published on Aug 09, 2022 by AirOps Team

More from the AirOps Blog