When to Integrate Analytics into Development Projects
A common scenario is to tack implementing digital analytics onto the end of a project at the last minute. While this is much better than flying blind, better outcomes can be obtained by spending that time earlier in the project.
- Providing hard data on what features are actually being used and what features are not. This arms the product owner with critical data used to prioritize the backlog
- Providing hard data on the business value the application is generating which is vital for stakeholders to make investment decisions
- Designers can gain important insights about the usability of the app. For example, monitoring user interface errors can provide insight into where users are getting tripped up
- Enabling experimental processes to make improvements to the app once it is launched
At the early stages of a project, the team is often more concerned with the volume of development work ahead of them, delivering the solution in a timely manner and the many decisions that need to be made throughout the project. As a result, the analytics requirements are often deferred.
However, taking a little time establishing analytics up front pays dividends.
A key planning activity and asset is a measurement plan. For more details on creating a measurement plan, you can read Digital Marketing and Measurement Model by Avinash Kaushik’s, How to Create a Measurement Plan and Why You Really Need One and the basics from the Google Analytics Academy.
Briefly, a measurement plan:
- Identifies the business objectives through discovery with senior stakeholders
- Identifies goals for each business objective
- Defines the key performance indicators (KPIs)
- Define parameters for success by identifying targets for each KPI
- Identify types of people (segments) and related desired behaviors and outcomes
The team should have a good understanding of the measurement plan early on in the project.
This will lead to enhanced alignment because knowing the key success metrics will crystallize the business value that the app is trying to deliver. The team will be better able to reason about features and priorities armed with this clarity.
Understanding the KPI’s may also have an impact on the designs the team chooses. For example, if engagement is an important goal, a design that requires users to flip pages or uses a “Read More” expander button could generate much more meaningful data than an app where the user scrolls and this behavior is not captured. Consider how the team may make different decisions if they were aware that the success of the project depended on any of these KPI’s or segments:
- Number of display ads clicked
- Average % of each article read
of answers found
of social media shares
- Net promoter score
- Return visits per month
- Seniors, tweens, newcomers to Canada
- Daily users, first time users
- Mobile users, desktop users
Knowing the data that must be emitted by the app may also impact the technical designs. For example, if time on task must be calculated this may influence the services that need to be developed and the design of reusable component libraries.
In summary, understanding the business goals and how the business goals will be measured are foundational to the design of a system. Better outcomes will be delivered with an appropriate amount of upfront discovery.