Analytics During Project Kick Off

Written by Ken Ono

Analytics During Project Kick Off

Building digital analytics features into web and mobile products is a small but strategically important set of activities when building out an MVP (minimum viable product) or rewriting a product. This blog post reviews where analytics planning intersects with other project kick off activities.

Preparing for the Kick Off

Much important information is gathered leading up to the project kick off. This could be in the sales process if a vendor is doing the development work or as part of the internal decision making in the case of an in-house project. Either way, prior to the formal project kick-off, it is good practice to review the project information from the perspective of those responsible for measuring the outcomes and running the data driven enhancement processes.

Some data points to look for are:

  • Why is the client/project sponsors embarking on the project and what measurable outcomes are implied?
  • What clarifications are needed to properly understand the business goals and how to measure the results?
  • What measurement defines the project success? (Number of users? Average user session time? Conversions? etc.)
  • What functional and nonfunctional analytics requirements have been identified leading up to the project?
  • What is the organization’s analytics maturity level and does the project represent an opportunity to improve data-driven processes?

Project Kick Off

Often the pressure is on to deliver the solution fast and the team and stakeholders are only thinking about user features. If analytics is presented for analytics sake, it will likely be deferred and the team will miss an opportunity to deliver higher value. However, if the team is correctly focused on outcomes (as they all should be), the need to measure these outcomes will become apparent to all.

There is a broad spectrum of capabilities across and even within organizations as it relates to leveraging the value in data. Assessing the stakeholder’s organization capabilities will allow the team to identify the optimal analytics activities during the project. For a detailed description of the Analytics Maturity Model click here.

Clarity Canvas

Clarity Canvas is Rangle’s project design discovery process that consistently defines the smallest version of the product that will most effectively progress our clients towards their goals. For details on clarity canvas see Project Discovery in a Lean Agile World: Introducing the Clarity Canvas.

Clarity Canvas has a focus on goals and outcomes and therefore is a natural way to establish understanding of the business objectives that will be needed to implement analytics. When the discussion turns to risky assumptions, look for opportunities where observation can confirm or refute assumptions.

Clarity Canvas should be high level and business oriented. The analytics tools and precise KPI definitions don’t matter at this stage. Some high level triangulation on analytics scope can be performed. For example, what requirements are there for personalization of the web site.

Set Up a Measurement Plan

If analytics will be an important part of the project (and this should be the norm), set aside time to do a measurement plan early on in the project.

Regardless of where the analytics resource comes from (customer side, vendor side), the analytics specialist who will run the measurement plan should be introduced to the project and the activity should be scheduled.

The senior stakeholders who need to be interviewed for the measurement plan should be identified. In complex organizations, there may be multiple stakeholders that need to be interviewed and the measurement plan will take longer.

Often due to scheduling issues, getting all the right stakeholders involved during the kick off is problematic. The measurement plan offers the team the opportunity to go back to senior stakeholders to refine the business objectives.

Analytics as Part of the Technology Plan

There is a huge number of click stream analytics tooling to choose from but in most enterprise projects, the tooling choices are pre-established. During the kick off, find out what the tooling is and who owns the data model. During the project, the team will need access to the analytics accounts and establishing a test and deployment strategy will require some coordination.

There are different types of analytics tools. For example:

  • The analytics engine. Google Analytics or Adobe Analytics are common and there are many other tools offering competing or additional capabilities.
  • The tag manager. Tag managers insert the analytics integration logic as the app is running by manipulating the DOM. The principle benefit of this approach is that non-developers can usually implement the analytics without a developer’s assistance. Some tools combine tag management and analytics so they can automatically deliver features like field level analytics or replaying customer interactions.
  • BI big data tools. Enterprises will often have established data warehousing or data lake strategies for capturing customer and prospect data.

What the team is comfortable with is often an important consideration on the tooling. If analysts will be working independently from developers, then using tag managers is a natural choice. If the developers will be responsible for integrating with the analytics, then a JavaScript centric approach, such as redux-beacon, may be warranted. See Two Good Approaches for Implementing Analytics Tags for a more detailed discussion.

Tooling and teams change, so the architects should consider approaches that provide maximum flexibility without over designing. For example, the initial instrumentation could be done in JavaScript with redux-beacon but later, new analytics targets can be configured.

Learn about the organization’s plans to build a single view of customers and prospects. Behavior on web and mobile is often just a part of the customer journey. What are the requirements to integrate with the organization's BI and big data strategies?

Set Up Analytics in the Quality Plan

The analytics footprint in the quality plan can range from trivial to foundational. For example, in a project that is delivering a new application without the expectation of a large number of users, the only analytics-related activity in the quality plan may be verifying that basic navigation is tracked.

On the other hand, a new version of the app may have specific goals in improving a metric such as task completion rate or time on task. In this case, implementing the KPIs are necessary to execute the quality plan and the quality plan and measure plan are intrinsically linked.

During the project kick off, ensure that the measurement plan and quality plan are coordinated where they need to be.

SEO Plan

Many application development projects will not be concerned with Search Engine Optimization (SEO). For example, if the app is for internal users or there is a related web site that attracts visitors to the app SEO will rightly be out of scope of the project.

When the app should be indexed by search engines, it is important that this requirement be identified early as it may require the inclusion of tasks such a server side rendering and an SEO audit to ensure all the links within the app are followed by spiders.

Prepare for Build, Measure, Learn

When real users begin to use the app, development enters a new phase of experimentation. Being able to quickly get improvements into production is vital for success. However, in many enterprise contexts, it is very difficult due to existing processes. When building the MVP or replatforming an existing product is an ideal time to assess the organization’s current capabilities and start the work of accelerating cycle times.

A BML readiness assessment will identify any issues that may negatively impact cycle times and establish a prioritized action plan to address them.

Executing a BML readiness assessment involves reviewing the functional and technology plans, reviewing the current deployment and testing processes and interviewing the key stakeholders.

Benefits of a faster innovation cycle time include:

  • Improved product market fit
  • Increased business value being generated by your property
  • Higher quality product and faster resolution of problems
  • Having hard data on devices, users and impact of experiments
  • Less stressful environment and high staff retention and productivity
  • A focused decision making process to create the right portfolio of innovation

Conclusion

Appropriately integrating analytics into the kick off process will increase the value delivered by the entire project team. During the initial build, analytics scope is important but small. Going live as soon as possible allows data-driven decisions to be made earlier and reduces the risk of developing something that isn’t needed. As real users begin using the app, the analytics scope increases and becomes an integral part of the development process.

Analytics During Project Kick Off