Both large and small analytics projects have their place in organizations. The current environment within an organization can influence whether one approach or the other is the best method for an early analytics project. Knowing whether the business is in a busy time period, open to new ideas, or in a mode where savings are needed to be delivered by end of year can all be reasons to choose one project size over another. The skill level of the person running the project is important as well because an analyst trying to run a project versus an executive or IT can create different dynamics for the project.
No matter the size of the project, resources will be needed in both time and effort at a minimum. A smaller project has the benefit of requiring less resources and less risk if the project fails and can be run by any level of individual. On the other hand, a larger project can gain efficiencies by pulling in more resources and using them in a more optimal way to extract value but require executive level buy-in to be successful.
Due to the increase in resources, larger projects will also create more areas of further analytic avenues by having deeper data penetration, as well providing more organizational change to ensure the project continues to drive success. However, with less resources needs and less complexity to manage, smaller projects can be easier and quicker to accomplish, delivering results in a much shorter time-frame.
Less complex smaller projects are also able to deliver results that are more customized to their problem focus. Larger projects can often get bogged down in the complexities involved in a comprehensive solution that can make them take so long that the project is no longer relevant, or that the standard business use case has changed since the initial ideation phase.
Large projects do deliver amazing results and are much more visible in an organization though. These win-stories can be critical for management to gain recognition, but also can be used as a marketing tool externally as well. Smaller projects typically deliver smaller results but can be a valuable proof of concept or direction for further path refinement. Many organizational leaders need something proved to them before they will approve large expenditures of time and capital, so a smaller project can be useful in those regards.
In the situations I have encountered, I would typically recommend starting on a smaller project to prove the concept, as well as learn the hurdles and issues, in a lower risk environment that can then easily be applied to larger projects later. Once executives get a taste of the wins and efficiencies, they are more likely to want to move forward on larger projects that solve bigger issues that the small project uncovered. They are also more likely to then understand some of the potential options for an analytics project that would be larger. I have personally also encountered rejection of even a small project. With a smaller project you can still work on it on the side if it benefits your day to day work and once it is completed it can be used to showcase a process efficiency even if originally denied (tread carefully). Although starting with a large project can deliver great results, only organizations in a certain position would be able to implement the project optimally so unless the organization is hungry for analytic growth I recommend starting small.
Author: Logan Callen
Covington, Daniel. 2016. Analytics: Data Science, Data Analysis and Predictive Analytics for Business. 5th Edition. South Carolina: CreateSpace.
Harvard Business Review Press. 2018. HBR Guide to Data Analytics Basics for Managers. Boston: Harvard Business Press.