June 2016 Data & Analytics Share Group Recap
Data & Analytics Share Group: Building a Sustainable Analytics Capability
Sponsored by PwC Strategy&
On June 6, during the 2016 Consumer Goods Sales & Marketing Summit in New York, attendees of the CGT Data & Analytics Share Group met in NYC to discuss best practices around building a sustainable analytics capability. PwC Strategy&, the group’s sponsor, led the discussion and presented use cases and demonstrations of how analytics are being leveraged to tackle big strategic imperatives.
Organizing for Analytics Excellence
CG executives see the benefits of data-driven decision, but are finding it difficult to incorporate analytics capabilities, process, and professionals into existing organizations. Experienced companies have figured out how to combine art with science to truly leverage the potential of data and analytics.
Developing the right analytics capability requires a review across six dimensions. (see presentation) An organization's operating model should align with its strategic intent, and four governance models were explored (see presentation). Share Group members fit into one of these models or were at least beginning to organize in this manner; there were only a handful of companies represented, however, with a chief analytics officer.
Best practices were discussed, many around the early phases of projects. For many participants, the recognition that this initiative is a marathon and not a quick sprint helped to drive decisions. Others recommended showing "the art of the possible” early on, to prove capabilities and demonstrate quick wins.
Share Group members broke into smaller groups to discuss three different questions:
1 - What is your organization's operating model currently?
Models differed across participants, but there was general agreement that the models represented most CG companies. One company is still experiencing a large disconnect and they are working on getting to a centralized model, which will hopefully make them more efficient and cost functional. Challenges of centralization include:
- The group may be disconnected from the business
- Questions arise as to ownership
- Groups who are set up in a functional model didn’t feel the need to re-structure into a central model.
Discussions ensued regarding consolidation of data, particularly across regions and where cross-functional brand teams are in place. In summation, several organizations started in the functional model and are now moving to a hybrid model because of "necessity" or "pain." Some found a need to start to do things in a more coordinated fashion. All agreed, however, that there needs to be a clear message from the top on what's important and what is coming in the future.
2 - What analytics model would work best?
The answer to this question different greatly and was dependent on the company and the culture. Any operating model might work in a given organization with a different structure. Most agreed, however, that there are benefits for centrally managing the data itself.
3 - Which element of the analytics capability does your organization struggle with most?
There was consensus amongst participants that the biggest challenge involves culture. Some have capabilities they are happy with, but some feel that what they do is an art. Other challenges involve balancing the need for a short term win versus the longer term investment. Departmental silos and competition between departments were also identified as major hurdles.
Identifying the Big Imperatives
CPG companies usually have a long list of use cases across the organization. One example of a performance dashboard enabled drill-down capabilities to the assortment in a specific store. In another example, group members noted that sometimes departments are wary about offering data that might be used to measure performance against a previous benchmark.
Additional considerations:
- Don’t underestimate the training component during an analytics reorganization
- Consider where individual people/skills will best fit
- Consider who to keep and what skills can be outsourced
- A team approach is often best, pairing a data scientist with a business-savvy individual
- Most organizations over-invest in data governance and under-invest in what they can do with the data.
- The model should grow and evolve with periodic changes (some large and some minor tweaks)
Future Topics
In the future share groups, members would like to hear more about the consumer drivers in the case studies reviewed, and the group wanted more details into the attributes themselves. They would like to get more technical in opening up the black box and specifically which models worked in which use case.
*We’d love to hear from you! If you have questions about the share group in general, ideas for future discussion topics, or comments, please contact [email protected].