In mid-2018, I had just taken over the responsibility of embedding analytics, as a way of working, in the Asia operations of a Fortune 500 company. I had previously a good mix of insights and analytics experience, in both agency and client settings, and was part of this organization for a few years.
We planned to standardize, automate and expedite the way in which syndicated marketing research information was being delivered across the continent. We also envisaged an overarching analytics project, that models data from various sources to deliver business KPIs. The great opportunity was also to tap into the already available, but unutilised sources of data.
We met with some success in the first year, deriving insights from an existing, extremely granular, but not easily accessible data source. This led to insights into behaviour patterns of sales teams. We were also able to embed a standardized system of monthly KPIs across the diverse markets in Asia. This was the start of the journey of change of mindset from process focus to analytics focus.
Unlike the title, this change was not a breezy one, but having gone through the process of bringing about this mindset change, I wanted to share a few things that might make it breezy(er) for you.
Need a senior sponsor within the organisation
The first and arguably most important thing we had on our side was a corporate sponsor who was considered an expert on data and analytics. He was convinced of the importance of analytics, and shared our vision. He also suggested a few modifications which made our plan stronger and more “marketable” with the broader leadership team. He opened doors, connecting us with key people within the organization (who may not have otherwise given us the time of day). Most importantly, his blessings allowed us to overcome occasional barriers, encountered in this mission.
Have a lean plan and think agile
Secondly, we approached the project with a very lean plan. The vision and objectives were described in broad strokes, covering only a few slides. We had visualized the stages leading to the objective, but did not spend time and effort defining the micro steps within each stage. This was truly an agile and unconventional approach, especially for an initiative as ambitious as this. Instead of detailed plans, we had pilot projects planned for each stage. These pilot projects would target a medium sized country and thus a relatively small dataset. But, it would always address the actual business need for the end market.
These pilot projects turned out to be a great device. They built our learning, established working relationships with the agencies, and served as a proof of concept. They also justified the investment for the analytics roll out. The projects planned were small enough that if we made mistakes, it would be easy to course correct. In all, it ensured regular pay offs instead of one big payoff at the very end.
‘Go local’ and think like an entrepreneur
Thirdly, an entrepreneurial mindset. We tried different methodologies on a small scale, in specific markets, before rolling it out on a large scale. We ran these studies simultaneously, with different vendors, creating some degree of controlled chaos. This was also necessary because Asia has seen a rise in the number of startup agencies and vendors with analytics capabilities, and we needed to identify the ones that could deliver. The agencies varied by their technical capabilities and tools, the ability to find human centric insights from the big data, and their reach (portability) across Asia. So we classified them accordingly, and assigned further projects.
Keeping our local stakeholders engaged was key in retaining their support for a system, that questioned the “that’s how we do it” mindset. We kept our local Commercial and Insights teams regularly updated and actively involved them in local execution. Although the project was run at a regional level, it needed to stand on its own legs in each end market. These interactions helped us understand local practices and avoid misinterpretations. For example, while analysing sell-in data, we noticed dips in sell-in data on Friday and spikes on Saturday. On speaking to the local teams, we realized that some of the sales team entered the data over the weekend, instead of on Friday (who wants to spend Friday night entering data, anyway?).
Think ‘human business’
Finally, we paid particular attention to the human aspect of business interactions. The most important learning was that you cannot change mindset and ways of working by cascading a process top down. We need to identify and convince key people, at pivotal moments. We took the time and made the effort to do this, resulting in entire teams, or country organizations, accepting the new way of working very quickly. In some cases after a good “discussion” we even modified our initial plans. While this set us back, and resulted in a scramble to catch up at our end, it was vital for the initiative to be accepted and successful.
Takeaways
In retrospect, it is relatively easy to recount what worked, but I have to admit to multiple lessons learnt in the journey.
We realized the importance of having specialized functions even within analytics. For example, the data scientist was often not the best person to share the findings with the commercial team. A “translator” is needed for this.
We also realized the importance of seeking market context from people on the ground at the start of the project. While we collaborated once the project got underway, we would have benefited from knowing the Go-to-Market plans and salesman incentive mechanisms while designing the study itself.
Finally, as a parting thought, I wanted to share an insight that I gained through this process, that unfortunately I could not put into practice (as yet). The analytics journey for any organization is a long and arduous one. So, it is much better traversed with a partner organization in similar stage of the journey. A non-competing, friendly one with a similar organizational value may be a great companion for the journey. If not resources, at least knowledge and experience can be shared.
