Big Data, Baby Steps: Organisational Buy-in Can Equal a Big Win.
Organisational leaders know they should be incorporating analytics in their operations but often find difficulty knowing where to start.
I’ve helped businesses and organisations of all sizes and across many sectors build analytics operations, and there is a consistent pattern: somebody with time and money hires a consultancy or asks in-house resources to mine an insight or two from organisational data. Those insights become a memo or slide deck presented along with a recommended action to the boss. Those insights are the first data spark. But, without leadership fuelling the fire – really the hardest part on the path to being more data-driven – the spark may extinguish.
As Data Manager for Barack Obama’s first presidential campaign, I was nestled in the IT department and responsible for care and feeding of our various databases. In June of 2007, the search engine marketing manager asked me if I could calculate total net contributions of donors who originally came in through an SEM placement. Looking at several separate databases, I reported our receipt of $165,134.08 from these people. Since far less than that amount had been spent on SEM up to that point, this spark of insight led to two fateful resourcing decisions by the campaign manager:
- Ramp up investment in SEM (and, ultimately, the whole New Media budget), and
- Pay some people to mine more data-based efficiency-finding insights.
In the above example, the key critical factor in determining the fate of the effort was buy-in from leadership. When the boss creates and enforces a mandate, the company will evolve around the data and realise benefits from becoming research-driven. Often, conflict between the established order and a new data-driven decision-making program will require more hands-on attention from leadership. But if an organisation is just getting started, analytics can become part of the established operational reality from the beginning. It’s not surprising that some of the biggest innovations around data-driven communications and marketing over the last decade have been brought into the mainstream by upstart, disruptive political campaigns. These ephemeral endeavours have nothing to lose and very little entrenched “wisdom” to protect: consumer segmentation and predictive modelling tactics were explored with great success by the George W. Bush campaigns of 2000 and 2004; A/B testing of e-mails helped drive fundraising numbers on Howard Dean’s 2004 campaign; real-time, in-cycle experiment informed programs were pushed and perfected by Barack Obama’s efforts in 2008 and 2012. The data and analytics operations embedded with these efforts shared certain characteristics: common mission alignment among staff, lack of entrenched opposition, high incentive to innovate, and low incentive to avoid risk. But on campaigns as with businesses and in non-profits, large-scale analytics operations begin as one person or one small team beginning to explore the quantitative possibilities of data they already work with on a daily basis, proving the concept and sparking the fire. The first step could be matching a customer list against social media to learn what customers pay attention to online; appending consumer data to accounting data in order to understand profit and loss (and opportunities) by customer gender, age, and income; or using e-mail subscription time-of-day to inform timing of marketing e-mail deliveries. Each of these bite-sized data research projects could yield small insights and in turn deliver significant bottom-line impact. Once the analytics seed is planted, the best first investment a company or organisation can make toward creating a successful analytics operation may not be in server hardware or consulting smarts, but in fostering a risk-tolerant atmosphere and creating an executive mandate for organisational adoption.