Some thoughts on the data conundrum

My earliest memory of consciously interacting with and relying on datasets professionally takes us back to 2012 when I was working in the food retail industry. The two key solutions I recall working with initially were a Point of Sale (POS) and Excel. The POS solution was critical to our ability to serve customers in a manner that was standardised so as to ensure that operationally we could achieve efficiencies and consistency where possible. Excel was a solution I started to use because the business had not adopted adequate data analytics solutions for people in my position - so before I knew it, I was learning how to do basic spreadsheet programming to create workbooks that could accommodate my data needs (I.e. data driven labour allocation proportionate to business performance and benchmarked against KPIs).

It didn’t take me long to become obsessed with the insights that can be derived from data, especially numerical data. I also quickly fell in love with having the ability to look at an operational activity or occurrence and knowing as precisely as possible in my head how that particular activity or occurrence impacted the business from the perspective of a spreadsheet or report.

Rather quickly, I learned that numerical data can tell stories.

As my career in the food retail industry progressed, I continued to develop my approaches to operational analysis by embracing a continuous improvement mindset. I was always asking questions such as:

“Why does this data matter?”

“How does this data influence this other data?”

“What does this numerical data tangibly represent for the operations of the business?”

Upon commencing work in the food retail industry, I was always told by experienced colleagues (who had been in the industry for far longer than I) “Labour as a percentage of gross turnover should be 30%, and so should COGS”.

Overtime I started to realise that the original rationale for such ‘industry norms’ could not be recovered. Whenever I’d ask, I was often informed “it just should be [that way]”.

So curiosity got the better of me and I started to invest time into developing theoretical standards for the operating model of this business. Specifically in relation to the “labour as a percentage” debate, I took the contentious approach of ignoring all of the industry standard KPIs I’d been informed of and decided to take [what I described as] a bottom up approach [contrary to top down; based on experienced egos]. In short, this approach involved me comprehensively analysing the theoretical operational requirements of labour on a per unit production basis. The short story is that this approach revealed, theoretically, that the labour demands of the operational model were equal to circa 4% of gross turnover. I was absolutely gobsmacked.

Particularly for those who believe the discipline of economics, for example, is garbage given it is largely reliant on theoretical models, please don’t race to the defense of “but it’s theoretical, and the real world isn’t theoretical Sally” - you’d be preaching to the converted. Plus, I’m not interested in debating the real world value of theory VS practical - let’s do that some other time.

If we go back to my theoretical discovery for a second, can you imagine finding yourself in a position where you work in an industry that you believe you have just realised is relying, operating and reporting on a completely flawed KPI? To clarify, the industry standard approach to this KPI was (and probably still is to be honest) that “it should be 30%” and I determined that for one [common] operating model in the industry the theoretical demand for that business metric was about 24% lower than what the industry was reporting “it is”.

This wasn’t a matter of me believing I was right and the industry was wrong. What this situation became, very quickly, was an absolute wake up call to myself (in particular) that how we interact with data is absolutely critical if we want to make decisions that are truly and reliably considerate to data. It dawned on me that the data immaturity of the industry, generally speaking, was at the helm of the self destruction of the industry.

I started to ponder “if an entire industry is establishing KPI benchmarks for numerical data without a fact-based approach to doing so; what does this mean for the longevity of this industry?”.

Almost immediately I realised that the capability of data to influence perceptions, beliefs and decisions was a capability that should not be ignored or exploited negligently.

This conversation requires a substantial amount of time and contextual consideration to adequately arrive at any sort of conclusion(s); the reason I am sharing this story is because of its impact on the overarching narrative of this article which relates to the multifaceted data conundrum. In the interests of time I will not endeavour to tackle every facet of the data conundrum conversation.

This experience in the food retail industry made me fall in love with data and taught me that a multifaceted data conundrum exists so it is absolutely critical for us to continuously seek to mature our approach to data (as individuals and businesses).

Different types of data, such as numerical and non-numerical data, are increasingly being enhanced by technology enabled processes which means that our ability to interact with and use such data is advancing. This is creating an increasing amount of overwhelm for businesses, hence the acceleration of debates relating to ‘effective data management’ and ‘data maturity’.

Innovators and practitioners are striving to enhance our human ability to use data to support effective decision making across myriad industries and disciplines.

I believe it is becoming increasingly critical that we enhance our capability to gather, analyse, interpret and respond to non-numerical data, such as written text. The democratisation of information distribution combined with competing algorithm biases has made communications networks extremely complex (hence the growing governance debates). As a passionate marketing and communications consultant, I’m excited about the growing number of tools that enable media and communications teams to equip themselves with the assistance they need to use big and complex datasets to enhance their work.

In particular, the ability to visualise stakeholder and narrative networks is becoming absolutely critical to be able to reveal conditions favourable for possible accumulation of strategic communications advantages in the public domain. It is only by visualising copious amounts of hand picked data (from relevant datasets) and revealing the relevant intelligence that media and communications teams can begin to more precisely and effectively enter important discussions in the public domain.

Finding the signal in the noise has never been more important to being able to cut through the noise when it comes to marketing and communications.

If you’d like to discover how we can work with you to reinvent your approach to marketing, media and communications - reach out any time, [email protected]

Best,

Sally A Illingworth