As digital transformation takes hold across the business landscape, organisations increasingly turn to big data to harness information to identify new opportunities. That, in turn, leads to smarter business decisions, more efficient operations – and eventually, higher profits and happier customers.
The COO is the person ultimately responsible for driving value from any big data analytics resource, whatever its level of sophistication.
Across a region, the dynamics are complex: different countries have different legal, political, digital and social attitudes to capturing data, ranging from privacy concerns to the actual task of acquiring data in scalable and meaningful ways.
Companies in mature or developed markets can have a wealth of data and a history of analysis. Emerging or growth markets can often have little data or growth that simply outstrips an ability to analyse where opportunities might actually be.
In all cases, our job as regional COOs is to make big data analytics consistent, accessible, valuable, relevant, meaningful and affordable.
Intent and strategy
The fundamental decision needs to be whether the customer is at the centre of everything
As with every business initiative, intent and strategy set the context for any big data operation. The fundamental decision needs to be whether the customer is at the centre of everything, in the sense of supporting “customer adhesion” – nurturing existing customers by providing value based on their needs, rather than doing the bare minimum to stop them leaving. One strategic direction embraces the customer as a valued partner, the other tolerates them as a necessary evil. (Needless to say, our stance at Lenovo is to regard customers as valued partners.)
This affects how data are captured, analysed and used within the organisation. For example, the operational style of a call centre, how its team members are briefed and equipped with information, even the tone in which they engage with customers – in short, their levels of empowerment to act with the customers’ best interests at heart – will change depending on the strategic position. Answering customer questions at the least-possible cost is a very different strategy, and data analytics requirement, to engaging with customers to up-sell and nurture relationships.
If you want to work with your customers, the quality and timeliness of your data will need to be the best they can be – with consequences for investment, training and analytics tools, systems and processes.
The challenges of integration
Customer journeys, even for the same product set, differ in each country, as will how customers interact. What is familiar and acceptable in India will not be the case in Japan or Indonesia.
Integrating big data across multiple countries and markets within a region will likely be a challenge. Customer journeys, even for the same product set, differ in each country, as will how customers interact. What is familiar and acceptable in India will not be the case in Japan or Indonesia. Data acquisition methods will similarly differ – and both characteristics can compound each other when it comes to making sense of data.
Even the choice of database software and tools may differ between countries, although the COO needs to be clear that the benefits of having different tools outweigh the benefits of having a single implementation across the region.
The accuracy, detail and completeness (or coverage) of data will vary. In super-dynamic growth markets (which need not necessarily be emerging markets), growth can be so great that it either obliterates meaningful data or outstrips the data-gathering process. This will apply to politically-stable countries or markets. In those that are unstable, the difficulties are again compounded – data may be incomplete, destroyed or compromised. For example, data on market sizes based on census information will suffer if records are incomplete or destroyed for political reasons (or even because of bureaucratic slackness). If a political coup imposes a curfew that restricts late-night shopping, as another example, but does so only for two months, will companies factor this into their analyses more as an outlier than a spending trend?
And market-shaping events can simply reset everything: what were norms and trends before simply no longer apply in any statistically meaningful way
And market-shaping events, such as Indian demonetisation, can simply reset everything: what were norms and trends before simply no longer apply in any statistically meaningful way. The old data are simply irrelevant.
One option for managing this intra-region complexity is to apply different analytical systems and techniques to different data sets, either within a particular country or market or in different countries within a region. As analytics software tool company Fractal notes, change the rules to apply analytical rigour. Whatever approach you take at a regional level, investing in data analytics early in a region’s or market’s cycle will provide consistent benchmarks and potentially save money via a marketing or sales budget that might otherwise be squandered on non-existing opportunities.
Where data danger lurks
Beware a political or developmental bias in the data. I don’t agree that data are biased – but data sources can be, especially as economies start to flourish and evolve from past corporate, commercial or national contexts.
The value of data is also in the discontinuities that can become apparent, either to colleagues on the ground or to consultants. If data and analyses don’t match personal experiences, one or the other is clearly wrong, but now the alarm has been raised and more analysis can be done, and more data captured, to resolve the anomaly. Anna Rosenberg and Lauren Goodwin, both of Frontier Strategy Group and writing in the Harvard Business Review, cite the example of the alcoholic beverage company Beta, which used outdated data to assess the potential size of the market for its products in Iran, comparing it with India. Problem was, Iran has been ‘dry’ since the 1979 Iranian Revolution.
What to consider when considering big data
● At the regional level, one size does not fit all when it comes to data. Instead, adapt and adopt data acquisition methods suited to each market or country, but ensure that these methods deliver data that are consistent, complete and timely. However, while data collection methodologies could differ, there is also an uncanny similarity in key business metrics one needs to track for a customer segment or business model irrespective of country and region. The key is to leverage the commonality while respecting the source diversity
● Be certain about the context and impact of your data. Operations management in mature or developed market is around managing costs and efficiency. In growth or emerging markets, it’s about creating capacity and scale, resource deployment, and investment. You can only do all of these with the best-possible data you can either buy or determine yourself
● Invest in the skills needed to analyse and understand data, whatever its source
● Legal positions around compliance differ in different countries. Consulting firm EYadvocates a single, global policy. I’m less convinced – as I note above, one size does not fit every situation when it comes to data
● However, I am in agreement that one needs to try and create ‘one version of the truth’ at a global / geo level while leveraging common data points and vectors to build strong analytical or visualisation engines
● Does your existing organisational structure contain silos? Can data flow across these boundaries? Do you need to consider larger-scale changes to structure ahead of any big data investment? Do you already have the data you need, but simply can’t see it, locked away within silos?
● Are you comfortable with a ‘black-box’ approach to data analytics, with the expertise locked away with a group of in-house specialists or external consultants, or do you want to understand the analysis as part of the overall data reduction processes?
● Do you want this expertise in-house (to protect your IP, and to build institutional expertise within the organisation) or do you want to outsource to expert consultants (to manage costs)?
● Finally, once you have it, can you integrate big data into your business planning and strategy development? This is easier said than done. Many leaders are still happy to see data in older formats like spreadsheets rather than getting used to using visualisation tools. There are also doubts about the accuracy of predictive analytics, and while no one doubts its potential, one needs to still experience the reality to be an advocate.
All of which makes the task of a regional COO more challenging (didn’t I tell you I have a tough job!)
As with many options in business, simply starting has advantages. Consider a big data pilot, particularly to test some of the points made above. Can you audit data you already have? Can you compare data veracity across different countries or markets, perhaps using the same product in both as your benchmark? If you have data capture processes in place, can you make them more-systematic, and how might technology accelerate that?
This whole area offers huge possibilities and opportunities for anyone driving or supporting business operations. Everyone will be using big data and analytics in the near future. The early movers are going to be winners for sure. Lenovo is on the right path and I am personally excited by the steps we are taking. Please share your thoughts in the comments below, or connect with me on Twitter!
Find the original article here.