In the second of three articles focused on the changing role of L&D, Nigel Paine looks at big data
At the beginning of David Weinberger’s book Everything is Miscellaneous he asks us quite simply to:
“Suppose that now, for the first time in history, we are able to arrange concepts without the silent limitations of the physical. How might ideas, organisations, and knowledge itself change?”
We live in a world that is drowning in information. Ask anyone how they cope with the flood of email, references, papers, articles and notes that they have to read; coupled with the videos, podcasts, audio clips and other media that they are expected to consume and inevitably you end up with a rolling of the eyes! Everyone is overloaded by information and behind that information is an infinite amount of data that sits on servers and computers everywhere. For example, in one week, Amazon adds the same amount of storage capacity to its server farms as contained everything Amazon did five years ago.
Weinberger calls this “information sprawl” but claims that this as “the natural topology of the miscellaneous.” However, his final exhortation is that “we [must] see past its mess to its meaning”. In some ways, this article will try to help you see past the mess of big data to its potential meaning for those in learning.
What is big data?
The Harvard Business Review (HBR) published three separate articles on big data in October 2012. The editorial said: “Businesses are collecting more data than they know what to do with. To turn all this information into competitive gold, they’ll need new skills and a newmanagement style.”
Huge amounts of data exist. This is a byproduct of the internet and the digital revolution. Everything generates data. Every time one of the new Boeing DreamLiners docks at a gate for example, it downloads 50 MB of data to Boeing and the airline. What neither Boeing nor the airlines have worked out yet, is how much of that data is significant. For example, if the plane tells you that one of the brakes was 6°C hotter after landing than another of the brakes, is that something to be alarmed at and to take action, or simply to be noted? As the editor of HBR pointed out, “the trick is finding new ways to make [big data] work for you”.
The key to understanding big data is not to see it “as an indicator of quantity” but “an indicator of complexity”. It is not about a single source of data but the ability to extract meaning from many data streams that are woven together. This is not trivial but with the right tools the output is visual, straightforward and easy to understand.
10 lessons for those working in learning and development?
- Don’t ignore this. As the rest of the organisation gets interested, show that you are interested as well.
- You do not have to generate all of the data that is useful to you. It is perhaps the exact opposite. You have to get access to the stream of data generated by your organisation to see what makes sense for you and what helps you do your job more effectively.
- You need friends! [pullquote]Talk to the people doing this already inside your organisation. Work out what you can learn[/pullquote]. An Australian bank, for example, seconds staff from across the bank into its social media and analytics centre for two or three days in order for them to understand how non-financial data impacts on the day-to-day business of the bank. In this instance, make sure you have your staff being seconded as well as other staff across the organisation.
- Find out what other people are doing. An idea that works in one organisation could well be an idea that works in yours.
- You will need to buy in expertise in data analytics either from inside or outside your organisation. Think about this when building your strategy. Work out who could help.
- Start to examine the data that your own systems generate. Could you use that data more effectively? Could you weave the separate strands together to gain more insight and increase your alignment with the business?
- Get your team enthusiastic. Show that this isn’t a daunting, difficult and overly complex world just another window on reality that will help rather than hinder.
- Do not get hung up on data alone. Jeffrey Berk from KnowledgeAdvisors talks about “roughly reasonable data”. You need to gather sufficient data that allows you to make your point and draw your conclusions. What is roughly reasonable data? You should not answer that question. The business will tell you when you have sufficient information to draw conclusions that will be widely accepted. Total certainty is rarely required in these circumstances.
- If you don’t get involved, soon you will begin to look like a dinosaur. Databased decision-making is becoming increasingly important – do not be left behind.
- Big data does not replace instinct, conversations or existing knowledge. It complements them. Use this as one tool in your armoury and allow it to enrich what you do and the measurement of the impact of what you do. This is the key to success.
Three case studies
To talk in abstract about the power of big data drives everybody crazy. What does that mean? What does it mean for me? What point are you trying to make? I have a lot of sympathy for people frustrated by the supposed power of big data without having a clue what it is and how to make sense of it all. Here are three examples which will clarify what it is and what it means for learning and development.
Case study one
An insurance company was using a whole variety of data gathering to try to understand what worked and what didn’t work for its customers. This was driven by the marketing department who followed up on customer plaudits, customer complaints, comments from sales teams alongside postings on Twitter, Facebook, Instagram and other forms of social media. They crunched this data to expose trends, issues with products and gaps in the market. This is something that many companies are doing.
The head of learning asked to be briefed on this data and when she looked at it and discussed it with her team she was aware that there were some fairly blatant learning gaps in the sales team, and where the development that was delivered to them was not quite fit for purpose and did not align with some of the key concerns expressed by customers. Her programmes were modified and all the sales team were put through the new development opportunities over a three-month period. They then watched the increase in sales performance as a direct consequence.
The learning team kept on analysing this data and where new trends emerged or where customer opportunities presented themselves they offered bite-size insights to the sales team to help them develop appropriate techniques to respond. This new alignment between business performance and the development offered to the sales staff was genuinely transformational. Here is a learning and development team revising what they do; focusing on the emerging needs of the business and having a massively increased impact on that business without generating a single piece of new information. All they have done is tap into the massive, complex data stream that the company generates on a daily basis and interpreting it from the learning capability perspective. Big data in action.
Case study two
A finance company realised that it was generating vast amounts of performance data from its 360-degree appraisals; feedback meetings; project review cycles and manager interactions with their teams. Each of these pieces of data was gathered by a different part of the organisation, sometimes by an individual who stored the data on his or her own laptop, and used sporadically if at all. Certainly nobody was adding this data up and trying to see what it was telling them about the organisation as a whole, and the performance of its key staff.
Data analysts moved in and collected this information into a single data stream which was then turned into a dashboard for each manager. At a glance, on a single screen, managers could review the performance of their whole teams, or by drilling down, into each individual direct report. The managers were able to see where things were going well and where things were not going so well. That data that had previously been fragmented and hidden became a daily tool in managing and leading more effectively. The end result was higher performance and lower attrition. Managers said that that dashboard was the most useful tool they had at their disposal to be more effective in leading and motivating their teams. Each individual also has a performance tool that kept current performance at the forefront of their mind and allowed them to journal new events, ideas and reviews that would ultimately be seen by their manager. Better and more realistic
discussion ensued.
This company created a vibrant and constantly updated development centre that worked for the individual, his or her manager as well as the company a whole. Only by collecting, refining and then visualising data was this possible.
Case study three
Most of us build leadership development programmes in abstract. One large technology company decided that this unfocused targeting of generic leadership development was not having the impact they had hoped. It was expensive, and worked well for their best leaders. For the rest it was either ignored or seen as a step too far in their own development.
Instead, this company collated and processed all the performance review and appraisal data that they had at their disposal. It was made anonymous to protect personal data. This revealed a list of eight criteria that described what their best leaders did and what their lower performing leaders did not.
The focus for leadership development became a concentration on their lower performing leaders, encouraging them to do more of what their best performing leaders did and therefore raising the standard of leadership inside the company and eradicating the worst leadership practices.
They continued to collect all that data so they were able to gather first-hand evidence that the leadership programme was working and that poor leadership was being eliminated in the organisation. The best leaders were encouraged to continue a process of self-development and they were used as role models for everyone else.
Conclusion
These are three simple examples of how existing data streams can be used to redefine learning and development within an organisation. And that is without turning our attention to the kind of data that can be specifically gathered on performance that will enable all L&D programmes to be focused and targeted to deal with core organisational issues that builds the right kind of capability within the organisation.
The sea change that is happening in front of our eyes is a shift from increasing the efficiency of learning – i.e. cutting costs, modifying delivery systems, doing more for less – to increasing the effectiveness of learning i.e. ensuring that the investment delivers for the organisation.
Elliott Masie in his book Big Learning Data says:
“Our hope is that you will be active learners, evidence-based experimenters, and explorers of both the opportunities and challenges the big learning data presents to the learning and knowledge world.”
And then in the final paragraph of the book Bob Baker sums up:
“Each organisation will have a unique context, need, and capacity to address these questions – and others that you might come up with. Move forward – as learners, experimenters, and explorers – creating and shaping big learning data to create big, personalised, and effective learning for employees and organisations!”
There is no better exhortation and call to action than that.
Next month Nigel will be exploring technology as a game changer for L&D
A fully referenced version of this article is available on request