In March 2020, just before the world went into lockdown, David James spoke with Trish Uhl (Data Science, AI & Advanced Analytics expert) on The Learning & Development Podcast.
This conversation highlighted where L&D was comfortable in dealing with data but also a huge piece of the puzzle that was largely neglected by our profession.
Read on for some of the insights Trish shared in that conversation, or tune in to episode 39 of The L&D Podcast for the whole conversation.
Gartner calls it being a Citizen Data Scientist. That’s somebody who is able to work with those who are innumerate and bring their domain knowledge in order to be able to help Data Scientists and Statisticians actually frame a particular hypothesis or problem and then using analytical output and numbers in order to solve it.
The first thing about the practice is, it doesn’t necessarily mean that we have to go from being people people to suddenly being really good with numbers. We can play in this realm with a Citizen Data Scientist. The second thing is it’s about qualitative data as much as it is about quantitative data. We often get persuaded or seduced by numbers. There are a lot of reasons why that happens. With humans across the board, if we’re going to get into things like predictive analytics, and we are, we can get there through qualitative data over quantitative data and a lot of people don’t know that in L&D.
The third thing is Formative Data. We have a tendency to look at data as a retrospective, looking back. Did we make an impact or not? We’re used to Summative Evaluation. We get to the end of a pilot, we get to the end of a Leadership Development programme and we do our Summative Evaluation. But the way we can really leverage data is formative. It’s as we go, it’s ongoing, so that we can influence the outcomes and the results.
We’re over that tipping point where this is no longer about aspiration. This is about expectation. In order to stay in the game of L&D, this is a capability that as an individual, we must develop and as a function, we must lead in our organisations.
This isn’t a wait-and-see opportunity anymore. The train has already left the station, it’s time to jump on board. What’s happening now in business and with organisations, whether it’s for-profit or non-profit around the world, is that we’re moving. We’re actually in the largest IT migration ever in the history of humankind, which means that we’re moving from information processing to cognitive computing.
What that means is, we’re moving from just being able to capture in-store and report on information to actually being able to leverage that information or that data, for intelligence, for organisational intelligence to be able to anticipate. How do we come up with algorithms? How do we use operational data? How do we use data that’s external to our organisation? How do we use real-time data in order to be able to have this operational agility? That comes into this intelligence paradigm that’s coming into our organisations. And it’s not only coming into our organisations, it’s coming into our homes. We’ve got video now at our doorbell. Somebody comes to the front stoop, rings the doorbell and there is a camera that’s recording everything that happens on the front stoop. We’re starting to put intelligence into our household, and we’re using that to make decisions about how we manage our households, how we operate and integrate in our neighbourhoods.
That same kind of intelligence, at that individual and social network level, is coming into our organisations. That’s what this whole thing about intelligence and cognitive computing is all about. Data and analytics is one component of a larger system that is setting the new framework in order to drive that intelligent capability.
It used to be in an old bureaucratic design model, which Trish understands many organisations still follow:
“We had decisions that were made at the top of the structure. We expected executives and leadership to be able to make decisions. Now, we’re pushing decision capability down to the point of work, we’re pushing decisions down to the frontline because organisations need to be able to move at the speed of change that’s happening within the markets that they operate, within the competitive landscape that they operate, that happens according to what’s happening with customer and client demands.”
That means that in order for L&D to support the organisation, which at the end of the day is our purpose, not to create learning or to have people learn something, but to support the organisation and its people in executing against its vision and mission and strategic objectives. Then in order to be aligned with that, we need to have the matching capability in order to facilitate that level of agility.
Any kind of analytics project, regardless of the domain, unless you’re actually using the data to translate that into insights, is not enough. You have to take insights and move that into action. To do that often takes an organisation a few cycles of Change Management in order to socialise that capability within an organisation. Most humans are not used to making informed decisions based on data.
That is a learning curve that many people from the frontline to the C-suite even need to learn within an organisation. That’s not unique to L&D. What Trish means by that is regardless of what you’re putting the analytical lens on – it could be operational analytics, business analytics, learning analytics, or people analytics – analytics methodology is exactly the same regardless of the particular domain.
What changes is the domain and that’s where, again, people are a little put off by the idea of data and put off by the idea of numbers and spreadsheets that are suddenly going to be in their lives. But the math is only really 20% of an analytics project. 80% is coming up with the right questions of value, hypotheses, identifying factors.
Then taking the analytical output, it’s the bookends. It’s the front end piece: What are we using data to solve? What threat or opportunity are we trying to address through data?
Then once we have that analytical output: How are we socialising that in an organisation to make sure people are using that analytical information in order to take action? Because you’re not done with an analytics project until you take action. We can be a Citizen Data Scientist. We can use our domain knowledge of learning and people and be able to then work with those. If we’re not completely numerate ourselves, we can actually complement our practice by working with those who are.
It was the systematic approach of being able to get all of the requirements upfront in order to be able to define a solution.
Since we’d do all of this initial assessment upfront, by the time we created the thing, we could have some level of confidence that we’d done due diligence and that thing would not only resolve whatever the issue was, but it would resolve that issue over a long period of time. We have a long history in L&D of wanting to come up with a solution that’s right. That the solution we came up with is the right solution to solve that problem. The challenge is, we don’t live in that world anymore.
That environment was static and routine. Where you could take weeks or months to come up with the right solution. Now, you can come up with the right solution on Monday morning, and it’s changed by lunchtime. The challenge has changed and the priority has changed. That Waterfall approach doesn’t work as well in this fast-moving business environment.
Find out more in episode 39 of The L&D Podcast.
We have to be willing to be wrong and we have to be willing to take the hypothesis or the solution that we think is right and make a small bet upfront and experiment to test and see if we’re on the right track or not. We have to be open and willing to take an action based on what that initial pulse check or that initial test comes back as. If it comes back and it tells us we’re off-track, we have to be willing to either tweak the solution or even eliminate it and start again.
The biggest thing in developing analytical capability is that paradigm shift in that mindset of being willing to be wrong and being okay with failing. Then learning from those failures and being able to iterate and improve or even when we need to, to start again.
That whole idea of being able to drop half a million dollars a year in a multiyear contract with ‘insert technology provider here’, is also at an end.
We can tell through business productivity tools, and not just L&D, and look at the data that’s flowing through our standard common, ubiquitous business productivity tools and be able to see how well we’re onboarding our managers or not. What Trish means specifically by that is, we can actually now see that in Office 365 data.
Microsoft, as a company, has a whole bench that is now dedicated to coming into organisations and meeting with executives and with business leaders in a line of business, and showing them how to leverage the data that they already create as a byproduct through daily use of their subscription software, in order to be able to address human performance issues, including and one of the big places that they’re able to leverage the data is by being able to take a look at how well we’re supporting our managers or not.
One of the things Trish says to L&D people all the time is, we are Grumpy Cat over the fact that line of business or stakeholders come to us and they go:
“Hey, we don’t mean to treat you as an order taker, but we’re treating you as an order taker, blah, blah, blah needs training right here. Can you go train these people because this type of performance is not happening?”
We get upset about this in L&D because there’s so much more we can do.
If they’ve got access to systems like Office 365, and can use Power BI and leverage the resources from the Microsoft team in order to be able to solve human performance challenges without our involvement, what happens when those orders stop? What happens when they stop coming to L&D mistakenly ordering training to fix a problem because they can see through operational data what the root cause of that problem is, and actually address it in an appropriate way, whether we’re involved or not? That’s the first thing on the manager induction training.
It’s not:
“Well, do we follow this trend or do we follow that trend?”
It’s:
“Yes and…”
As an example, we can use Design Thinking and we can use something like Journey Mapping in order to be able to map out that manager’s journey. We can take a look and say:
“Okay, over that 12-month period, what are the conditions of success, not only at the end, (Summative) of 12 months, what makes the manager successful?”
Leading indicators, (Formative) leading up to that: How would we know if that manager is on track or not? We can do that by plotting out that first-year manager’s journey, and we can use Design Thinking as a discipline in order to be able to do that. When we do that, we can then look at those milestones. We can say:
“Okay, there are different milestones that are either time-based or capability-based or that can be observed in that person’s behaviour or maybe there are other assessment instruments that are already embedded in that 12-month run.”
We can take a look at what the milestones are at which we need to be able to measure again to be able to see on-track or not and we can lay that out. That’s why it’s really important like some of these disciplines that are coming in now, like Learning Engineering, which is something from IEEE. Being a Learning Engineer means being able to take learning science to your point.
We want to know about Learning Science, we want to know about the neuroscience of learning, we want to know how it is that humans learn, but alone, it’s not enough. We need to then take that and add that in with Design Thinking, with Human-Centered Design, Performance Consulting, and Data & Analytics.
It’s all of these things coming together into this multidisciplinary approach in order to be able to have these outcomes, not outputs, it’s not about creating assets anymore, like training materials or a training course, those are assets. It’s not enough. We need to be able to create outcomes. We need to be able to drive outcomes.
It’s that shift from outputs to outcomes and being able to tell if we’re on track. Formative, in addition to Summative, as we go. It’s about being able to make those small bets over time until we find out that something is working and then be able to scale it. Then we can make the investment and we can actually bring that into our organisation and scale it across the enterprise or to whatever population that that serves.
The way that organisations are using Artificial Intelligence to begin with is to process data. We’re using Artificial Intelligence in order to be able to do things like Factor Analysis. We’re using Artificial Intelligence in Operations to be able to analyse data at speed and scale. Not just data that we’re using to structure data like in a spreadsheet and numbers. We’re using unstructured data like video, like images, like sound. That’s all data now.
That can all now be processed and especially when we’re using Artificial Intelligence. Qualitative data, text mining, sentiment analysis, being able to look at language patterns, being able to look at the language patterns in an organisation. That’s where a lot of the Predictive Analytics is coming from. It’s not coming from the quantitative data sets, it’s coming from the qualitative data. The qualitative data, if you look at where our conversations are happening, is being generated and we can see it on communication tools like Skype, Slack, and Microsoft Teams, which is a big data repository now. Email is a data repository now. Where are conversations among people happening? How might we be able to leverage that data, analyse that data in order to be able to predict future human behaviour, which links to a particular organisational result?
A lot of the data sources that are being used on the HR Analytics side and the broader HR Analytics or Talent Analytics side or People Analytics is actually taking a look at those types of repositories. That’s why we’re starting to hear a lot about Organisational Network Analysis or Social Network Analysis is where we’re figuring out that it’s in the conversations in and among people. That’s where the work gets done.
Microsoft as an example actually has a lot of research now that says:
“You know what, it’s not about the individual skills of a person that actually determine success. It has more to do with that person’s ability to be able to build an internal network in their organisation across different stakeholder groups. That’s how work actually gets done.”
We can see in conversational data now through these different platforms, how those conversations are points of influence, like who has influence in your organisation and oftentimes it has nothing to do with the org chart.
Specifically, we need to be ready by building our own analytical capabilities, so we can go and meet the business in its ability to build its analytical capability. It’s a play within a play. Again, going back to that Boston Consulting Group quote about:
“We know that entire industries now are being not just disrupted but deconstructed.”
They’re being disassembled.
If we look at how industries are being disrupted and deconstructed then that means the business models are changing, which means the operating models are changing and we in L&D are supporting the ways that people work. If all of that is shifting, then we need to raise our perspective beyond L&D, that’s the first place to start. What is the greater context? What is happening in the world? What is happening in our industry? What is happening in our organisation and how do we align to that first to make sure we have repositioned ourselves and our value proposition to support the ongoing future state rather than our old traditional existence within an organisation. How do we redefine ourselves and our identity within the greater context of what’s happening in our organisation’s industry and our organisations specifically?
Learn how to solve real problems using data in The L&D Disruption Playbook or speak to our L&D experts.
About Trish Uhl
Trish is a globally experienced strategist and consultant leveraging data science, artificial intelligence (AI), emerging technology, analytics and evidence-based actionable insights to deliver business value. Her work sees Trish keynote speaking at conferences, teaching workshops and working directly with client leadership teams across North America, Europe, Asia, the Middle East and Africa.
Chat to our L&D experts and find out how Looop by 360Learning gives you your days back
© 2022 Looop by 360Learning Online Limited
