Whenever I think about “the future”, I inevitably remember all those crazy movies and videos made in the 80’s and 90’s, showing people in weird-looking silver suits and helmets, with all sorts of pointless and hilarious gadgets. What’s really funny to me is how incredibly precise some predictions were – we’re now talking into our watches and can play music or a movie by talking to a tiny little device, but also how completely outlandish others were – smellevision, flying houses, and snail-mail sent via rockets.
In reality, “the future” is actually far more subtle than writers and moviemakers had anticipated: a lot of the advancement we are witnessing now comes from humans gaining the ability to collect, store and interpret data in an automated way. William Gibson, a fiction writer widely credited with the creation of the cyberpunk subgenre, said the following: “the future is already here – it’s just not evenly distributed”, referring to the fact that some people already live in a time ahead of most, whereas others are still a significant number of years behind.
In spite of its repeated use in a socio-economic context, which was arguably not its intended purpose, the phrase applies exceptionally well when we look at the level of technology and the broadness of its application in various work domains. We’re currently witnessing a fundamental retooling of the enterprise: automation drives people towards creative thinking roles, making data-driven decisions. Product managers optimize platforms by analyzing user interactions, marketers optimize conversion through customer behavior insights, recruiters optimize the hiring process with the help of candidate funnels and the list goes on. Every function of the enterprise is now growing faster with the help of automated data collection and real-time data insights.
So what about Learning & Development, what’s happening here? There are a plethora of buzzwords that you may have heard in recent years: digital transformation, future of work, re/up/microskilling, future-proofing the business, digital intelligence, adaptability quotient, gig economy, and so on. And there are many experts who talk about what these big ideas mean, their impact on our work lives, what we should strive toward, and how bright the future ahead of us looks like.
Although I agree in principle – we should indeed be knowledgeable in what these concepts mean – I would argue that there is a huge gap between how these abstract and complex terms are presented to wider audiences and the very basic routine of daily L&D management, where they should ultimately be applied. It’s wonderful to think about the importance and impact of being tech-savvy, but how does it specifically apply to basic L&D work, like creating a training plan, extracting weekly reports, scheduling classes, centralizing feedback, calculating budgets, analyzing skill gaps, and understanding learning needs?
Initially, when I started outlining the content for this article, I had built in my mind a very straightforward sequence of what automation is and how and where it could be applied in Learning & Development management, based both on past personal experience, but also as a result of our customer research efforts at Nifty Learning. Speaking to L&D professionals, we discovered that as much as 30% of their time is currently wasted in tasks that a machine is very much capable of doing and, in some cases, the number goes much higher. For other business functions, this type of work has already been successfully taken on by software – there’s sales automation, marketing automation, recruitment automation, you name it. So how come, when you search for “learning & development automation”, the results vary greatly and the term isn’t yet established? Based on the variety and complexity of information that an L&D professional needs to process, the countless daily decisions that need to be made, the need for automation always seemed like a default to us.
Therefore, I’m using this opportunity to take a step back and start from the beginning: before we dive into the “what” and the “how”, let’s have a look together at the “why”. Of course, depending on the organization you are currently a part of, you may find that the reasoning applies to you to a wider or narrower extent. As the saying goes, there is no one size that fits all.
I started out with some research on the topic of learning & development automation, in an effort to not say what has already been said. What I discovered felt a bit odd: people aren’t really searching for “learning & development automation”. Sure, L&D is a very wide domain, with many, many specific subdomains and, yes, automation is all the hype right now, more recently with the spectacular rise of RPA companies, which has forever put the term on the map. But the combination of the two isn’t something that L&D professionals look for on a weekly or even monthly basis. And the resources I could find, while performing this search myself, mostly focus on the learner-facing aspect of learning & development automation: how to identify the best kind of content in the fastest way possible, how to generate and publish that content, how to optimize engagement with the content, advertising the learning offer and so on. These are, of course, very valid and important aspects of Learning & Development in any organization.
My favorite search results were the ones where the L&D function is itself being prompted and expected to help employees understand automation, which I found somewhat ironic. But the results felt lacking in another area of L&D: management, administration, overview, strategy, data gathering, decision making. Let’s call this the behind-the-curtains perspective, that of the L&D professional who has a population of employees to support in their professional development and relies on their tech infrastructure, team, and wider organization to perform optimally.
Could this be a case of the shoemaker’s child always walking barefoot? There are lots of search results on how employees themselves should be learning about digital intelligence, to align with the times and maintain competitiveness, both for themselves, as individual contenders on the work market, but also as part of their company, to support the growth of the business. The implicit expectation is that L&D, as a critical support function, is responsible for driving this momentous adaptation to technology forward for the entire company. So how come L&D people aren’t searching for ways to make their own work-life easier and to take advantage of current technological advancements?
Based on numerous conversations with L&D professionals, I’ve found that there is a set of possible causes for this, that unfortunately influence each other in a negative way. Historically, L&D has always been considered a sub-domain of HR, assuming a more passive role within the organization. Only in recent years and maybe with the extra nudge of 2020 have we started seeing L&D take center stage, pushing the agenda of business agility and digital transformation forward, as a matter of priority.
Within this hierarchical setting, L&D hasn’t really had a strong enough voice to demand its place at the table and explicitly ask for technological advancement: apart from a couple of very dedicated pioneers, L&D tech is a follower, not a leader in the enterprise tool kit. That could also be happening because, for a very long time, L&D hasn’t been able to prove its business value, being perceived as more of a “money spender”, rather than “value creator” in the company – this, happily enough, is changing as we speak, with countless companies and individuals working on defining and measuring L&D’s positive business impact.
And with L&D sitting under HR, it’s often the case that L&D professionals aren’t necessarily tech aware or savvy, sometimes relying on “the way we’ve always done it” to keep doing their job, other times overwhelmed with the massive amount of (administrative) workload to even consider any sort of technology adoption.
When ultimately and inevitably the time comes for L&D Operations to become more “cost-effective”, what ends up happening, especially in larger companies, is that the current HR solution’s L&D module is added to the tech suite, giving the L&D teams a pre-built, often rigid technical framework, that doesn’t necessarily fit with the culture and dynamics of how learning management actually happens within that organization. This leaves L&D professionals with a never-quite-right kind of technological support and a lot of loose ends to manually tie, in order to stay afloat.
Even worse, I’ve personally seen a case of a company attempting to cut costs by attempting to reduce its learning catalog to a quarter of the normal size, because “fewer courses means fewer costs”, disregarding the very basic fact that no two courses are the same and they address different needs. The administrative chaos this decision has generated is absolutely indescribable.
This being the situation and the wider context, I’m hoping that the information below can help shed some light on the potential benefits of having an increased interest in automation and the many ways it can make a difference in both L&D job effectiveness and, why not, satisfaction.
Let’s start with a look at “what” you would need to automate and “how” you could get started.
Have a look at your current work and identify tasks that are repetitive, fairly manual and take up a lot of your time. Try to capture – in a spreadsheet or a text document – the specific actions that you take, but also write down the triggers (what prompts me to perform this action), inputs (what information do I receive that must be processed), and, of course, outputs (what is the expected result after I finish performing this action). This, in short, is called business process analysis.
Such an internal audit of your day-to-day work, maybe over the course of one month, will start showing you some patterns. You will very likely discover two things: you have quite a high number of workflows that are, in essence, the same, but that only vary in inputs and outputs, and you will also identify process exceptions and what causes them. This pattern-and-exception identification is the perfect starting point in your learning and development automation efforts.
Before we move forward, though, I’d like to address the possible perception that automation couldn’t really work or add value to the specific work performed in L&D, because the type of activities require a lot of human intervention and don’t lend themselves well to being automated. After all, L&D is about understanding employees’ needs and the company’s strategic direction, looking for the best kind of learning opportunities, performing learning needs analyses, catering to the very diverse and very specific needs of each learning group or individual.
Here I would propose a quick mental exercise: if you are currently in an L&D role, do you use spreadsheets on a daily basis for any part of your work? If so, do you use multiple sheets for multiple purposes? Do you find yourself copy/pasting one type of data in at least two different places, more than once a day? Do you have a big ol’ spreadsheet that centralizes your work, together with your L&D colleagues’ work, for that “big picture” overview needed to make sense of things or that “give me a number” demand from management? Do you have processes that you must track over the course of weeks or months that are comfortably nestled in a table with at least 15 columns? I believe you know where I’m going with these questions. As a rule of thumb, if you’re regularly using spreadsheets for one or more of your daily tasks, chances are you are a good candidate for learning & development automation. As you’ll notice, most of the questions above are around data and reporting, but the same line of thinking can be applied to other types of work: scheduling, cost tracking, survey interpretation, compliance monitoring, etc. And this applies even in spite of having more advanced technology that you already rely on, like an LMS or LXP, performance management or employee survey tools, maybe some budgeting or resource management technology. The bottom-line question is: how much of the work happens in between or outside the software solutions that you have at hand, as opposed to being independently executed by them?
If you’re that fortunate L&D professional who’s got everything under control, has all the necessary data at hand, and always knows what stage any learning initiative is in, congratulations! This is no easy feat and you are clearly on top of things. My only hope is that you end up reading these lines and want to share your good practice story with others as well.
But one common theme that I’ve encountered in now hundreds of conversations with L&D professionals is that, in spite of finding it quite easy to specifically define what their job consists of and how to execute their tasks, they find it hard or sometimes impossible to explain why the workload is overwhelming or why tasks don’t end up getting done on time or at the level of quality they would love them to be – of course, barring the situation when the team is understaffed. And when I ask what the most common time traps tend to be, the answer focuses more often than not around menial, repetitive tasks, not being able to get to the right kind of data in time to make the right decisions, and just low-value busywork that is hard to put into words or measure precisely.
An awesome revelation came to us when we interviewed the Global L&D Manager of the Services Division in a large bank. Her scope of work extended to a couple of thousand employees and accounted for about 10% of the bank’s total workforce and she had a team of 3 people supporting her. When we met, she was just about to outsource the kind of work that I am referring to above to either their HR admin provider or to an intern/junior role. She had specifically built a business case, categorizing and time tracking the type of work she did every day by splitting it into value adds – tasks strictly related to getting employees closer to good quality, useful L&D content – and time traps – copy/pasting data, extracting the same type of report repeatedly, sending the same reminder to managers every couple of days, formatting tables, sending calendar invites, booking resources, etc. It was an absolute goldmine! She was sketching out her process on a piece of paper lying between us on the table, in the conference room, explaining her method and, in the end, told us her very saddening, but not at all surprising conclusion. After painstakingly tracking her effort over a longer period of time, she realized that she spent at least 3 months out of her entire year in time traps. Then, as she raised her head from drawing on the paper, she let out a very frustrated sigh and said “but I swear it feels more like 6 months, I just can’t figure out how else to measure it and prove my point further”.
One very likely answer to that dilemma is the sneaky and very wasteful context switching time cost that has become the attention of psychologists in more recent years: you’re completely blind to it, but it’s still time that passes without you actually making progress in your work. Going back to learning & development automation and its relevance in this particular example, it becomes quite obvious that there are very important, measurable benefits to leaving the repetitive work in the “hands” of a supporting tool – whether it’s just being able to stay on top of things and to stop falling behind on your current work, or saving precious time that’s currently being wasted in low-value tasks, in order to focus on more creative, enriching L&D activities. Or rather, to frame it a bit differently, it’s not that certain tasks are less important than others. After all, all of the work needs to be done, in order to consider the scope of your L&D work completely addressed. But some of these activities can take as little as 30 seconds to complete, could very well be done by a machine, but generate that context switching time cost which can sometimes be as high as 20-25 minutes for every single switch.
There is, of course, one potential showstopper that could prevent learning & development automation from being successfully adopted: inconsistent signals coming from the various sources and stakeholders L&D interacts with. Essentially, every time a piece of crucial data is missing or is delayed during communication, it creates an exception to the very business process you want to automate, which can make the adoption of technology hard or impossible.
So how can you control and maintain data inputs clean? Just like keeping your workspace tidy: if you plan to do it regularly, it’s relatively easy, less time-consuming, and produces consistently good results in more ways than one. A good example is employee attributes: if there are people signaling to L&D that they’re not receiving the right email notifications, it’s likely that they have a missing or wrong attribute in their profile: maybe their manager was promoted and this left a gap in their org chart; maybe they’ve changed job roles, but they’re still getting information that’s relevant to their previous role; it can even be as simple as having changed the last name, which can result in reporting errors. It’s worth keeping in touch with the various teams that L&D interacts with and establishing a data correction and maintenance protocol. Ideally, you don’t want to be making changes directly in your L&D systems, since that doesn’t correct the source error and, even worse, adds another manually entered variable into the whole flow that the L&D professional becomes the “owner” of. Otherwise put, it’s a great quick-fix, but isn’t sustainable in the long run, just like plugging a hole in a boat can be temporarily done with a stick, but we all know that boat isn’t traveling very far.
The same goes for any other kind of data that is stored in or, by way of integration, comes in contact with your L&D process, be it the course catalog, vendor data, rooms or other training resources, cost items, employee groups, assignment schedules, quiz & survey questions, and the list goes on. These errors ultimately trickle all the way down into your reporting, causing a lot of damage and missed opportunities along the way. And reporting, as a lot of L&D professionals repeatedly confirm, is one of the least favorites, but very time-consuming parts of the job. After all, very few people choose L&D as a professional path because they just love formatting that spreadsheet date field in that assignment completion report over and over again, every Tuesday at 4 pm.
Going into the “how” of it all, let’s have a look at a couple of ways in which you could put automation to work in your organization. This is by no means an exhaustive list, but here are a couple of examples of what you could do to make your work life easier, depending on the most typical tasks that you have to perform on a daily or weekly basis. Some of these might require more technical savviness on your part or for you to start a collaboration with your IT team or a new provider, but know that there are always options available. We are now navigating within the realm of BPA – Business Process Automation.
If you have tasks in Excel that you repeat ever so often, you can record a macro to automate those tasks. A frequent type of activity that an L&D professional may perform in Excel is centralizing various employee attributes from multiple spreadsheets, in order to unify the information in one single “master” report (using vLOOKUP or a similar formula) and then potentially creating a pivot table or chart, to extract L&D KPIs.
A macro is basically a set of actions that you can run as many times as you want and that never changes in structure. When you create a macro, you are recording your mouse clicks and keystrokes while you are performing the task in Excel for the first time. For all the following reports of the same kind, all you have to do is get your data sources organized, then click on your macro, and the work is done in seconds.
It’s a great way to save a good 2 to 4h of your time each week, depending on how much of your work happens in spreadsheets, and it’s very likely that you’ll be able to do it all by yourself, with the help of some nifty online tutorials. It’s completely free, provided that you already have an Excel license, but it does take a bit of an initial learning curve to get the hang of it.
For the various types of activities that an L&D professional performs on a daily basis, there are a lot of software solutions that can automate specific workflows, typically accessible via a subscription model (SaaS meaning Software-as-a-Service). The range is pretty broad here: starting with your standard LMS features, feedback/survey tools, performance management systems, quizzes and compliance testing, employee onboarding software, training needs analysis, scheduling tools, budgeting & cost tracking, reporting & KPI measurement, vendor management, or tracking trainers’ billable hours. Solutions can also be grouped by the goal or business function that they serve: sales training, customer service, leadership, coaching, compliance & audit – you get the point.
These kinds of software solutions are typically easy to set up by yourself, via their self-service function, but they often come with a price caveat: whether it’s a monthly subscription or you pay per usage, there is a possible conversation to be had with the Procurement department or the HR Director, to get approval for this budget. Some solutions do come with a forever free tier, as long as you work within specific usage limits.
These web applications are excellent in dealing with their narrow area of focus, but there is a possible hidden downside to having a lot of tools that each cover only one part of an otherwise complex L&D landscape. The L&D role thus becomes a human data integrator, since these systems don’t “speak” to each other or to a centralizing data reporting and analysis tool. You might find yourself doing multiple data entries in between these software solutions, or you might have to manually use the outputs of one tool as inputs for another: one example could be that the outcomes of the training need analysis are the starting point for scheduling activities and contacting vendors or trainers for offers and availability.
The upside, however, is that the value proposition for each of these tools is quite easy to identify, validate and explain to your team or your manager. Depending on how much time you spend performing a specific activity – let’s say you send out and centralize a lot of surveys regularly – it should be easy enough for you to measure how much time you could save and how much faster you could produce the needed results with the help of a tool.
In comes the next layer of technology that can help support and automate some of the L&D activities: business intelligence tools. Their purpose is to look at your multiple data sources and help you extract the right kind of trends that inform your L&D strategy. The power of such tools becomes all the more evident with their predictive capacity: not only can you see what kind of historical or current data you have in your systems, but they can also anticipate what could or should happen next, thus becoming an objective, data-driven decision making support to the L&D function.
The wonderful thing about using BI tools is that they give you real-time insights into what’s happening in your L&D operations, without you having to constantly extract and manipulate reports. As a complex business function, L&D doesn’t have a great track record of being consistently managed in real-time – it’s likely that a lot of important L&D work happens at the end of certain processes, not concurrently (a good example here is L&D cost tracking and being able to reallocate unused budgets throughout the year, as the money is being spent, and not just once, at the end of the year, a mere month before the budget’s expiration date – I’ve heard about plenty of cases when “we have to spend it quickly, or we can’t request it next year”).
A possible disadvantage of BI, though, is that it requires a lot of specific, technical knowledge to do the initial set-up, especially when identifying and defining the right way to calculate L&D metrics and KPIs. If this particular step isn’t done properly, you might be setting yourself up for failure by looking at the wrong kind of data or by looking at the right data, but in the wrong way. Partnering with your company’s IT team is the best way to get over this hill and the good news is that, once you’re done with the implementation, the BI tool consistently churns out impeccable data at the simple click of a button – better still, you might not even need to click a button, since you’d be getting reports directly in your inbox, based on a schedule that you define.
There are also some financial considerations to think about, as BI tools typically do come with a high subscription cost. Pairing that with the fact that a BI tool does require a dedicated implementation effort, it might be difficult to succeed in adopting this technology on your first attempt. You could, however, try to find out if your organization is already using some sort of BI technology in other business areas – very likely in Sales, Marketing, or Customer Service, and, if you’re part of a company that’s creating physical items (cars, food, clothes), chances are that the Manufacturing line is already well equipped with BI software to keep delivery and quality on point and costs low. You might find out that the know-how and software are already within your reach and all you have to do is build a business case to get that technology onboarded in the L&D department, too – more on that later.
If the name sounds unfamiliar or too technical, don’t worry, it’s really not. What it means is that there are software companies that have pre-built integrations between the different tools that you use on a daily basis (spreadsheets, calendars, communication platforms, social media, survey tools, etc), to help you automate some of your business processes. Essentially, application integration software allows the different tools you use to “speak” to each other.
Let’s have a look at a possible scheduling flow that touches multiple tools, to explain the mechanics behind the concept – assuming that a dedicated scheduling tool is missing from your tech suite: when I need to schedule a new training session, I ask for availability from trainers via an online form. When a trainer submits the form, a calendar invite is scheduled and sent to the Teams or Slack channel – where learners spend most of their time – to allow them to register. After the session takes place, the trainer marks attendance also via an online form, which then feeds into my centralizing training attendance spreadsheet. All of these steps can take place without the L&D professional’s intervention in the highlighted tools.
There is, of course, a bit of work to set things up. First, you would have to define and implement a specific collaboration protocol that involves the other stakeholders – in our example, our trainers and our learners – and ask them to communicate with you via this newly agreed method – process inputs and triggers should be consistently the same so that your workflow can indeed be automated. This means there is a change management component to getting things set up, apart from just the technology itself. Then, you would need to build the actual integration between these systems and to test it out a couple of times, first by yourself and then with the involvement of a small group of trainers and learners, before you do a full rollout. It shouldn’t take you more than a couple of weeks of intermittent work to fully go from the initial idea all the way to successful implementation and, before you know it, you’ll have a system set up to take over big chunks of work on your behalf and give you back your precious time. All you have to do is to check every once in a while if any adjustments need to be made to the rules that you’ve defined and if people are consistent in their interactions with L&D.
You can expand and elaborate on these flows quite a bit, depending on the type of work you want to automate. The more complex your workflows, the more time you’ll have to spend in that initial process mapping step, to break down each flow into trigger-action-output groups. Ask yourself what is the event or piece of data that triggers a specific action, where does that input go and what is the result of executing that workflow. You might sometimes need to manually allow the tools to communicate with each other via API, which can take you again into IT & software territory. What is very common, though, is that these integration software providers have extensive documentation and really friendly and supportive user communities, that can help you get set up independently. Whichever path you choose, doing it by yourself or asking for help from a software developer colleague, make sure that the solution you are about to use is accepted by your IT team in terms of security and data protection. It’s often just a formality, but it’s better to be safe than sorry when you give access to your email or storage space to an external tool.
In terms of cost, depending on what and how many flows you are trying to automate, you might end up not paying anything at all or a hefty monthly cost that can end up making a dent in your budget. The good news, however, is that many of these application integration solutions come with a free tier, too, as they are just another kind of SaaS. It’s important to keep in mind, though, that if you start going into the tens of integration flows, spending lots of money, and still struggling to get significant time back from your daily schedule, it might actually be worth investigating a more complex, maybe even bespoke solution for your specific needs. Chances are, it’s already been built and will fit like a glove.
Robotic Process Automation (aka software robotics) sounds very futuristic, but it works quite similarly to a concept we’ve already visited earlier: macros. RPA functions more or less the same way, only it can perform actions not just within or in between spreadsheets, but also outside of Excel, on top of the various software systems you access during your daily work. This works wonders for massive amounts of repetitive, administrative tasks (data entry, course scheduling, complex reporting), and is more suited for large L&D operational teams (in the tens of team members or above). The main reason is that RPA technology can get expensive pretty fast and likely requires a dedicated implementation project if you don’t have the knowledge in-house. The price for one such “robot worker” can range from 5.000$ to 15.000$ (or more) per year, though the true cost doesn’t necessarily lie in the robot’s price itself, but the implementation effort, and the various tweaks made along the way, as you’ll most likely have to work with an outside consultant. There is still the option of you learning to use RPA technology by yourself, to keep costs low, but this type of knowledge takes a lot longer to master than learning how to build macros and will likely take you away from your work even longer than you might currently afford – after all, the goal is to give you back time, not take it away.
A more cost-effective alternative to RPA is using a standard BPA tool – Business Process Automation. Essentially, what this means is that you have to manually program the BPA software to perform actions based on a series of if’s – the trick here is to be as granular as possible in defining all the if’s at the process mapping stage, right at the very beginning. This way of doing things is more similar to the application integration software option above, instead of macros. Granted, it’s not as spectacular as RPA and doesn’t rely on screen scraping or recording mouse clicks to capture your entire workflow – you literally have to tell it what to do, instead of it quietly observing your actions on the screen – but the outcome is very much the same as with using RPA. The costs could end up being smaller by at least one order of magnitude than RPA, but there is the same learning curve or consultancy aspect to take into account.
If RPA/BPA sounds like something you might want to implement in your L&D management work, I would recommend doing a bit of research internally first, similarly to BI tools, to find out if another department has already taken those steps before. You might even discover that some people in your organization have taken some RPA/BPA courses and are now skilled in designing software robots, in which case the consultant cost could even be offset altogether.
Moving on to an even higher level of technical complexity, Artificial Intelligence is a term that you have heard repeatedly in the past years, with various applications in virtually every domain imaginable. In L&D, AI is used for multiple purposes: creating a personalized learning offer, generating smart content, automating tasks, teaching, interpreting information, making data-driven decisions.
A more recent and really amazing application of AI comes in the form of skill gap assessment and training needs analysis. By way of NLP – Natural Language Processing, and ML – Machine Learning algorithms, you are able to delegate some of this very detailed and time-consuming work to technology. The heavy lifting can be done by the software, while you act as a validating filter, making sure to adjust the outcomes and correctly interpret the results of the analysis before you actually go ahead with the L&D planning part. Pairing that kind of power with an adaptive learning mechanism typically found in LXP platforms and then allowing a Data Analytics tool to not only track metrics, but also give you predictions, and you’re all set for a fully automated L&D management set up.
With the right kind of technology set up, you can get back as much as 50% of your time spent doing administrative work, possibly even more. A disclaimer must be made at this point, though: technology has come a long way and we’re now able to rely on it for an immense variety of tasks. But right now and also for the foreseeable future, learning & development automation can only augment and not fully replace humans in the workplace. It’s less a matter of reducing headcount and more of an opportunity to put your time and intelligence to better use in service of the organization you are part of.
Of course, there are certain types of tasks (and roles) that are very good candidates for extensive learning & development automation – more so for those types of administrative work that focus only on one or a small group of repetitive actions: scheduling, budgeting, reporting, etc. But, as is the case for most L&D professionals, there are significant parts of the job that fully depend on the type of intelligence that is inherently human, and which computers and algorithms are not yet (and for a long time won’t be) able to replace. Sci-Fi movies and books, thought-provoking as they may be, do sometimes contain fear-inducing apocalyptic scenarios and entertain this fear in us that “the robots will take over”. In reality, there is no genuine threat that machines will be fully replacing people any time soon.
There’s only so much technological advancement that we can accommodate in our life and work, before it’s moving too fast for our brain to process and thus begins to render itself useless to us. An engaging and effective professional growth experience for an employee can’t happen in the absence of key people within the organization, to give them the right kind of support in crucial moments in their journey. And if there’s one thing that our recent global circumstances have shown us, it’s that, with the ever-stronger presence of technology in our lives, which has even enabled us to work completely remotely, doing incredibly complex work, the one thing that keeps us going is human contact and support.
However advanced and powerful, a dedicated learning & development automation mechanism that can integrate with your work and acts as a virtual assistant will always remain a support system, not a contender for your job. We are a very long way away from allowing technology to take over the ineffable things that make us human, like creativity or empathy.
Automation works best in a human-in-the-loop setting, especially since we’re talking about processes that apply to and directly involve and impact people, and not raw materials, production lines, or quality control. As a rule of thumb, when humans are the end beneficiaries of a specific type of work, then humans must always be involved in the execution of that work. So what are possible ways in which learning & development automation fails?
Allowing a machine to take over completely is one possible failure scenario. For example, when performing a needs analysis or interpreting free-text survey responses or test answers, in spite of how advanced NLP currently is, there are very specific context-related factors that only you, as an L&D professional, are able to understand. It’s the equivalent of being able to read between the lines, something which machines simply cannot do.
Trying to automate communication flows where a human touch is needed is another way in which technology hinders more than it helps. It’s great to have chatbots or user guides that allow learners or managers to figure out by themselves how to interact with a new tool or how to take charge of their own learning path, but a conversation with an actual person will always be essential to deploying a quality L&D strategy. The L&D function acts as a bridge between employee professional aspirations and the company’s strategic direction of growth. L&D professionals should constantly be in tune with what the business is looking for and should be able to articulate and support the deployment of a strategy that meets the business’ goals. This means that technology should be a facilitator of communication, but never a complete replacement.
Exceptions are also a possible cause for automation to fail, but they shouldn’t be the reason you give up on even getting started. Sure, there are certain special, last-minute, or critical business requirements that must be handled in a very particular, hands-on kind of way – maybe it’s something that is time- or budget-sensitive, maybe it’s a specific intervention as a result of bad customer interaction or a faulty product, or it’s a manager-team dynamic that needs not just some learning, but also some facilitation. Trying to fit exceptions into an automated process just for the sake of time or cost gains is a very bad idea, as it produces cascading negative consequences later on and paints a less-than-flattering portrait of automation, which then becomes the default scapegoat. But this is where the creative and empathetic nature of the L&D role comes into play: the reason why we automate the repetitive parts of the work is so that we are able to handle exceptions in as much detail and with as much attention as needed so that the overall outcome of L&D work is always of high quality.
If you’re quite new to the topic of learning & development automation, I’d suggest you start small. It might even be a fun exercise to see for yourself and your own work program how much time you actually spend doing repetitive, administrative work versus what your current perception of your work is. You can start by initially writing down what the purpose of your role is, what your current perceived activities are, and how much time, in percentages, you spend performing each group of tasks. You can then actually start the time tracking itself, let’s say over the course of one week or a whole month.
Create a minimally disruptive time-keeping discipline, that doesn’t get in the way of doing your work – for example, at the end of each task, go into a very basic time tracking spreadsheet, write down the name of the activity, the duration, and whether you feel it was a value add in your day or more of a time trap. You might even want to record how that doing that work makes you feel, with the help of a very simple emoji or thumbs up/down indicator. After all, your own motivation and engagement also influence how well you perform your job. After you’re done with this time tracking experiment, apply a couple of filters or just generate a pie chart and see what the outcome is. It might surprise you to discover that your initial, subjective self-assessment was way off and you actually spend your time very differently from what you thought. If you find that a quarter or more of your day is spent in repetitive, manual tasks, it’s very likely you could benefit from adding automation to your daily workflow.
If you’ve already dabbled in trying to automate some of your work and you’ve succeeded, then you might find it useful to take a step back and try to perform a more holistic analysis of your role. You could devise your own strategy for how to allow technology to substitute as much of your work as possible so that you are left with the parts that you must in fact always be present for – those tasks requiring your involvement, attention, and your critical thinking. It doesn’t have to happen “now” and it doesn’t have to cost you an arm and a leg, but it’s a great opportunity to objectively observe your own role and figure out how you can minimize time traps so that you have more room for your value-creating work.
The follow-up to the above is actually going ahead and implementing the automation mechanism itself and then letting it do its magic. If you are able to do it completely by yourself, with minimal time and cost investment, you are a very scrappy self-starter and might even secretly be an entrepreneur at heart – I salute you! If, however, you discover that you’re spending a lot of time in low-value work, but the pressure on you just keeps mounting, and you’re feeling somewhat lost as to how you might start, you could consider building a business case, to present to the decision-makers in your organization. How do you go about doing that?
Start with the facts: track your time, your process inputs/outputs, and the outcomes of the various types of activities you perform. Any time that you spend with someone else in your day is also time spent on their side: discussing an allocation with a trainer, approvals with a manager, or handling a last-minute and costly cancellation from an employee, etc. Ask these stakeholders if they also perform any L&D-related work and try to understand how much of it could be taken over by an automation mechanism. Make sure that your tracking is objective and accurate, as it is the foundation of your business case.
Next, you should assess the value and criticality of those groups of activities. What is more crucial to your role and to the business? Can you make a comparison between how important it is to do on-going class scheduling compared to discovery interviews with the business? Can you give weight to these activities in terms of how business-critical they are versus how much they can be automated? This step should paint a clear picture of how within reach it is for you to automate certain tasks combined with the urgency of dedicating more of your time to focus on those critical, human-dependant activities.
Lastly, if “urgency” is too vague of a term to help argue your case, try a proxy measuring unit, typically expressed in time or money. How much time are you (and other stakeholders you engage with) spending to perform specific L&D activities? You might discover, for example, that you spend almost half of your time going back and forth with employees or managers on purely administrative tasks. On the other side of this coin, those same employees also spend maybe 1 to 4h per month, their managers maybe 2 to 5h and trainers probably a lot more, engaging with you. If you put together all these hours and then also add salary or opportunity costs into the mix, the possible time & money wasted might become a surprisingly high amount, which ultimately serves to strengthen your business case.
Ideally, this kind of analysis should be done within your team – provided that you are not a one-person-L&D-show in your organization, of course. This helps eliminate biases in how you assess the data and gives the business case even more validation. Once you have all your data neatly gathered and analyzed, prepare a presentation that is to the point, contains a strong value proposition, and shows a specific goal that you want to achieve with the help of learning & development automation. If you want to save the company time and money, highlight those numbers. If you want to get more time to engage with the business and create a more relevant learning offer, point out that aspect. If you want to make better, faster decisions, focus your business case around the impact of having access to reliable, real-time data and how that affects the relevance and timeliness of the L&D strategy.
And one last, but very important, though sometimes overlooked detail is this: L&D professionals aren’t in the habit of speaking up and making their voices heard, especially where there’s always a business fire to put out that takes precedence. I’ve attended countless events with really amazing L&D speakers and I’ve unfortunately discovered that this is a recurring theme: the L&D function is passive, in part, because it doesn’t know how to make its voice heard. A very good way to support your cause is by building a business case, of course, but it doesn’t go very far if you’re shy or reluctant to speak up, not very confident, or if you always allow other priorities to come first. One of the ways in which building a business case can really help you out is by putting that unbiased data in front of you – it might even act as an internal courage trigger. All of a sudden, you are no longer on the shifty territory, unsure how to get your point across; on the contrary, you now have very solid data to back up your claims.
Whatever the case may be in your organization and for your particular flavor of L&D activity, one thing is absolutely certain: understanding the applications, power, and impact of automation is no longer optional in our domain. It’s simply no longer a matter of IF parts of your L&D job will be automated, but rather WHEN and what effect that has on you, the individual doing the work.
The best way to stay on top of this wave is to be informed and prepared. Make a conscious effort to understand your current work environment, take a look at the bigger context your company operates in, improve your business acumen and always be aware of the impact that you have in your organization, as an L&D professional. Technology is an excellent way to help you augment and enrich your and your colleagues’ work and life and it’s well within your reach to make that happen.
I'm Co-Founder and CEO of Nifty Learning - a tech solution that helps L&D professionals stay on top of their overwhelming workload by digitizing manual tasks, shortening decision-making loops, and optimizing training budgets. I love exploring how technology and data can help people do more fulfilling, impactful work and I'm very passionate about L&D processes and analytics, in particular.
Each Sunday we compile the best L&D resources we find and send them right to your inbox.