Training is not bad. Bad training is bad.

This is what happens when Instructional Designers do not know Instructional Design

If I told you I was training for the Olympics what would you think I was doing?

Chances are you’re not thinking about me reading about the history of my sport, being shown a PowerPoint presentation of technique, or answering quiz questions in an eLearning module.

Think of any type of training outside of the corporate world. You’ll likely think of people learning skills through practice, feedback and being assessed on their performance.

And you should. That’s what training is.

For many different reasons the idea of training at work has been watered down to represent communication. Lectures in physical and virtual rooms, and digital content.

Oh, content. What a soulless, non-specific word that is.

My intention in writing this for you is to see a new perspective. I want you to learn a few things from this but this is not training. What I’m writing now is communication. It’s not training, but if I were to add a few graphics, drag and drop activities and a quiz at the end a lot of people would refer to it as such.

They’d be wrong.

Two type of workplace Training

To make it simple let’s categorise training like this:

on-the-job training, and
off-the-job training.

On-the-job training is straightforward to most people. People are learning while doing the work.

Off-the-job training is where things have way more room to go wrong. It’s where the biggest mistakes are made. That’s what most of this essay is focused on.

Off-the-job training is a replacement for doing the work. Off-the-job doesn’t mean ‘as far away from the work as possible’. In fact, the aim should be to make it as close to the work as you can.

I’ll show what I mean while we take a look at two crucial aspects of training that are often ignored.

1. Practice.

Training must provide practice. That practice must be as close to the context of the desired performance as possible. If it doesn’t, then it isn’t training. The closer to the context the better. There are times when we can’t practice the real thing exactly as we’d do it. In those cases we have to get the practice as close as possible to the real thing.

Let’s say you’re learning how to box. The best practice is sparring. Throwing punches, dodging, punches, moving around the ring.

You get feedback in two ways: Your coach will tell you what’s working, what isn’t, and how to change. You’ll also get real-time feedback by hitting your opponent, avoiding being hit and getting hit.

You can’t spar all the time. You need an opponent, you need space, you need a coach to watch, you don’t have full control on what gets practiced, and it’s really not that good to get hit in the head too often.

So then you’d use a punching bag. Pick specific punches or combos and hit the bag. You can self-assess based on what you know and your coach can watch and give you advice. It isn’t the exact context of a boxing match but it’s still very useful practice. You’re working within logistical constraints and you’re also able to zone in on specific things you have to work on to get better.

It’s the same at work. We have constraints and we also need to be able to provide practice on specific skills and actions. We can’t wait around for someone to have a heart attack every time we need to teach someone CPR, so we use manikins. Astronauts can’t practice flying a rocket every day, so they use flight simulators. The customer service team can’t practice provisioning products to customers in the new software so they use a sandbox environment with fake data.

Training must include practice and that practice has to be as close to the real thing as possible.

2. Assessment.

Training also needs assessment. In fact, I’d go as far to say that if it doesn’t involve assessment it isn’t training.

Not quizzes or knowledge checks. Assessment of the actual performance.

Q. How would you assess someone’s proficiency in building a brick wall?

A. You get them to build a brick wall.

They build it and you assess their work on important criteria for a brick wall. Is the wall stable? Does it look good? How long did it take? How much material was used and how much was wasted? Was it built to the specifications provided?

Proper assessment isn’t asking someone if they understand. Assessment is judging their performance and output. Just as we’ve started equating workplace training with content, we’ve started to think of assessment purely as exams, drag and drops, pop quizzes and the multiple choice questions at the end of an eLearning module.

To ease people into this pro-assessment idea, and away from years of anti-assessment sentiment, I explain it like this: Assessment is practice with a score.

I do the thing, I get assessed and given my result. That “score” is a binary pass or fail, and/or a grading in terms of the criteria. In the brick wall example let’s say we have three 2 metre high walls to assess. One is at exactly 90 degrees, another has a 2mm variance, and the last has a 5mm variance. The building regulations in that area may allow a 4mm tolerance so it’s a pass for the first two walls. The third fails. Their adherence or variation from the optimal result is given to them. All three apprentice bricklayers know exactly how they performed for that part of the criteria.

The only thing that separates the assessment from practice is that we don’t get that detailed score (or results if you prefer that term) from an expert during practice. I can practice laying bricks and building walls by myself. That’s useful, but without an assessment I do not have a true measure of my performance.

There should be no difference from that example and any other workplace training. Assessment means we know exactly how proficient someone is. It also helps you learn how effective your training is.

And it saves time and money.

Instructional Designers have a professional (and I would argue, a moral) obligation to not waste people’s time.

We waste their time when training is not effective (they did not learn) and inefficient (they took longer to learn than needed).

eLearning is an especially big risk for this due to the size of the audience. eLearning is appealing because it scales. You can reach people quickly and you don’t have to book in time with them, schedule rooms or organise any specialised equipment. So extra special care needs to be taken with eLearning. Creating useless games or over-engineered interactions because they’ll be “fun” a clever decision.

If you waste 10 minutes of time for 250 people that’s over 40 hours of paid working time wasted. A full work week. Think about 30 minutes for 3000 people. It’s a lot. That doesn’t even include the time spent making the content.

If we assess we know exactly how proficient people are. From that we can also tell if it was or wasn’t effective. If we assess we also can also be efficient. Why should someone do more or less training than they need to?

There are times when someone has to do training for something they can already do. Like a new starter with plenty of industry experience in other similar orgs. Or example of people who can do most things they’re being trained on already. Like someone with transferrable skills, or a natural aptitude for what is being trained.

There are also times when someone takes longer to train. Someone who doesn’t get it straight away, but will be good with more time. A bad hire. Someone having a bad day, or distracted by something else. When we do not assess throughout the training these people are not given the appropriate visibility of their progress.

This creates a problem for people on either side of the spectrum. The people who need to be trained less have to do more than they should. People who need more get less than they need.

Fix it with Assessment

Assessing someone’s performance throughout training makes it more efficient. The proficient can “test out” earlier than other people who need more practice.

It follows a loop like this:

Create that loop for every aspect of the process you’re training and you have a full training course. It can be linear, branch off, be locked down in some areas and not others, or combine tasks into full end to end scenarios. That all depends on what/who/when/where/why you’re training.

What’s important is that the person going through the training has a clear sense of how well they’re doing. The trainer (or system) can see when someone is proficient and can progress from training into the next level of training, or move on from training and into the work itself.

They learn by doing and getting feedback. Include consequences in that feedback and show why they’re right or not. The only other information they need throughout the training is:

What is expected of them: Goals, process and criteria. (what good looks like)

Anything they need to do the work: e.g., product information, data, SOPs, reference material (what they would need on the job anyway, give them access to the real thing)

Backwards Backwards Design

When designing training we should follow a process of creating objectives > assessments > activities and information.

That doesn’t tend to happen these days.

A lot of the time it’s something like: information and activities > assessments > objectives. The objectives get written last as an afterthought because “well, we’re supposed to them”. This is exactly how you end up with bad content.

Useless learning objectives thrown into the face of the person you’re training, activities far removed from the context fo the work and way too much information.

Your objectives are the goals of your training. When written correctly they give you everything you need to design the training.

The brick wall example from before would not be “By the end of this training the learner will understand/describe/explain how to build a brick wall”. It will list out the desired performance, the criteria for it and what’s needed to achieve it.

Here are some simplified examples that should be relevant to most people reading this.

Example Objectives

Objective: Making double espressos in a cafe

Given a calibrated espresso machine, grinder and fresh beans during a service shift, the barista will pull a double espresso that runs 25–30 seconds and yields 36–40 g, with a consistent crema, on at least 9 of 10 consecutive shots.

Objective: Operating a forklift in a warehouse

Given a counterbalance forklift and a loaded pallet, the operator will retrieve the pallet from a 3 m racking bay and place it in a marked floor zone within 2 minutes, with no contact to racking, stock or pedestrians.

Objective: Handling a complaint in call centre

Given an inbound call from a dissatisfied customer and access to the CRM, the agent will resolve or escalate the complaint within one call, logging the outcome and next action in the CRM before ending the call, on 90% of audited calls with an after-call customer satisfaction survey rating of 7 or higher.

Notice how the objective spells out exactly what the person needs to learn and do? It sets clear expectations and shows them how they will be assessed. For the designer it practically spells out the assessment for them.

All you have to do is build it.

Example assessments:

Assessment: Making double espressos in a cafe

Setup: calibrated machine, grinder, fresh beans, during a live or simulated service shift. Assessor has scales and a timer.
Task: “Pull 10 consecutive double espressos.”
Measured each shot: shot time (25–30 s), yield (36–40 g), crema present and even.
Pass: at least 9 of 10 shots meet all three. Fewer than 9 = not yet competent; re-sit after practice.

Assessment: Operating a forklift in a warehouse

Setup: counterbalance forklift, loaded pallet in a 3 m racking bay, marked floor zone, timer, observer positioned to watch clearances.
Task: “Retrieve the pallet and place it in the marked zone.”
Measured: completion under 2 minutes; pallet fully within the marked zone; zero contact with racking, stock or pedestrians; seatbelt worn, horn at blind corner.
Pass: all criteria met in a single attempt. Any contact with racking, stock or a person is an automatic fail regardless of time.

Assessment: Handling a complaint in call centre

Setup: live or recorded inbound complaint calls with CRM access; assessor reviews a sample of audited calls against a checklist.
Task: handle the call to resolution or escalation.
Measured per call: resolved or correctly escalated within the one call; outcome logged in CRM; next action recorded before call end.
Pass: 90% of the audited sample meets all three measurements and an average of 7 for the satisfaction score.

The most important work is defining what the desired performance is and how it will be measured. Once you do that everything falls into place.

I’ve seen some people describe this as ‘Backwards design’ which honestly, seems backwards to me.

This is Instructional Design. This process has been around in various forms for over 50 years now.

Content ≠ Training

We should push back at misusing the word training at every possible opportunity. We should also push back on making this kind of content-disguised-as-training.

It will continue to get made when we people do not understand what training is.

Training cannot close every performance gap but unfortunately it is the easy go-to in business. “Numbers have dropped/We have more errors/We need more output… they need more training!” is an assumption that gets made often. Yet there are still many situations in which training is needed. Without having a firm grasp of what good training looks like we’re stuck in a cycle of training being overused and used incorrectly.

When training is needed it’s not designed properly. How can someone design something they do not understand?

When training isn’t needed we may still get content-disguised as training. If you don’t have an understanding of training then how can you ever determine if training is the right option?

You’ll end up saying yes to every training request. And you’ll end up and making something that isn’t training.