If you make a copy, they will learn…

Our TalentSprout Technologies group has two very talented people that make their living copying software.

But don’t worry, David and Roberto neither pirate software nor spend their days copying and duplicating files from one device to another. No, these guys do something quite extraordinary: they expertly replicate and build new software simulations that look and function just like an existing piece of software.

But in a world where it’s possible to duplicate software by just using the “copy” command on the computer, why go through the process of painstakingly replicating a piece of software and building a simulated version of that software?

To answer that question, it is useful to first understand what we do at TalentSprout. Our mission is to develop training that enables our nearly 50,000 customer service agents to do their job, and to do it well. A big part of the job for these agents is being proficient with complex software packages. And if you ever tried to master on your own a software package like Excel, you know how challenging a training task like this can be.

Of course, if we were training people to use Excel, we could let them practice by simply running Excel, doing practice exercises, following hands-on tutorials, and the like. With desktop software (meaning software that runs without an Internet connection) users can’t really screw anything up outside of a few sample files that can easily be thrown-away and rebuilt. But if you are training people to use large, database-driven systems that contain product, customer, and company information – the kind of software most big companies use – the training world gets much more complicated.

For example: if trainees use a real customer-ordering system to learn how to place an order, then when they practice placing an order, they will actually place an order! Clearly this is a no-no. Instead, those agents need to train using a copy of the system (a copy that is exactly like the “real” thing, but doesn’t place orders!).

But if you copy the software (in the simple “copy and duplicate” sense from above), you still face serious problems. Most significantly, the copy would contain the same data as the original, meaning you’d be revealing personal information to trainees. And if that information is financial (a credit card number) or medical (a patient record) in nature, there are strict laws against doing that.

But even if you copy the software and create a database full of fake information, you’re left with yet another training challenge, namely, that all trainee actions will be what we call “preserved.” This means if you give one trainee a practice assignment to cancel the fee on Fake John Smith’s account, you can’t give that assignment to anyone else until you reset the database. Worse yet, if your copied software uses a third party system to run a credit check on a potential customer, then a trainee using the copied software could still issue a very real request to run a credit check – again, an unacceptable training situation.

To address these challenges, some organizations end up building complex workarounds that feature recurring database resets and partial system lockouts, but this ends up being very undependable and risky. Other organizations solve this problem by essentially giving up: rather than let trainees gain experience through real, hands-on practice, employees sit in a classroom and simply watch an instructor talk through a manual full of diagrams, charts, screenshots, and step-by-step procedures. Now ask yourself, did you learn to use Excel by reading a manual or listening to a lecture on how to use Excel? I didn’t think so!

At TalentSprout, we believe that effective training requires learning by doing, not learning by listening or learning by reading. While this is neither a new nor heretical concept in education research, the fact remains that most classroom training around the world is in fact done via passive learn-by-telling or learn-by-reading.

In our world of customer service, ineffective learn-by-telling strategies have serious consequences, measured by both the large number of agents who quit within days of completing their classroom training (thanks to being “thrown into the fire” of using the real system!) and from those that remain, poor customer service metrics.

This is unfortunate from pretty much every angle. Hard-working new employees lose jobs not for lack of effort or skills, but because a company failed to give them the training they needed to succeed. And well-meaning companies lose significant money (paying trainees for weeks of training only to have them subsequently quit) and business (thanks to the dissatisfied customers who received poor customer service).

Conversely, learn-by-doing strategies have consistently proven to produce superior outcomes, weather measured by employee retention, agent performance or other key business metrics.

So now you get a sense why David and Roberto are so important: by creating robust simulated copies of the software systems that agents use in their day-to-day jobs, TalentSprout provides new employees hands-on, real-world training experiences that enable those employees to master that software and subsequently go on to deliver superior customer service.

In a future post we’ll make the case for how simulations can actually be better than the real thing. Stay tuned!