We all got needs

Several years ago, the FDA ran a campaign to reduce sales of tobacco to minors. One sign showed a convenience store clerk with a cartoon-style balloon:

You need cigarettes?
I need some ID.
Hey, we all got needs.

As Cathy Moore said the other day, when it comes to designing training, I too am a big advocate of writing the questions before you write the main materials.

When I say “questions,” especially at the outset, I’m really asking, “What do you want people to do?”

Take, for example, those convenience-store clerks. A challenge in these stores is preventing tobacco sales to minors.

I worked on a project to redesign training for the clerks. The original test items were real stumpers like “Is it legal to sell cigarettes to someone who’s 15?” A good example of a poor albeit fact-based question.

Poor, because if people can pass the test without taking the training, the training’s not worth much.

What’s more, convenience-store managers know that a brand-new clerk can answer the question correctly. A good thing, since turnover can be as high as 400%. But there are lots of things the new hire can’t do (and that some current clerks can’t do well) even if the manager’s standing there.

Things like:

  • Decide when to ask for ID
  • Identify acceptable forms of ID
  • Check for the validity of the ID
  • Refuse a sale in as customer-friendly a way as possible
  • Respond to specific difficult situations like drop-and-run (customer drops money on the counter and runs out of the store with the smokes)

I’m simplifying here — for example, individual states have their own specific requirements. A state might require a particular type of sign, for example. Signage, though, isn’t really the clerk’s responsibility, except perhaps to replace a damaged sign.

By focusing on what we want the clerks to do, we can work toward figuring out what information they need, what judgments they have to make, and what strategies can most effectively get them there.

Take the objective, “decide whether to ask for ID.” One approach to getting there is “decide if customer looks younger than 27.” Why 27? Because it’s too hard to guess age, as in, “does she look like she’s over 18?”

The client’s instructors knew that clerks sometimes resisted this reframing, and had developed exercises to underscore the difficulty of guessing if someone’s over 18. After eight or ten attempts to guess age based on photos, the clerks generally agreed on the usefulness of “asking down.” And on the value of consistently asking for ID from anyone who looks younger than 27.

Those design decisions flow from questions that make up a high-level assessment for the training program. No, they don’t appear as Mager-style objectives. The audience is implicit (convenience-store clerks), and the conditions aren’t detailed (certain types of customers in certain transations), but the performance is clearly observable.

Which makes it easier for the client and the developer to agree on what a properly trained employee should do.