I haven’t done a ton of pre-employment testing in my career, but whenever they were assigned? It was always really inconvenient. One team had me taking an assessment the day our movers came to pack the truck for a cross-country move. Talk about chaos. I was literally sitting on a folding chair in the middle of the living room as they took boxes from the area behind me.

In that particular case, this test made no sense. First of all, I already had a signed job offer so quizzing me on grammar and SAT style questions felt unnecessary. Second, it was supposed to help see if I was a fit for the team. Again: already had the job. Seems repetitive.

But what pissed me off the most about the assessment? They never even told me how I scored or what they learned about me after taking the test. I didn’t even get an automated report. I just spent hours typing away because that’s what they made everyone do.

The Old Assessment Approach Doesn’t Work

Looking at the sheer volume of applicants most people are getting now thanks to AI, I’m hearing more companies talking about using AI for pre-employment testing. When you get 1,000 people for 1 job, we need some way to sort the stack. A short test seems like an easy way to do that.

However, the power of the traditional pre-employment testing has vanished with the introduction of AI. Now, every answer is one search string away. How do you automate the sorting of a big pile to a small one when everyone can cheat the game by Googling their answer? I don’t think it’s going to be by creating a list of questions anyone can answer by Googling or asking ChatGPT.

Don’t even start with monitoring the candidate during an assessment. Want to creep a candidate out before they even start? Go ahead. Try that one. Note: sarcasm.

4 Simple Ways To Make Your Pre-Employment Testing Better

A sea of sameness lies ahead with traditional assessments and if you’re going to keep approaching assessments the same way you always have? Prepare to hire some unqualified people. News flash: I don’t think AI can solve this problem. On a high volume role where you hire more than 100 people each year? Sure. Get an AI tool. Standardize the hell out of that pre-employment testing.

If we want to improve assessments for everyone else we hire once or twice every few years, it starts at building blocks - not technology. This process is already broken. You’re not going to fix it by making it faster. Here’s what you need to do.

1. Audit your pre-employment tests. Make sure you’re giving assessments that actually align with the role. Why is someone who would never be customer-facing doing an asynchronous video interview? When will their success rely on their ability to answer unpredictable questions on camera under duress? Writers should write. Software engineers should code. Assessments are for showing proof of work.
2. Tell people upfront. Be clear in your job posting or the post-apply email confirmation that you’re going to have people take an assessment. Tell them how long it will take, the style, and when it will be due.
3. Start over with AI. If you’re not sure what a good assessment looks like, ask AI for ideas, not outputs. Here’s what I mean: Don’t ask it to write you a pre-employment test. That’s how you end up with really predictable stuff that won’t measure the right outcomes. Prompt AI with questions like “let me see 4 examples of this” or “give me 10+ ideas for questions that assess the following skills..." 
4. Show people the results. Maybe this is just for me, but c’mon. There are a lot of type A people who want to know their results, damnit.