Every compliance program is struggling to demonstrate the effectiveness of its employee training. By effectiveness, I mean that the training changed employee behavior.

The struggle comes from two places.

First, there is no baseline. Most teams start measuring after the training has already been delivered, with nothing to compare against.

Second, there is the temptation to prove the effectiveness of the entire curriculum at once. That is an impossible task, and it paralyzes the team.

The way forward is to stop trying to measure everything and start with one small piece. The most important piece.

Here is how to do it:

1. Identify your biggest risk.
2. Identify where in the business that risk is most present.
3. In that location, identify the employees closest to the risk.
4. Identify the specific tasks that are critical to managing it.
5. Pull together the history of mistakes and violations tied to those tasks. If there is no history, pause and build one.
6. Train those employees on those tasks. Embed the training in the workflow whenever possible.
7. Measure mistakes and violations after the training.

You’ve just measured training effectiveness.

For one tiny piece of your business. But an important piece.

Now rinse and repeat with your next biggest risk (or risky task). Then the next.

Do this relentlessly. When DOJ eventually knocks at your door and asks for proof of effectiveness, you will have something real to show them.

Your insider trading policy probably covers securities.

Does it cover a $500 bet on whether your company will close a major acquisition by year-end?

Prediction markets — platforms like Kalshi and Polymarket — now offer contracts on events like product launches, earnings results, regulatory approvals, and M&A activity. Anyone with access to material nonpublic information (MNPI) can place a wager and profit.

No securities change hands. But the legal exposure is very real. I wrote briefly about it last month (see here) but today’s article on NPR (link) about campaign staffers betting on their own political candidate prompted me to explore this a bit more.

The CFTC has already issued an enforcement advisory. The SDNY has met with platform operators. DOJ is watching. Enforcement test cases are coming and the companies whose employees are named in them will wish they had updated their policies sooner.

Here’s what E&C teams should do right now:

Update the policy. Explicitly cover prediction market trading. Most existing insider trading policies are written around securities and simply don’t get there. Great article here.

Expand the definition of MNPI. Product launch dates, earnings scripts, cybersecurity incidents, and merger timelines can all move prediction market contracts. If it’s confidential and material, it’s covered regardless of the instrument. Another great article here.

Widen the net. The risk isn’t just in the executive suite. Engineers, investor relations staff, communications teams, and legal assistants all carry information that could move a market.

Address tipping. An employee who shares a tip with a friend who then places the bet may be just as exposed as the one who bet directly. Make sure your policy says so clearly.

Train. Then certify. The most common defense will be “I didn’t know this counted.” Training eliminates that defense. Annual certifications create accountability.

One more thing that often gets overlooked: a policy alone accomplishes nothing. If you update the language but don’t train, monitor, or enforce, you’ve just created a paper control.

Prediction markets are a compliance blind spot. The good news is they’re not a hard one to close — if you act now.

Early in my business ethics career I was designing the annual hour-long employee training. Then I learned that our largest business unit would take a $1M hit to its bottom line because of this training.

I was shocked. But the math was simple:

Number of direct charge employees

X

average hourly wages

X

length of training

=

$1,000,000

I saw the financial impact we could have by creating a shorter – but just as impactful – training. Millions of dollars across the organization.

Do the math for your organization. How much could you save if, without sacrificing quality, your employees spent 25% or 50% less time completing online modules?

Share those numbers with your GC, your CFO, your CEO – and enjoy the nods of approval.

A critical goal of compliance training is to influence employee behavior and prevent compliance violations.

To measure training effectiveness, it is therefore necessary to measure the targeted behavior before and after the training.

If your organization doesn’t know what behavior it is trying to influence, or if it doesn’t try to measure the before-and-after, it’s possible that the training is there simply to check the box.

In a study by Kouchaki and Desai (2015), employees who reflected on the importance of ethical behavior were found to be less likely to engage in unethical behavior.

To create the conditions for such reflection, companies often provide off-the-shelf ethics training. Unfortunately, most employees find the traditional online modules to be relatively boring (and the end-of-module quiz too easy).

Perhaps a more engaging approach is to provide employees with opportunities to discuss real-world ethical dilemmas. A quick look at the morning news should offer several examples of cheating, lying, and stealing. Any supervisor can start a discussion with this phrase: “This just happened at Company X. Could it happen here? If not, why? If yes, how could we respond differently?”

Doing this on a regular basis is sure to align the moral compass of most employees. Give it a shot.

The concept of minimal viable audience (MVA) works well for most businesses. You don’t have to convince everyone that your product is right for them. Instead, you can focus on customers who already want what you have to offer.

When you try to please everyone, you risk creating an average product. This might be why ethics and compliance education, pushed to all employees, is often considered mediocre (and not that effective). What if instead we focused on the MVA?

The MVA for E&C education can often be our supervisors. They usually represent 10-15% of the employee base, having on average 8-10 direct reports each. It’s easier to tailor education for this small group, and it’s easier for them to pass the information on to their direct reports, whom they know well.

I’m not suggesting that all E&C education should be in the form of supervisor-led training (SLT). But I do believe that SLT has its place, and that not enough companies are making a good use of it.

Today I asked GPT3 to write a 200-word blog post about business ethics. It wasn’t original but it was surprisingly good.

In a few years, AI will be able to write codes of conduct, corporate policies, training modules, and probably answer many ethics questions from employees.

E&C professionals should embrace this fact and anticipate the changes it will create. The creative aspect of our job is what still remains out of reach for machines.

Do you know how to boost your creativity? GPT3 just served me 9 suggestions.