The Decision Intelligence Gap: Good Intentions to Consistent Execution

Every organization knows what a good decision looks like on paper. Far fewer can say with confidence that the right decision will actually be made at 4:59 on a Friday, when the pressure is on and the policy is twenty pages long. That distance — between what we intend and what we actually do — is the decision intelligence gap, and it is the subject of this week’s Ethicsverse session.

Host Nick Gallo of Ethico was joined by Hema Lomax, author of the forthcoming book The Decision Intelligence Gap, and Chelsea, a former attorney turned compliance leader. Together they dug into why the gap persists even in mature, well-resourced programs, and what it takes to close it — not by writing more rules, but by building better decision-making. The conversation ranged from toddlers learning to walk to the Delaware business judgment rule to Waze, but it kept returning to one idea: the last mile of a great ethics and compliance program is making good guidance invisible and instant, delivered in the moment a person actually needs it.

Key Takeaways

The Gap Is the Distance Between Intent and Operational Reality

Hema defined the decision intelligence gap as the distance between an organization’s stated intent and its operational reality — the space where good values quietly fail to translate into consistent action.
There is a deliberate pun in the acronym: DIG. Digging matters because we rarely make our own reasoning — the path between intent and action — inspectable, even to ourselves.
The timing is pointed. As organizations rush to export human judgment into AI tools, we first have to understand that judgment ourselves. Human judgment is not an intangible mystery; it is something we can inspect, and this is the moment to do it before we scale it into machines.

Why It Persists Even in Mature Programs

Chelsea explained that employees constantly encounter situations their policies never clearly addressed. No policy can anticipate every gray zone.
Even A-plus policies fail when they are long, inaccessible, and hard to navigate. Nobody facing a high-pressure decision at 4:59 on a Friday has time to read a twenty-page document to find the one line that applies.
A program can be excellent on paper and still leave people stranded in the moment. The failure is often not the policy’s content but its usability when urgency is built in.

It’s an Opportunity, Not a Threat

Asked whether the gap is a threat to the profession or an opportunity, Hema was firmly “team opportunity.” This is not just about compliance — it is about decision-making writ large.
She offered a reframe: the old line is “the only constant is change,” but the truer version is “the only constant is uncertainty.” You are never 100 percent certain of anything; you take calculated risk all the time.
The successful companies we admire are defined by the big bets they took under uncertainty. Compliance, as a field, is often reluctant to sit with that unease — and then exports its own insecurity to employees, amplifying their survival brain until they are afraid to act without sign-off.

The Curb Metaphor: Fear Creates Two Bad Decision-Makers

Hema described telling her three daughters they can’t cross the road “because you’ll die” — the easy, threat-based shortcut. It produces two kinds of children: one who will never cross without her, and one who will always cross the moment she isn’t looking, and probably too fast.
Organizations do the same thing with employees, creating reckless renegades on one side and low-agency dependents on the other — both the product of a fear-only system.
The goal is not to stand at the curb holding every employee’s hand for every decision. It is to teach them to cross safely, swiftly, and strategically on their own, and to come to compliance only for the mega-highway or the broken bridge — not every routine choice.

Build High-Quality Decision-Making, Not High-Quality Decisions

Hema drew a sharp distinction: she is not interested in high-quality decisions, but in high-quality decision-making. If you fixate on the decision, you reverse-engineer your judgment from the outcome.
A decision muscle means thinking about the operating conditions, being clear about what you are optimizing for, and staying open to coaching and ideation so nobody is flying solo.
Compliance too often “punishes the outcome and assumes the process.” When something looks like a violation, an investigation returns a finding and labels the person — even if the real problem was that the policy wasn’t clear, wasn’t fit for purpose, or was never truly understood.
The alternative is to allow bets that fail. A high-quality process that doesn’t get the outcome you wanted is still a good process — and the missed step is that we rarely go back to the disciplined employee to explain what happened and extend a little grace.

From the Office of No to Unlocking Responsible Growth

Everyone says they don’t want to be the “office of no.” The stated intent is to be the office of unlocking responsible growth — but you have to examine what you are actually optimizing for.
If you are optimizing for what you think the regulator might want, you are probably not optimizing for high confidence in the partners and decisions you actually make. Naming the real objective is the first honest step across the gap.

The Business Judgment Rule — For Everyone

Hema invoked the Delaware business judgment rule: courts will not punish directors or executives for a bet that goes wrong if they can show well-reasoned judgment based on the information available at the time.
The jurisprudence already protects good reasoning over good outcomes for the C-suite. Her question: why not extend that logic to employees? If a person can show high-quality reasoning — the same instinct behind a business-justification memo for a gift — that should cover the organization when a defensible bet doesn’t pay off.
She shared a personal example: on a big decision, she asked a teammate how confident he was, he said 95 percent, and instead of fixating on the missing 5 percent, they inspected their reasoning, accepted the odds, and moved forward. The 5 percent has partly materialized — and they are still fine with it, because the odds were transparent and expectable going in.

Design Humans — and AI — to Learn

Responding to an audience question about whether AI is just “throwing a dart at the dartboard blindfolded,” Hema argued that AI, like a hedge fund, suffers from a trust problem — and the fix is the same as for humans: make the reasoning inspectable.
Both AI systems and people should be designed to learn: take the bet, then ask what could have been done differently, what should have stayed the same, and what would have made no difference.
We are good at this as toddlers. When a child falls while learning to walk, we don’t declare walking off-limits — we create a safe space, mitigate the risks (round off the sharp table corner, add gates, watch closely), and let them learn. We don’t walk for them. Hema tied this to Viktor Frankl’s writing on the gap between stimulus and response: the space between intent and output is where the real work lives.

Meet People in the Moment: Embed Decision-Useful Guidance

Chelsea’s answer to “what do you do when people have every cheat sheet and still go ask five colleagues and get five different answers?” was to embed prompts and decision lanes directly where decisions happen — a renewal workflow that asks, in the moment, whether due diligence was done or whether compliance pre-approval was obtained.
Real-time, in-workflow reminders beat expecting employees to remember to hunt down an SOP. The failure point is the moment someone turns to a neighbor instead of the guidance — so fix the system, don’t just blame the person.
Hema extended this with her favorite analogy: Waze. It crowdsources the road ahead and gives you decision-useful information for the specific turn you’re making — practical wisdom in the moment, not a mandate to memorize the entire highway code. People turn it on voluntarily because it is genuinely useful, not because they were made the villain for driving without it.

Stop Bothering Everyone to Catch the Few

Hema challenged the reflex to survey, test, and question the entire population to find a small percentage. The lobbying questionnaire sent to every employee every month, or blanket compliance training “just in case,” disrupts everyone to reach a few.
A better question: what would it look like if people had the capacity to flag when they are actually in scope? The shift is from knowledge transfer, which the profession is already good at, to capacity building.
She offered a metric-check: if 91 percent of people pass your “challenging” training, the win isn’t saved staff time — it’s a signal to change the training next year into something people actually need to learn.

Make Policies Bookmark-Worthy: Motivation, Ability, Prompt

Chelsea described a quarterly, bite-sized compliance newsletter, timed to the calendar — gifts and entertainment guidance landing right before the holidays — that links to the full policy but leads with a digestible reminder of what to do.
Hema mapped this onto BJ Fogg’s behavior model (of Tiny Habits fame): behavior happens when Motivation, Ability, and Prompt converge. Chelsea’s work covers ability (make it easy) and prompt (deliver it at the right time); motivation is what makes someone bookmark a policy because it was genuinely useful, not because they’ll be fired.
The practical design: put the three things people need to know and the three things they need to do at the very top of the policy. Keep the detailed, legally precise material underneath for the moments — and the regulators — that require it. And in an era of policy chatbots, you no longer have to keep policies short; you have to make them readable for the AI, using retrieval-augmented approaches so the bot answers accurately without hallucinating.

Compliance as Catalyst, Not Cost Center

Chelsea framed the evolution as becoming a thought partner rather than a gatekeeper: instead of being the final “let me get approval from compliance” step, get invited in at the beginning, when the guardrails are still being drawn.
The path there is visibility — get on the committee, get in the meetings early, and make sure the organization knows who you are and how you can help. Usefulness earns the invitation.
On spotting trouble early, Hema reframed “warning signs” as “intelligent insights trapped in the process.” The language people use to bypass a control — “we’ve used them before, it’ll be fine,” “the approval takes too long” — is information the system is handing you, if you widen your aperture to hear it.

Boldness Is Decision Intelligence in Action

Both speakers closed on boldness. Chelsea’s version is simple: make yourself visible and ask. If you wanted to be invited to a hundred meetings, how many times would you actually hear no? Roughly zero. Become the gut-check person people want in the room.
Hema uses retrieval cues to inhabit the mindset — “remember who you are,” “what else could be true?” — and reframes the fear of seeming self-promotional as a calculated risk: a 5 percent downside against a 95 percent chance of building trust and impact.
Her final line captured the whole session: boldness isn’t luck, instinct, or blind faith. It’s decision intelligence in action — inspecting your reasoning, accepting the odds, and moving forward anyway.

The throughline from start to finish was a call to spend real time in the gap between intent and execution — to make our reasoning inspectable, design our organizations (and our AI) to learn, and deliver practical wisdom in the moment people actually need it. Hema’s forthcoming book, The Decision Intelligence Gap, promises to go deeper on exactly that terrain.