Marketplace
You connect buyers and sellers.
The Bias Advantage
Free with VIP Club membership
Executive Brief
Where AI will disrupt your business — and where it won't.
Five questions, a ten-minute diagnostic, and a shared language for the AI conversation with your team.
Adapted from The Bias Advantage by Liat Ben-Zur
LBZ Advisory
Inside this brief
- 1Five questions every team should answer
- 2Where value moves when information is free
- 3A ten-minute AI exposure test
- 4The human advantage inventory
- 5Toolkit roadmap
Purpose
Most leadership teams know AI will change their business. Fewer can clearly state which parts of the business are exposed, which parts are defensible, and where human judgment still creates value that AI cannot copy.
This brief gives your team a sharper way to have that conversation. It includes five questions for leaders, a short explanation of where value is moving, a ten-minute diagnostic for one revenue stream, and a map of the full toolkit that gets unlocked with the purchase of five or more copies of The Bias Advantage. You can use this brief to help facilitate AI strategy conversations in your board meeting, offsite, or next business review.
1
Five questions every leadership team should answer
These five questions are designed to help you understand whether AI is about to disrupt your business, and where your real advantage sits. Work through them with your leadership team before a competitor forces the conversation for you.
01What are customers really paying us to do?
Put aside the mission statement. Strip away the brand language. Name the actual job the customer pays you to perform.
Most businesses make money in one or two of four ways. AI is already aggressively disrupting two of them — controlling access to information, and packaging what already exists.
Gatekeeping
You control access to information.
AI is disrupting thisCuration
You package or synthesize what already exists.
AI is disrupting thisEfficiency
You sell speed and convenience.
Weak answer
"We deliver world-class solutions."
Useful answer
"Customers pay us to connect them with suppliers they can trust."
02Can AI do the core work cheaper, faster, or well enough?
This is the question underneath the whole discussion. AI is unusually good at a narrow set of valuable tasks: finding information, producing first drafts, spotting patterns in large datasets, and running routine analysis. If the center of what you sell sits inside that set, you should treat it as highly exposed.
Do not ask whether AI can do the work perfectly. Ask whether it can do enough of the work for a customer to question your price. A competitor does not need to replace you cleanly. They only need to make your premium look unnecessary.
The useful version of this answer names the part of your business that depends most on speed, volume, or access to information, and makes an honest estimate of how quickly it could be replaced.
03Where does AI win, and where do we win?
It helps to draw the line explicitly.
Where AI wins
- Speed
- Volume
- Pattern recognition
- First-draft analysis
vs
Where we win
- Judgment under uncertainty
- Ethical reasoning
- Genuine relationships
- A feel for context
The common mistake is to keep competing on AI's side of that line, trying to be faster or cheaper at the very things it does best. The more durable move is to shift the contest onto the human side. As the book argues, this is a matter of redesigning where you compete, not of sharpening your marketing.
This is a strategy question, not a messaging exercise.
04What can we do that AI cannot, and where is the proof?
It is easy to claim your competitive differentiation or your "human advantage." It is harder to prove it. "We are a human-centered company" is a claim, not evidence. Evidence is a specific thing your people did that a model could not have done.
Look for proof like this:
- a decision made under real uncertainty that proved right for reasons the data did not yet show;
- a moment when someone overruled an efficient and profitable AI-powered solution because they sensed the second-order effects could be far more detrimental, though not yet actualized;
- a relationship that survived a serious failure or fire because the interpersonal relationships were strong, and trust had already been earned;
- an idea that broke an assumption your industry treated as settled.
If you can point to the moment, the advantage is real. If you can only point to the value statement, you have an aspiration.
05Are we promoting the people this era actually needs?
The AI era rewards traits that stable hierarchies often miss: curiosity, willingness to abandon old ways of working, comfort acting with incomplete information, and the habit of seeing second-order effects before they show up in a dashboard.
Some of the people with those instincts may not look like the obvious heirs to power. They may be the ones who learned to read weak signals, build trust without formal authority, and recognize unspoken biases — because they had little choice. The question worth asking is whether your structure advances those people, or continues to reward fluency in a playbook that is already being disrupted.
2
Where value moves when information is free
AI makes information cheap. It can gather it, sort it, summarize it, and remix it at speed. All this makes judgement even more valuable. Business value now moves toward work that requires taste, context, trust, and consequence. The book describes this value migration in four shifts. Together they give a leadership team a shared vocabulary for deciding where to invest its people's time and resources.
The new value hierarchy
As AI floods the bottom with cheap information, human value climbs to the top.
01
From information to insight
Facts are no longer the scarce asset. AI can retrieve them quickly. The advantage moves to the person who can frame the right question, interrogate the answer, and turn it into a decision that holds up in practice. That means catching the bias in an AI's output, checking its sources, and supplying the context a model cannot see: the incentives, the trade-offs, and the edge cases.
02
From volume to discernment
AI will flood every market with more content, more analysis, more options, and more noise. The scarce skill becomes judgment about what deserves attention. Who can separate the most critical signals from the noise?
The valuable executives become those who can review forty metrics and identify which two matter, which one is misleading, why, and what decision should follow.
03
From competence to mastery
AI is good at broad competence because it learns from what already exists. What it cannot easily reproduce is lived depth.
A junior manager with a strong prompt can produce a decent plan. A veteran knows which customer segment will revolt, which constraint will break at scale, and which obvious idea will erode trust six months later. The book calls that scar tissue. This goes beyond pattern matching as it comes from lived experience. It is accumulated consequence.
04
From transactions to trust
AI can draft the proposal, summarize the account history, and calculate the pricing. It cannot sit across from a nervous buyer whose career depends on the decision and earn the sentence, "I trust you to deliver when this gets messy."
The same is true inside the company. After a round of layoffs, a model can write the talking points. It can suggest a new organizational structure. But it cannot look a team in the eye and rebuild belief. The higher the stakes, the more trust becomes the differentiator.
"Value moves to those who can frame the right problem, interrogate the answer, and translate it into a decision that holds up in the real world."The Bias Advantage
3
A ten-minute AI exposure test
This exercise condenses the Arbitrage Audit, the first tool in the toolkit, into something you can complete in about ten minutes. Choose one revenue stream. Do not average across the whole company. Pick a specific offer, product line, or service and assess it honestly.
Step A · 2 minutes
Name what your customers actually pay for
Write one sentence that describes the specific activity a customer pays you to perform. No mission language. No brand promise. If the sentence gets too long or too vague, rewrite it until it is crisp and clear.
Weak
"We empower customers with innovative, end-to-end solutions that transform their business."
Better
"Customers pay us to reduce the time it takes their finance team to close the books each month."
Step B · 3 minutes
Identify the kind of value you sell
Mark the type or types that describe how this revenue stream earns money. Most businesses rely on one or two.
- Marketplace: you connect buyers and sellers.
- Gatekeeping: you control access to information.
- Curation: you package, rank, interpret, or synthesize what already exists.
- Efficiency: you provide speed, convenience, or operational ease.
The greatest AI disruption exposure usually sits in gatekeeping and curation, because AI now performs much of that work at little cost.
Weak
"We sell trust, innovation, and quality."
Better
"We sell curation and efficiency. Customers pay us to sort through hundreds of options and get to a usable recommendation quickly."
Weak
"We are a technology-enabled services company."
Better
"We sell efficiency. Customers pay us to complete a manual workflow faster and with fewer errors."
Weak
"We monetize our proprietary ecosystem."
Better
"We sell marketplace access. Suppliers pay us to reach buyers they could not easily reach on their own."
Weak
"We are in the knowledge business."
Better
"We sell gatekeeping and curation. Customers pay us because we have access to information they do not have, and because we turn that information into a ranked set of choices."
Weak
"We help leaders make better decisions."
Better
"We sell curation and judgment. We gather the relevant data, pressure-test the options, and recommend the move we believe the leadership team should make."
Step C · 3 minutes
Score your compression risk
For each value type you selected in Step B, ask the central question: can AI generate roughly 80 percent of this value for free, or close to it? Score each from 0 to 3. (0 = AI cannot meaningfully do it · 1 = rough version, but trust/regulation/complexity protects you · 2 = good-enough version exists today · 3 = AI already does it better, cheaper, or faster and customers know it.)
0Compression
risk score
risk score
DefensibleExposedHighly exposed
Score each type to see where you land.
Only score the value types you actually sell.
0 to 2 · Defensible for now. Your value still rests on ground AI does not own. Revisit this every two quarters.
3 to 5 · Exposed. At least one core activity is already compressing. You should begin moving the offer toward human advantage now.
6 or higher · Highly exposed. Assume a free or good-enough substitute will arrive within a product cycle. Build the offer that would replace you before someone else does.
Step D · 2 minutes
Name the human advantage
Name the one capability AI cannot replicate, and identify the last decision where that capability changed the outcome. If you cannot name a decision, you have a claim, not an advantage.
Weak
"Our advantage is that we are human-centered."
Better
"Our advantage is judgment under uncertainty. Last quarter, we chose not to enter a market that looked attractive in the data because two senior operators saw early signs of channel conflict. Six weeks later, the largest distributor in that market changed terms, which would have destroyed the margin case."
Why it worksIt names the capability and ties it to a decision. It shows the scar tissue behind the judgment.
Weak
"Our people build deep relationships."
Better
"Our advantage is earned trust in high-stakes moments. When a major client rollout failed, the account lead kept the relationship because she had already built enough credibility to tell the truth quickly, own the miss, and negotiate a recovery plan before procurement reopened the contract."
Why it worksIt does not claim relationship strength in the abstract. It points to a moment where trust changed the business outcome.
Weak
"We have a strong ethical culture."
Better
"Our advantage is ethical reasoning before harm reaches the customer. In a pricing review, one manager caught that the recommended model would have raised costs for smaller customers who had fewer alternatives. The team changed the rollout, even though the original plan would have produced more short-term revenue."
Why it worksEthics is easy to claim and hard to prove. This answer proves it through a trade-off.
Weak
"We are more creative than our competitors."
Better
"Our advantage is reframing the problem. Instead of building another dashboard for store managers, the team changed the workflow so the system flagged only the three actions that mattered that day. Adoption rose because managers did not need more information. They needed fewer decisions."
Why it worksIt shows creativity as a business decision, not a personality trait. It also makes the AI point sharper: more output is not always more value.
Weak
"Our team understands the customer."
Better
"Our advantage is contextual intelligence. The data showed that customers were abandoning the product during onboarding, but the customer success team knew the real issue was fear of looking incompetent in front of their own teams. We changed the onboarding sequence, gave managers language to introduce the tool, and retention improved."
Why it worksIt connects numbers to lived context. AI can see the drop-off. The human advantage is knowing what the drop-off actually means.
4
The human advantage inventory
The book identifies five human capabilities that AI cannot currently, and may never, authentically replicate. What this section asks of you is honesty about the difference between what feels uniquely human and what genuinely is. The test is whether your organization can prove where it has the human advantage.
Strategic judgment
What it isMaking a consequential decision with incomplete information, while seeing the second-order effects others miss.
Why AI strugglesAI works from patterns in existing data. It struggles when there is no clean precedent, or when its input data is incomplete or inherently biased. It often cannot read the ripple effects that are not yet showing up in the data, but which seasoned leaders can predict.
Proof to look forOne decision your people made under genuine uncertainty that proved right — a market bet, a hire, or a product pivot that proved right for reasons the data could not yet show.
Good example"Our advantage is strategic judgment under uncertainty. When the data showed strong demand for a new enterprise product, the team still delayed launch by one quarter because sales leaders saw that customer success could not support the implementation load. The decision looked conservative at the time. It proved right when two early pilot customers needed far more hand-holding than the model had forecast. Had we launched at full speed, we would have created churn inside the very accounts we were trying to expand."
Ethical reasoning
What it isRecognizing fairness, harm, and accountability in unfamiliar situations, and catching harm before it reaches a customer, employee or other stakeholders.
Why AI strugglesAI carries the bias and blind spots in its training data. It cannot reliably notice harm it was not built to see.
Proof to look forA time someone overruled an efficient answer because it was wrong, not because it was unprofitable.
Good example"Our advantage is ethical reasoning before harm reaches the customer. In a review of an AI-assisted credit model, the output recommended faster approvals for applicants with longer digital histories. On paper, that improved efficiency and reduced fraud risk. One leader caught the problem: the model was quietly disadvantaging younger customers, immigrants, and people with thinner digital footprints. The team paused the rollout, added human review for edge cases, and changed the approval criteria before the product reached the market."
Deep relationship-building
What it isEarning trust through candor, consistency, and the ability to read what someone needs but will not say.
Why AI strugglesAI can summarize a client's history and draft the message, but it cannot sit across from a nervous buyer whose career depends on the decision and earn the confidence that closes the deal. After layoffs, it cannot look a team in the eye and rebuild belief.
Proof to look forOne relationship that survived a serious failure because trust had already been earned.
Good example"Our advantage is trust earned before the crisis. During a major system rollout, a client's implementation went badly. Deadlines slipped, the client's internal sponsor was exposed, and procurement started asking whether to reopen the contract. The account lead kept the relationship because she had built credibility long before the failure. She told the client exactly what went wrong, named what the company would fix, and stayed close through the recovery. The client did not leave because the trust account had already been funded."
Creative problem-solving
What it isBreakthroughs that come from questioning the status quo and rethinking stale assumptions instead of merely optimizing the current model or improving workflows that already exist.
Why AI strugglesAI learns from what exists. It can remix the frame, but it rarely questions the frame that has been provided as context or training data.
Proof to look forOne idea from the past year that challenged an industry rule and actually shipped. The book points to Perplexity as an example: it did not try to build a better Google page; it changed the way people expect answers.
Good example"Our advantage is reframing the problem. The team was asked to build a better dashboard for regional managers. Instead, they challenged the assumption that managers needed more data. Store leaders were already drowning in data. What they needed was a short daily prompt telling them which two actions would matter most: call this customer, move this inventory, fix this staffing gap. The team shipped the prompt-based workflow instead of the dashboard, and adoption jumped because they solved the real problem, not the requested one."
Contextual intelligence
What it isReading the story behind the number and the unspoken politics in the room.
Why AI strugglesAI misses lived depth, hidden incentives, and the quiet signals that experienced operators notice. A junior manager with a good prompt ships a competent plan; a veteran knows which segment will revolt and which obvious move erodes trust six months out.
Proof to look forThe person whose read on a room, customer, or metric repeatedly beats the dashboard.
Good example"Our advantage is contextual intelligence. The dashboard showed that usage dropped sharply after onboarding, so the first answer was to simplify the product. But the customer success team knew the product was not the real problem. New users were afraid of looking incompetent in front of their own teams, so they avoided the tool after training. The company changed the rollout: managers got language to introduce the tool, teams got a lower-stakes practice period, and adoption improved. The metric showed the drop-off. People who knew the room understood the reason."
5
Toolkit roadmap
This brief is the starting point. The full toolkit contains twelve interactive frameworks across five modules. Each one is built for teams to easily use in a live discussion. The summaries below describe each module in the toolkit.
Module 1AI-Ready Strategy
Assess where AI may disrupt your advantage before the market does.
Arbitrage Audit. Pressure-test a revenue stream you believe is safe and measure how exposed it really is. The audit forces the team to name what customers actually pay for, identify whether the business is selling access, curation, efficiency, or trust, and test how much of that value AI can now reproduce at lower cost. It also pushes the team to name the human advantage that remains after and alongside AI.
Module 2Influence and Power
Build real influence, and understand how decisions actually get made, regardless of the org chart.
Power Map. Map who genuinely makes a decision, who appears to make it, and who quietly blocks it. Useful when something keeps stalling and the reason is unclear.
Three Levers Influence Audit. Identify where your influence comes from: the people who trust you, the results you can point to, and what you are known for.
Trust Ledger. Track trust with the people who matter as though it were an account you can overdraw, so that before a high-stakes effort you know whose trust you have built and whose you have spent.
Module 3Culture Under AI
Keep the team intact while AI changes the work, the leadership and often the culture.
Anti-Fragile Culture Canvas. Don't let AI create culture cracks in your teams. Design team habits that make people sharper as AI changes the work. Produces a 90-day plan.
AI Trust Gap Diagnostic. Find where people have stopped trusting the tools, the decisions, or one another.
Adopter / Resister Bridge. Close the gap between the people racing ahead with AI and those resisting it, when a team has split into two camps.
Module 4Responsible AI
Decide when to use AI, when not to use it, and who owns the consequences.
AI Decision Discipline. Use this when a team is treating AI output as if it came from an oracle. The tool forces one real decision through a simple test: What did the AI recommend? What evidence supports it? What context might it be missing? Who is accountable for the final call?
AI Risk and Harm Check. Use this before an AI tool reaches customers, employees, or partners. It asks the questions teams tend to skip when they are excited to ship: Who could be harmed? What could the model get wrong? What data might be biased or incomplete? Who owns the consequences if the system fails?
Board AI Briefing. Turn scattered AI activity into a clear board-ready view: what is live, what is being tested, where the risks sit, who owns governance, and which decisions need board attention. It moves the conversation from "we are doing AI" to "here is where AI is creating value, here is where it could hurt us, and here is how management is controlling it."
Module 5Leadership for the AI Era
Recognize and develop the leaders this moment actually rewards. These tools help leaders turn real results into a clear case for trust, advancement, and influence.
MAKE IT COUNT Canvas. Use this when strong work is not translating into visible leadership credibility. The tool helps a leader name what they actually delivered, why it mattered, and how to tell that story in a review, promotion case, board interview, or executive conversation.
Leadership AI Fluency Scorecard. Assess leaders against the capabilities AI-era decisions require, and see who is ready and who is not. Some leaders will sound AI-ready because they know the language. Others will be AI-ready because they know how to make the call when the model is useful, incomplete, biased, or wrong.
In closing
Stop speaking about AI in generalities.
AI will not disrupt every part of your business in the same way. It will compress some sources of value, expose weak assumptions, and make certain forms of leadership much more valuable than they used to be. The useful conversation is specific: where are we exposed, what do we still do better, and who inside this organization has the judgment to lead through the shift? This brief gives a leadership team the language to start the conversation. The complete toolkit gives them a way to act on it.