Too many leaders still treat sales compensation like a set of payout mechanics (or worse, a cost center), when it's actually one of the few levers that can successfully change behavior at scale, creating predictable revenue.

Which is why this episode of The Sales compensation Show with Sid Ganguly, Global Comp Partner & Head of Sales Incentive Design at Expedia Group, will resonate.  

Expedia Group—a company most of us don’t get to peek inside—has enough scale and complexity to stress-test any compensation philosophy. And having designed comp across Dell/EMC, VMware, AWS, and now Expedia Group, Sid bringing a rare mix of systems thinking and field pragmatism to his role.

With experience in cloud-first environments, navigating complex org transitions, and now operating with a broader total rewards lens, Sid’s view is that sales comp is how you operationalize trust. If sellers can’t predict outcomes, trust erodes. And when this happens, everything else (enablement, forecasting, retention) pays the tax.

From aligning comp to strategy, to the hardest balancing act in the game, Nabeil and Sid's discussion unpacks what really holds up in large, complex orgs. They also dig into risk posture, and the cross-functional reality of getting comp design right with Finance and Field Ops in the room.

They even tackle the question some comp leaders are quietly beginning to ask: can AI unlock more dynamic, individualized incentives at scale? And if so, what needs to be true first?

Below we've curated the most actionable ideas from this episode, so you can benefit from the convo even if you can't queue it up right away.

‍Episode resources

• Connect with Sid—a founding member of our Sales Comp Think Tank— on LinkedIn  
• Book recommendation: Debt: The First 5,000 Years by David Graeber

Alignment above all else (aim to steer just one ship)

Throughout this discussion, Sid underscores that incentives aren’t just payouts, they’re distributed instructions.

If your instructions contradict the executive narrative, the field isn't “misunderstanding.” They’re simply doing what you paid them to do.

And as you've likely experienced first-hand, misalignment rarely shows up as a single giant failure. It appears as:

• a pipeline shape you didn’t intend
• a product mix drifting from priority bets
• a territory strategy that looks “rational” for reps and “irrational” for the business
• a comp team drowning in exceptions because the plan is fighting the strategy

As Sid observes, a pattern emerges where teams with great intent optimize locally…all while the C-suite is intending to steer globally in a totally different direction.

But, alluding to his time at AWS where they had 200+ products, Sid maintains it's possible to lead “one ship, heading one direction.”  

A practical tool: Try the “One-ship test”

Overall, it's critical your decisions and prioritization as a leader reinforce the org’s wider strategy. In a working session with Sales, Finance, and RevOps, answer these five questions:

1. What is the company optimizing for this year? (Growth? Profitability? Retention? Penetration? New logos?)
2. Which field behaviors most directly create that outcome? (not KPIs—behaviors)
3. Does the plan explicitly reward these behaviors?
4. Where might the plan accidentally reward the opposite? (e.g., cross-sell when you need greenfield)
5. What would a rational rep do if they only cared about their payout? If that answer makes you flinch—you’ve found the gap.

Sid’s personal prioritization approach is simple and ruthless: when he has five comp issues on the table to manage, he tackles the ones that move the needle against strategy in the next 3, 6, 12 months—and parks the rest without pretending they aren’t valid.

On “risk appetite” as a missing design input

Another observation in this discussion is that many orgs talk about comp as if it’s purely mechanical: pick measures, set rates, choose thresholds.

But Sid notes that it's often more about organizational psychology: plans are typically tuned to culture, leadership, and role expectations.

And, in reality, teams thrive on very different risk profiles:

• some team members respond to stable, clearly articulated pay (you get ZYX for doing ABC)
• others thrive in high-risk, high-reward environments
• and the same structure can succeed or fail depending on where a team sits on the “risk appetite” curve

This has a huge impact on who the org needs to hire according to the risk profile that aligns with the strategy overall. In today’s market (budget scrutiny, changing growth motions, shifting product bets), the cost of mismatched risk is huge:

• If you have too conservative a plan → sellers won't push into the messy growth work
• With too aggressive a plan → attrition spikes, trust erodes, and you end up “fixing” outcomes with exceptions.

The fix? Before you touch plan mechanics, build a “risk profile map”

For each role segment (or geo, if needed), align on:

• what success requires (hunting? farming? expansion? retention?)
• what the org can tolerate (variance, overachievement risk, underperformance risk)
• what seller profiles you want to attract/retain (and whether your pay mix matches this reality)

Sid’s “total rewards lens” matters here: he noted the shift from changing “gears” (tweaking KPIs) to changing the “engine” (pay mix, leverage, role design, and risk posture). This is where comp starts acting like a leadership tool versus a payout formula.

‍Design for trust, not maximum precision (and keep complexity on a leash)

During the episode, Sid shares something comp leaders should print and tape to their monitor:

“We’re not paying money. We’re generating trust.”

And, as he sees it, the more complex a plan becomes, the more trust dilutes because uncertainty grows.

Sid’s vivid example is the seller doing commission math mid-customer conversation. At that moment, the plan isn’t motivating performance. It’s hijacking attention:

Here, Sid and Nabeil unpack the real cost of complexity: when sellers spend mental energy interpreting payout math, they’re not spending it creating customer value.

The fix? Create a “complexity budget”

Practically, you can aim to treat plan complexity like engineering treats latency: you only spend it where it pays you back.

This is where you decide upfront:

• What is structurally fixed in the core plan this year? (core measures, primary crediting logic, role families)?
• What are the approved “release valves”? (SPIFFS, spot bonuses, one-time adjustments)?
• What requires exec-level approval to change mid-year?

And, while you're at it, add a simple rule: if a new lever requires a calculator explanation, it must earn its keep.

At the end of the day, sellers don’t need a perfect model. They need confidence that if they do the job, they’ll get paid—cleanly, predictably, and fairly.

AI doesn't justify completely dynamic pay just yet. But it does change what’s possible in comp design

When the conversation turns to AI, Sid shared a grounded take that avoids both hype and fear.

Nabeil teed up a future state that most comp leaders have thought about, but few say out loud. That is: 14,000 unique comp plans designed by humans is indeed chaos—but a system that can generate individualized incentives consistently might actually be the cleanest form of fairness. 

Not because customization is the goal. But because the “same plan for everyone” often ignores the reality of different:

• territories
• seller strengths and tenure
• market conditions
• motions (hunter vs farmer vs hybrid)
• capacity constraints across geos

Ultimately, as Nabeil sees it, the future isn’t “one plan versus 14,000 plans.”

It’s designing the rules engine that can produce incentives at scale—intelligently, consistently, and explainably. Something that doesn’t collapse into one-off exceptions.

But while “AI-generated individualized comp plans” sounds elegant in theory, Sid is skeptical about its application near-term—both because the behavioral data isn’t entirely accessible, and because comp teams deal with PII and must be responsible about what data touches external systems:

Where Sid lands? He's keen on AI as an input engine (with guardrails), not an autopilot.

Sid is keen on AI as a platform to absorb more signals than comp teams typically use today and convert them into better inputs for comp plan design:

• pattern detection across segments and roles
• scenario modeling to stress-test levers before rollout
• behavior indicators tied to risk tolerance and performance
• “what actually moves outcomes” beyond blunt input/output metrics

And then Sid shared an operating principle every comp leader should staple above their AI roadmap:

“Garbage in, garbage out.”

Ultimately meaning: AI doesn’t rescue messy systems. It'll only amply them.

The actionable takeaway? Earn your way toward personalization

For sales compensation leaders, the path ahead on incorporating AI is staged, not binary:

• Start with data quality + governance (especially with PII and payout data)
• Use AI to consolidate and interpret what you already have (CRM, pipeline, attainment, tenure, segment, role changes)
• Stress-test incentives through scenario modeling before you roll anything out
• Build “sense-and-respond” capabilities when external shocks hit (market shifts, region disruption)

This is the real shift Nabeil is pointing toward: comp moving from an annual planning artifact to an increasingly responsive system. Wherein decision-making with the help of AI can finally become faster, better informed, and more explainable.

AI won’t save a broken planning engine. But it will make strong engines faster and over time, it'll create the conditions where thoughtful individualization becomes feasible.

Ultimately Sid advocates for the opportunity where AI first helps you see what’s driving performance (behavior → outcome), before it helps you automatically pay for it.

The most strategic comp teams don’t chase perfection, they protect alignment

Sid’s perspective is a reminder that mature incentive design isn’t about piling on nuance. It’s about building a system sellers trust and leaders can defend.

If you’re heading into a plan cycle (or stabilizing one mid-year), steal Expedia Group’s underlying operating principles from Sid:

• Map risk appetite before mechanics
• To avoid needless complexity, build a shared comp structure, then layer controlled levers
• Use comp to clarify strategy, not to patch local drift
• Treat trust as the real output (without it, you're back to square one)
• And use AI to accelerate insight, not to automate accountability

Want more insights like this? Subscribe to The Sales Compensation Show on Spotify or Apple Podcasts, or YouTube for bi-weekly episodes featuring the revenue leaders behind today’s fastest-growing companies.