Setting Compensation Triggers: When to Act on Market Pay Changes

For decades, comp ran on an annual rhythm. You bought a survey, aged the data, built ranges, ran your cycle, and repeated. The market moved between surveys, but you only saw it move once a year. That had a strange built-in protection: you couldn't overreact to a number you couldn't see.

Live data takes that protection away. You can now watch a job family shift week to week, and the first time a recruiter asks to go above range "because the market moved," you can actually check. The data says the market is up 6%. 

And then everyone in the room has a different read on what that means. Someone calls it noise. Someone calls it the start of a trend. Someone's worried about the budget. The number is sitting right there, and the conversation still stalls.

That's the trap. Visibility is not the same as a plan. The teams getting the most out of live data are not the ones reacting fastest. They're the ones who decided in advance when to act, when to wait, and who gets to make the call. This piece is about doing that work before you need it.

First, decide how live data fits

Before you set a single trigger, get clear on the role live data plays in your benchmarking. There are a few ways to use it, and they're not mutually exclusive.

• You can use it as a primary source, alongside or eventually in place of traditional surveys. 
• You can use it as a signal layer, a way to see how compensation and labor markets are shifting and to identify emerging jobs and skills. Compa’s Offers data provides early signals when a market is changing.
• Or you can weight it by condition, leaning on it most heavily in markets seeing significant demand or movement, and increasingly the further you get from your last point-in-time source.

Most teams land on some blend. The point is to name yours, because how you use the data shapes every trigger you set on top of it.

Anchor your triggers to your talent strategy and compensation philosophy

Here's the part that's easy to skip: live data doesn't have a strategy. How you respond to it has to come from your talent strategy and comp philosophy, not from the data itself. 

In practice, we see a range of postures, plus a factor that cuts across all of them:

• Top-of-market teams react fast to upward movement. Companies paying at the top of market, or carrying a small number of high-leverage critical roles, treat live data as a continuous input. 
  
  • For example, if the market has moved up and they’re now below the desired 75th percentile positioning from their compensation philosophy, they open a review. 
  • They’ll make off-cycle adjustments to protect their stated position. They're slower to react when the market slows. 
  • Note - a faster reaction is also typical for frontline roles, where talent will move for as little as a fifty cents an hour increase.
• Threshold-based teams react to conditions. A more balanced posture: move once a sustained trend shows up, or once the market has shifted by a material amount. 
  
  • The work is defining “material” in advance (more on this below). For example, market offers running more than 8% above midpoint for three consecutive months, or any critical or high potential employee is more than 10% below the latest market data. 
  • This works for companies and areas with a mix of high and average performers. They watch for sustained shifts rather than weekly movement, and they bias toward annual recalibration unless something material moves.
• Budget-constrained teams. Reframe the question from “peanut butter adjustments” to “where do we actually spend the budget that we have?” If a role is up 15% and critical to the business, that is where the investment goes. Live data turns into a strategically sound allocation instead of a guess. 
• Manage exceptions only (the most conservative). Use live data to handle exceptions, but don't change employee pay or structures off the back of it. This suits companies focused on stability or sitting on deep talent pipelines. Be honest about the cost, though. Held too long, it risks pay equity problems, slower hiring, higher turnover, and widening market gaps.
• When your ranges are public. Pay transparency adds another layer. In jurisdictions where you’re posting salary ranges, stale data creates visible problems on both sides. Ranges built from data that’s six to twelve months old can be too low to attract qualified candidates, which shows up immediately in your application volume and top-of-funnel drop-off. They can also be inconsistent with what you’re actually paying internally, which creates legal and employee relations exposure. Live data should be a part of the range audit before anything goes public, not an afterthought once candidates start pushing back. 

None of these is the right answer on its own. The right answer is the one that matches your employee value proposition and how you've already decided to pay.

Defining parameters

A few parameters are worth pinning down before you need them:

• What size of move triggers action? A 1% shift is noise. A 5% shift in a single month might be a watch item, or it might be the start of a sustained trend. Where's your line, and does it differ by job family? You can ask the same question from the other side: how far above or below market can your pay drift before you act?
• How sustained does the move need to be? A two-week spike on a thin data sample is not a three-month trend. Saying that out loud in advance keeps you from chasing every blip.
• Write these down. A threshold you defined in a calm moment is worth far more than one you negotiate under pressure.

One clean tool for the first question is the Offer/Employee Ratio: median offer divided by median employee pay for a given role, level, and geography. 

• A ratio above 1.0 signals a heating market, with rising exception pressure and attrition risk. 
• Below 1.0 suggests a cooling market and the potential for overspend. 

Monitored consistently, these give you and your leadership a clean reading rather than a fresh debate every time. The catch is the word "consistently," and that's the thing worth automating.

Let Compa Agents do the watching

All of this raises an obvious problem. Live data is exciting right up until you realize someone has to actually watch it, interpret it, and decide what happens next. That isn't sustainable by hand across hundreds of roles.

The monitoring layer, though, is exactly the kind of work our Agents are built for through Compa Watchlists. A typical sequence looks like this:

1. Monitor live data against the parameters you set.
2. When a threshold is crossed, check your ranges. Are they still defensible?
3. Compare to your actual employees. Who's now below market, and by how much?
4. Document the finding, decide on an action, and route it to the right decision-maker.

The first three steps are repetitive and rule-based, and they reward speed. The fourth, deciding what to actually do, is where human judgment earns its keep. The teams getting live data right are offloading the watching so they can spend their attention on the deciding.

Live Data Rewards Preparation, Not Reaction

Live data doesn't tell you what to do. It tells you something changed. The work that makes that useful happens before the change shows up: deciding your posture, setting your thresholds in both directions, and handing the watching to something that won't get tired of it.

Do that, and the next time the market moves 6%, you won't be debating what it means. You'll already know.

Stay tuned for Part 3: a deep dive on how to respond when the data moves materially, matching the move to the moment, and planning for the case most teams put off - the market that's falling.