Salary Benchmarking for EA and PA Roles: What It Actually Involves

Most businesses that come to us for salary benchmarking already believe they know the answer. They want the data to confirm what they’ve assumed for a while: that their pay is competitive, that the market hasn’t moved much, and that any attrition they’re seeing is down to something other than compensation.

More often than not, the data tells a different story.

What benchmarking is meant to do

Salary benchmarking should answer one question clearly: where does this role actually sit against the live market, right now, not two years ago. Done properly, it accounts for location, sector, seniority, and the real scope of the role rather than just its title. It should be uncomfortable when it needs to be, because its value comes from accuracy, not reassurance.

How it actually works

Proper benchmarking isn’t a single average pulled from a public salary survey. Those figures have their place, but they’re usually too broad to be useful on their own, blending together roles with very different scopes under the same job title. A useful benchmark starts with the actual remit of the role in question: what the person is genuinely responsible for day to day, not just the title on their contract.

From there, it’s cross-referenced against live market activity, roles currently being recruited, offers currently being made, and counter-offers currently being used to retain people, rather than historic averages that may already be a year or two out of date. Region matters too. A PA role in central Manchester and the same title in central London can sit meaningfully apart, and a benchmark that flattens the two into a single national figure isn’t giving a business anything it can actually act on.

It’s also why we publish a fresh Salary Survey every year rather than treating it as a one-off exercise. Given how quickly this market moves, a data set that isn’t refreshed regularly starts working against the businesses and candidates relying on it. An annual survey gives both sides an unbiased reference point for where the market currently sits, rather than a figure that’s already ageing by the time it’s needed.

Common mistakes we see

A number of recurring issues emerge when businesses attempt to benchmark independently, or have not revisited an existing benchmark for some time.

Benchmarking by title rather than by scope. Two “Senior EA” roles at two different companies can be doing genuinely different jobs. Comparing pay by title alone, without accounting for what the role has actually grown to include, tends to understate what a role should be paying.

Treating it as a one-off exercise. A benchmarking report commissioned once, at the point of hire, has a shelf life. Role scope shifts gradually, market rates shift with it, and a figure that was accurate at the start of someone’s tenure is rarely still accurate two or three years in.

Relying on public salary surveys alone. These are a reasonable starting point, but they’re often self-reported, broad by geography, and slow to reflect sudden market movement. They tend to lag behind what’s actually happening in live recruitment by a considerable margin.

Assuming low attrition means pay is fine. People who are settled, engaged, or simply not actively looking will often stay below market for a period without saying anything. That’s not the same as being paid fairly, it’s a delay before the gap becomes visible.

What we actually find

The most common issue isn’t that businesses have no data. It’s that the data they’re working from is out of date, sometimes by a surprising margin. A salary band that looked competitive eighteen months ago can fall behind quickly, particularly in a market where role scope has been expanding faster than pay frameworks have kept pace. An EA hired two years ago to manage a diary is often doing meaningfully more today, coordinating projects, managing stakeholders, sometimes running parts of the operation, without the compensation review that scope change should have triggered.

Internal salary data tends to be built once and left largely untouched. External market data moves continuously, and the gap between the two widens gradually, without any single event marking the point where it becomes a problem. By the time it surfaces as a resignation, it’s already too late to address cheaply.

The part that’s harder to hear

The uncomfortable reality is this. When we share benchmarking data that shows a business is behind the market, it isn’t always welcomed. Sometimes it confirms a suspicion the business would rather not have had confirmed. Budget constraints are a legitimate factor, and adjusting pay across a role or a team is rarely a small decision.

Disregarding accurate data doesn’t remove the problem, it just delays when it becomes visible, usually in the form of a resignation letter rather than a conversation. The cost of that delay is rarely factored in properly. A business that decides not to act on benchmarking data isn’t avoiding a cost. It’s postponing it, and usually increasing it.

Signs a role is overdue a fresh benchmark

A few practical indicators are worth monitoring, rather than waiting for a scheduled review date that may already have been overtaken by events.

The role has grown since it was last priced. If someone’s responsibilities have expanded, project ownership, stakeholder management, team coordination, and their pay hasn’t been revisited alongside that growth, the gap is likely already there.

It’s been more than twelve to eighteen months since the last review. Given how much movement we’ve seen in Business Support pay recently, a benchmark from two years ago is a historical document, not a current one.

Similar roles elsewhere in the business have become harder to fill. If recruitment for comparable positions is taking longer or requiring higher offers than it used to, that’s live market signal worth applying to your existing team as well as new hires.

Someone on the team has recently left, or given notice, for a role that looks similar on paper. That’s often the clearest and most expensive signal of all, and by the time it arrives, it’s already about damage control rather than prevention.

The real cost of getting this wrong

Replacing a senior EA or PA is expensive in ways that don’t always show up on a single line item. There’s the recruitment cost itself, the weeks or months of reduced executive support during the gap, the onboarding time for a replacement to reach the same level of trust and efficiency, and the risk that the next hire faces the same underlying pay gap and leaves within a similar timeframe. Backfilling a role that left because of pay, without adjusting the pay, tends to be a cycle rather than a one-off fix.

There is also a less visible cost worth naming. A senior EA or PA typically holds a significant amount of institutional knowledge: how the executive prefers to work, how the business actually operates day to day, which relationships need careful handling. None of that transfers with a job description. It has to be rebuilt from scratch with every replacement, and that rebuilding period is where inefficiency returns to a business that believed it had already solved the problem.

In our experience, the cost of retaining a good EA through a fair, timely pay adjustment is almost always lower than the cost of losing them and starting again.

Using benchmarking data properly

The businesses that get the most value from benchmarking treat it as an early warning system, not a box to tick once a year. That means checking role scope against pay regularly, not just at the point someone hands in their notice, and being willing to act on what the data says even when the answer isn’t the one they were hoping for.

Benchmarking only works if the business is prepared to use it. Otherwise it’s just a report that confirms a problem nobody addresses.

If you’d like an honest read on where your EA, PA, or Business Support salaries actually sit against the current market, get in touch with the Lily Shippen team. We’ll give you the real picture, not the one that’s easiest to hear.