Compensation is a complicated topic for both companies and candidates alike. There’s a number of factors to take into consideration when creating (and when evaluating) an employment offer: experience, role / title, benefits, location, company size, cash vs. equity split…
It gets particularly complicated when candidates are trying to compare offers from companies that are very different in size and trajectory. Of course a Series A company’s offer isn’t going to measure up to a Google offer, but that doesn’t mean the former should be a lowball offer! So… how do you tell whether it’s a “fair” offer?
Conversely, how do companies ensure that they’re making fair but competitive offers to candidates across the different roles and departments? And over time, as they mature and bring their cash compensation closer to “market”?
There are many gaps in the way that compensation is currently managed and it results in some pretty inequitable outcomes:
“Compensation studies” have significant shortcomings
Compensation studies cost (tens of) thousands of dollars and say things like: “For a company that has raised $3-10M, a Staff Engineer should be paid $140-180K and 0.5-2.0% in equity”. That is such a wide range that it’s barely useful!
Which brings us to takeaway #1: Benchmarks in the absence of a framework can be harmful, not helpful.
Compensation "Ranges" introduce bias, and aren't the only way to go
This is a controversial topic, so we'll present both sides of the argument.
1. The argument against Bands: When “Bands” are ranges, they allow for bias.
- The same hiring manager might make two offers, and one candidate negotiates while the other doesn’t. Now what? Candidate #2 ends up getting $10K extra / year just for asking, even if both candidates are “in band”
- Two different hiring managers might make different decisions about the same person. As we all know, this is all too common with subconscious gender and racial bias; but even with good intent, they simply might have different backgrounds (e.g., manager 1 is coming from BigCo while manager 2 from Scrappy Startup), or perhaps manager 1 has more budget. Both offers might be “in band” but it doesn’t mean it’s the same or even equitable.
This is why a “band” promotes inequity. If you want to pay more for performance, this should be captured through promotions or bonuses. If you think the offer for a candidate is “light” then it shouldn’t be “fixed” by flexing on the comp for the level… it’s probably a sign that the level was set wrong to begin with, or the candidate is being misleveled for their experience or capability.
2. The argument for Bands: a range of compensation allows for smoother transition between levels.
- For example, if the Total Comp / Total Rewards between levels is $20,000 apart, that's a big raise to offer in one go; and then, the employee might have to wait 2-3 years before their next raise. Many companies reasonably consider the tenure of an employee in deciding if they should be placed at the bottom, middle, or top of a compensation range.
- For example, a freshly promoted L5 might get a $10K raise, then a $3K raise next year, then a $5K raise next year to end up at the top end of the band – at which point they'd be up for their next promotion. It's a smoother process with a feeling of continuous improvement and upward mobility.
Takeaway #2: It's worth deeply considering whether you'd like your compensation ladder to be set with "Levels" (one single number for total compensation, per title), vs. "Bands" (a range of acceptable total compensation, per title).
Companies End up Creating Dozens of Levels Across Functions
Levels vary across functions; a Director of Product may not be the same compensation as a Director of Support, for example, even if they are both “Directors”. This is typically due to the career ladder for the functions, the nature of the role, and simply how competitive the talent market is for a role.
But there is no consistent way to normalize across different functions, so companies end up with dozens of compensation levels (10+ each in: engineering, product/design, operations, support, sales, marketing, finance...), adding an incredible amount of overhead for HR to manage. So redoing bands ends up being a massive undertaking, and companies do it very infrequently (12-24 months, and fall behind on market norms in the meantime).
Instead, creating a single set of levels across the company -- as few as 10-12 is usually sufficient -- and map different titles to each level. For example, a Director of Product may be L8, but a Director of Support might map to L7. This may seem like a counterintuitive gap, but it ends up being closer to market salaries anyway, and will significantly reduce your overhead in managing your levels.
Takeaway #3: Create a single set of levels that normalizes across departments, and map different titles to it.
Offers across Company Stages leads to Confusing Comparisons
There is no existing way to translate across the different stages in which a company can be. It makes intuitive sense that a “Sr. Engineering Manager” should make a different base salary and different “total rewards” comp (base + equity + bonus) at a Series A (perhaps $300K) vs. public company ($700K+). But how do you account for that? And how do you gradually increase it over time? Right now, companies dramatically update their framework once every couple years, and it’s time consuming and very laggy... and underpaid employees leave in the meantime.
Takeaway #4: Take a continuous improvement approach to compensation, and update your compensation framework every 6-12 months; with fewer bands (see #3 above) it’ll be possible for you to update your bands much more frequently.
I think we can all agree that compensation is a tricky subject to nail down. Everyone wants to feel that they’re being offered (or are offering) a fair compensation package. For startups that means a lot of variety in the way that compensation is structured making it harder to compare offers. With the advent of remote work, the location of the employee also adds another facet to consider when making an offer.
This brings us to takeaway #5: “fair” vs “unfair” is a better way to think about compensation than “right” vs. “wrong”. A company should set a framework that feels fair, and apply that in a uniform manner, without exceptions.
Okay, so if you think all of the above makes sense, you’re probably wondering how to implement this at your company (and you’re probably wondering how far off your company is already from a rational, well-designed framework!).
That’s where we can bring the magic. We’ve developed a proprietary framework we’ve backtested against thousands of offer letters spanning dozens of companies ranging from early stage, to pre-IPO, to public companies.
Want to hear more? Shoot us an email to hear about our early access: [email protected].