Compensation Bands

USV

LOU CONSULTING

THE PEOPLE DESIGN HOUSE

Focus Demographics

1-150 Employee Count

Series Seed - B

Novice

Creating Compensation Bands

KEY STEPS

Where are we pulling the data from?

DETERMINE DATA SOURCE

OPEN COMP

Determine Job Architecture

HOW MANY LEVELS DO WE HAVE? DO WE, IF SO, HOW, DO MANAGERS AND IC’S ALIGN?

Determine Job Architecture

WHICH FUNCTIONS/ROLES DO WE PROVIDE BENCHMARK/LEVELS FOR?

DEFINE YOUR

“TALENT MARKET”

CHOOSING MARKET PERCENTILES

Determining Market Competitiveness

Once you’ve established the talent market data you’ll need to choose a target percentile to build your benchmarks around (i.e. 50th %ile, 75th %ile, or top of market). The higher the target percentiles the higher the cost but the easier it is to attract talent

Some companies will use different percentiles for different functions (i.e. 75th %ile for tech, 60th %ile for non-tech)

​

Your talent market will be key to determine who and how you should be paying. �This is where you will be competing for candidates.

In order to target your companies talent market you’ll need to filter market data to best represent this group;

1. Industry, Valuation, Employee size, Location ect.
2. Peer Groups - competitor companies

Creating Compensation Midpoints

After creating benchmarks evaluate against current employee base to evaluate outliers

1

2

3

4

5

Align market data with job architecture for benchmark creation

Set natural connections points between market data and job architecture

Create initial benchmarks making sure there is the proper “distance” between levels (smoothing)

Evaluate benchmarks with other groups to ensure differences between ladders fit business strategies.

Focus Demographics

1-150 Employee Count

Example Spreadsheet

Novice

COMP GRADE

1 2 3 4 5…

SALARY RANGE

JOB FAMILY: SALARY RANGE BY COMP GRADE

1

2

3

First compensation benchmarks

Determine the width of your range (ex. 15%)� A) Compensation Philosophy� B) Level or Function

Apply to benchmarks

Creating Compensation Ranges

The Covid-19 pandemic has prompted companies to evaluate their work models. Companies have adopted fully remote or hybrid models depending on their lines of business. Successful location based pay strategies factor in the following considerations:

​

• Where does the company want to be strategically located?
• What is the business need for talent and where are they located?
• Does the workforce need to be centralized or can it be distributed geographically?

​

EXAMPLES OF LOCATION BASED PAY STRATEGIES BELOW:

NATIONAL RATE

TIERED MODEL

DIFFICULTY

SPECIFICITY

Locations Impacts on Comp Benchmarks

GEO SPECIFIC

• Most companies still use a tiered approach anchoring to 3 or 4 geos

​

• Some companies are beginning to use one geo �for the US

​

• How many tiers and the geo diffs are company specific based on their location strategy

​

• There is a lot of buzz around having 1 geo but current market data does not support this idea yet

What are other companies doing?

Questions?

Appendix

Creating Midpoints (exercise)

​

​

Creating Midpoints (exercise)

​

​

Application of Benchmarks

GREAT!

You’ve created you’re first set of benchmarks.

Now, here is what you can do with them:

Create Compa-ratios (CR) = Salary / (Midpoint or Benchmark) which help normalize pay

Evaluate internal pay equity based on Compa-ratio

Have the basis for more advanced compensation planning. For example:

• Merit Matrix
• Pay Equity Analysis

EX. MERIT MATRIX MODEL

1

2

3

Creating Benchmarks (Tips)

HOW CAN I CREATE BENCHMARKS WHEN MARKET DATA IS SUFFICIENT

Determine Target Market Percentiles

What % of market rate do we target to build our benchmarks around?

The higher the target percentiles the higher the cost but the easier it is to get talent

Least Expensive

Least Competitive

Most Expensive

Most Competitive

25%

50%

75%

90%

• Type 1: Data is missing for a level but not all levels in a job family
• Use the typical differences between levels to infer level progression. SWE 1 market data is 100k SWE III is 135k but SWE 2 is missing. What could be a placeholder for SWE 2?
• Type 2: Data for a whole job family is missing
• Find similar job family and extrapolate based on experience or company need.
• Type 3: Market data does not capture our specific problem properly
• Strategic business decisions should drive compensation strategy unless it’s to the detriment of the compensation philosophy. For example if the company can’t hire ML engineers and it’s the reason the company won’t hit their targets then you might have to pay more than the market data suggests to ML engineers.
• Create a structure to systematically track how you’re intentionally deviating from market data. Ex. (ML engineers is x% higher than SWEs or we price ML engineers y% higher than their market data)

​

Creating Benchmarks (Tips)

HOW CAN I CREATE BENCHMARKS WHEN MARKET DATA IS SUFFICIENT

Create Compensation Ranges

How do I calculate the low and high end of the band?