Investors use alternative data to spot changes inside a company before those changes show up in its reported numbers. Each type measures one thing well and misses others. Enrich Layer provides one type: data on who works at a company, who joined, and who left.
What is alternative data?
Alternative data is any information used to judge a company that doesn't come from its filings, its earnings calls, its share price, or government economic releases. Examples include card transaction records, store foot traffic, app downloads, job postings, and the work histories of a company's employees.
Its value comes from timing. Alternative data can show a change inside a company before that change appears in reported numbers.
Types of alternative data and what each one measures
The main sources of alternative data each stand in for something an investor can't see directly:
- Card and receipt data stands in for consumer revenue. It only covers businesses that sell to consumers, and it reflects whoever is in the panel of cardholders the data was collected from.
- Location and foot-traffic data stands in for store visits. It covers physical locations only.
- Web and app activity stands in for product demand. It says little about businesses that sell to other businesses or sell offline.
- Reviews and social posts stand in for customer sentiment. A small number of very vocal people can skew it.
- Job postings stand in for planned hiring. A posting is a plan, and plenty of postings never become hires.
- Work histories stand in for actual hiring and departures. They're self-reported, so they arrive late.
How hedge funds and venture investors use alternative data
Hedge funds mostly invest in public companies, which already report their numbers every quarter. They use alternative data to estimate those numbers before the report comes out, or to check them from a second angle. Card data can give an early read on a retailer's sales. Headcount data can show whether a company is hiring at the pace its management described on the last earnings call, and how that pace compares with its competitors.
Venture and growth investors mostly invest in private companies, which don't publish their financials. For them, alternative data is often the only outside view of how a company is doing. They use it in three places: finding companies before those companies raise money, checking a founder's claims about team growth during due diligence, and keeping track of companies they've already invested in.
Why people data belongs on the list
Hiring is a spending decision a company makes months ahead, based on internal plans an investor can't see. A company adds engineers before it ships a product, salespeople before a sales event, and compliance staff before it enters a regulated market. When senior people leave, that often comes before an announcement about the reason.
When people change jobs, many of them update their own professional profiles, and those profiles are public. Put the work histories of a company's staff together and you get a dated record of decisions made inside the company, available well before those decisions show up in revenue.
Hiring velocity shows which teams in a company are growing
Hiring velocity is how fast a company's headcount changes over a period. It's often broken down by function: engineering, sales, operations, and so on.
"Is this company growing?" is a basic question, and total headcount answers it bluntly. A more useful question is "Which teams is this company adding people to, and is that faster or slower than last year?"
A company that adds ten salespeople and no engineers in a quarter is making a different bet from one that adds ten engineers and no salespeople, even though total headcount moves by the same amount.
Measuring hiring with headcount data
Headcount data can answer that question in two ways: how many people worked at a company on past dates, and who joined and who left.
Headcount at past dates
The Employee Count endpoint takes an at_date parameter that returns a company's headcount on a past date:
GET /api/v2/company/employees/count
?url=<company profile URL>
&at_date=2025-10-01
Call it for a series of dates, such as the first of each month for the past two years, and you have a headcount curve.
Each number is a count of the people whose work history says they worked at the company on that date. The endpoint rebuilds the past from the work histories in the data, rather than looking up a number someone recorded at the time.
As a result, people who never had a professional profile aren't counted, and older dates tend to be less complete than recent ones. The shape of the curve is more reliable than any single point on it.
The count only includes people whose work history links to that company's profile, so for some companies it's far below the real headcount, today as well as in the past. Compare a company with its own past rather than reading the number as a total. A count of zero usually means the company wasn't matched, so check the company URL before reading it as a company with no staff.
Who joined and who left
The Employee Listing endpoint returns the people who work at a company. Set employment_status to past to get the people who used to work there, or all to get both:
GET /api/v2/company/employees/
?url=<company profile URL>
&employment_status=past
&enrich_profiles=enrich
With enrich_profiles set to enrich, each person comes with a work history that has start and end dates. That lets you break headcount change down by job title and seniority, so you can see which teams grew, which shrank, and whether any senior people left in the same few months.
Job postings show intent, but not history
The Job Search endpoint returns a company's open job postings, each with a title, location, posting date, and link to the posting. Postings show what a company plans to hire for before anyone is hired, which makes them earlier than work histories.
They come with three limits:
- A posting isn't a hire, and some postings stay open for months or are put up to collect résumés.
- Job titles come back exactly as the company wrote them, so grouping them into functions like engineering or sales is your job.
- The endpoint returns postings that are open now, without a record of past ones.
To see how a company's hiring plans change over time, save what the endpoint returns each time you call it, with the date. The history starts on the day you start saving.
What people data can't tell you
It's self-reported. People update their profiles late, and some never do. Recent changes and very small companies are the most affected, which is a problem because early-stage companies are often the ones investors care about most.
It's uneven across jobs, industries, and countries. Engineers and salespeople keep their profiles more up to date than warehouse staff or field technicians. A manufacturing company's headcount will usually read lower than its real headcount. Compare a company with its own past, or with companies in the same industry and region, rather than across very different businesses.
It approximates headcount. It isn't a payroll figure, and the numbers shouldn't be quoted as if they were. Use them to rank, sort, and spot change.
It doesn't tell you the why. A drop in engineering headcount can mean layoffs, a reorganization, or a team moved to a contractor. People data shows that something changed. Finding out why takes other sources.
Frequently asked questions
Is alternative data the same as alternative investments? No. Alternative investments are assets outside stocks, bonds, and cash, such as private equity, hedge funds, real estate, and commodities.
Alternative data is information used to research an investment of any kind. The terms get mixed up because the same funds often use both. Questions about the drawbacks of alternative investments, such as higher fees and how hard they can be to sell quickly, are about the assets themselves and fall outside this piece.
What is alternative credit data? Lenders use alternative credit data to decide whether a person can repay a loan when that person has little or no credit history. Common examples are rent payments, utility and phone bills, and the money moving in and out of a bank account. It shares a name with the alternative data in this piece, but it's used to judge individual borrowers, while investors use alternative data to judge companies.
What is alternative data in AI? AI describes how the data gets analyzed. Funds use machine learning models to find patterns across large alternative datasets, such as millions of job postings or work histories, and some funds use alternative data to train those models. The limits described above still apply: a model trained on self-reported headcount has the same gaps as self-reported headcount.
