Thinknum alternatives for web-sourced alternative data

Funds follow companies between reporting dates with alternative data collected from the web: job listings, store counts, prices, technology, announcements. Thinknum Alternative Data is one supplier of it. This guide is for the analyst or data lead who uses Thinknum, or is weighing it, and wants to know what else covers the same ground. It sets out what Thinknum's own site says it offers, what it is built for, four other sources, and a test that shows whether a second feed agrees with the one you hold. Statements about other companies were read on their own pages on 4 October 2026, and each page is linked.

Start with what each is built for. On that day Thinknum's Job Listings page stated a coverage period of more than 11 years, and its site names datasets on store locations, product pricing and car inventory, worked through an application of charts, dashboards and alerts. Fokals is built for research that needs hiring, technographic, announcement and intent data on one company index, with every listed company mapped to its security identifiers and daily and weekly datasets written once after the period closes and never revised.

What Thinknum offers, as its site describes it

Thinknum's About page says that companies create data trails as they move their operations onto the internet, and that Thinknum indexes those trails in one platform for investors. Its home page shows example datasets: store locations, job listings, product pricing, property pricing, car inventory and Facebook app active users. The dataset catalogue adds its Apple Store Ratings dataset and several views of products by vendor. A page for corporate teams states 450,000+ US and international companies, public and private, and the Job Listings page stated more than 4,600 public and more than 6,800 private companies on the day it was read.

The product is an application as well as a feed. The home page describes tools to build and share queries, ready-made maps and charts, dashboards with widgets, and alerts sent by email when a metric crosses a threshold. The API documentation lists a Query API, a Historical API, a Company API and an Upload API. According to those pages a query names a dataset, with tickers to limit it to certain companies; results can be saved as CSV, XLS or XLSX; and the Historical API returns a dataset's full history or a daily update feed. The documentation also carries guides for a Google Sheets add-on, Power BI Desktop, Tableau Desktop and a Python package.

No price appears on the pages linked here. The site asks you to request a demo, and the documentation says API credentials are issued by a Thinknum account manager.

What each is built for

• A long record of job listings. Thinknum's Job Listings page states a coverage period of more than 11 years.
• Datasets on stores, products and cars. Store locations, product-level pricing, property pricing, car inventory and Apple Store ratings are on Thinknum's site.
• An application. Thinknum documents charts, maps, dashboards and alerts, a Google Sheets add-on and guides for loading its data into BI tools.
• App and social engagement. Thinknum's home page lists Facebook app active users.
• One company index with security identifiers. Fokals resolves every company to one stable company ID. A listed company carries its ticker, MIC, ISIN, LEI and share-class FIGI, and a brand carries the identifiers of its listed parent, so the join to a security master needs no name matching.
• A point-in-time record. Every observation is dated. Daily and weekly datasets are written once, after the period closes, and are never revised, so a query as of any day in the record returns what was known that day.
• Hiring with labels. Every role a company publishes is labelled by job function, seniority and ten role flags, and Hiring Activity gives daily open, new and closed postings for each company.
• Company actions beside hiring. Technology Changes, Company News, Company Funding and Intent Scores sit on the same company ID, and the weekly Market Series give the sector context on same-store cohorts.
• Delivery into your own research stack. A REST API of 25 endpoints and bulk exports of any dataset for any period, with a manifest naming the period, label versions and licence.

Why a team looks for an alternative

These are reasons of fit, and they say nothing against the product.

• Scope. The research question is about which functions a company hires for, which technology it runs or what it has announced, and you want those on one company key.
• Delivery. Your research runs in a database of your own, beside prices and fundamentals, and you want files and an API that your jobs fetch on your timetable.
• Licence fit. Internal research is one use and a product for clients is another, so tell each vendor, Thinknum included, which you intend.
• A different kind of data. A compliance review asks where each row came from, and a back-test asks when it was known. You want the source and the observation time on the record, and documents you can hand to a reviewer, such as the sourcing statement.

Thinknum and Fokals side by side

What is said of Thinknum comes from the pages linked above, and what is said of Fokals from its data dictionary and methodology.

Four other sources to weigh

Each gets one sentence, from its own site as read on 4 October 2026.

• LinkUp. LinkUp says its job market data is taken straight from employers' own websites and refreshed daily, with a record that starts in 2007, and it delivers its RAW dataset through cloud platforms and FTP. The guide to LinkUp alternatives goes further.
• Revelio Labs. Revelio Labs describes its Revelio Terminal as unifying professional profiles, job postings, employee sentiment and layoff data and tracking headcount, hiring, attrition and skills for any public or private company, with APIs for access to its workforce data.
• Similarweb. Similarweb offers Stock Intelligence, which it describes as web, app and user engagement trends on public and private companies, delivered through dashboards, an API, S3 or Snowflake.
• Coresignal. Coresignal offers company, employee and job posting data that it says is collected only from publicly available sources, through APIs and as datasets in formats that include JSONL, CSV and Parquet.

Each is built around something of its own: LinkUp a record that starts in 2007, Revelio Labs the workforce measures built on profiles, Similarweb the traffic and app engagement, Coresignal the employee records. Fokals is built around the company: hiring, technographic, intent and announcement data on one company ID, with security identifiers on every listed company.

Testing a second feed on the companies you track

Job postings are where the two catalogues overlap most plainly, so that is the place to test. The steps use Hiring Activity, in the Fokals hiring dataset, and a series you already hold, keyed by ISIN after you have mapped its tickers in your security master. The query below calls it the incumbent series.

1. Check coverage first. Every Fokals row carries the ISIN, FIGI, ticker and MIC of a listed company or of its listed parent, so the join needs no name matching. Count the names in your list that have at least one row in Hiring Activity. Brands under a parent, delisted names and check digits are treated in mapping alternative data to a security master.
2. Roll up before you compare. Brands and subsidiaries are filed under their listed parent's identifiers, which means one ISIN can stand over several company IDs. Sum open postings by ISIN and day before you set them beside a series keyed by ticker.
3. Compare changes, not levels. Two vendors rarely read the same job boards or close a posting by the same rule, so their counts will differ. What should agree is the direction of the week.

-- incumbent_jobs(isin, day, open_jobs): the series you already hold
with fokals as (
  select isin, day, sum(open_postings) as open_postings
  from company_hiring_daily
  where isin is not null
  group by isin, day
)
select
  f.isin,
  count(*) as days,
  corr(f.open_postings - f0.open_postings,
       i.open_jobs - i0.open_jobs) as weekly_change_corr
from fokals f
join fokals f0         on f0.isin = f.isin and f0.day = f.day - 7
join incumbent_jobs i  on i.isin  = f.isin and i.day  = f.day
join incumbent_jobs i0 on i0.isin = f.isin and i0.day = f.day - 7
group by f.isin;

Date arithmetic differs by warehouse. Read the correlation name by name as a check of direction, and judge it over a full quarter of overlap, so that one reporting cycle sits inside the window.

Three properties of the record explain most differences in level.

• A first observation is a baseline. A posting already open when a company enters the dataset is counted as open and never as new. New postings are roles observed to open, so the series of new postings counts the roles that opened in each period.
• A closure is confirmed before it is written. Each vendor confirms a closure by its own rule, so closed counts differ more than open counts. The Fokals rule is set out in the methodology; ask your other vendor for its rule before you compare closed counts.
• Roles come from the company itself. Fokals records the roles a company publishes on its own careers pages. A series that also reads job boards and staffing sites counts listings from those places too, so levels differ while direction agrees.

Revisions are a separate question, and one to put to every vendor on your list. Fokals daily and weekly datasets are written once, after the period closes, and are never revised. That makes the series point-in-time: a row holds what was known when it was written. Once the feed is loaded, the guide to using hiring data in equity research shows what to build on it.

Frequently asked questions

The queries and code on this page are examples to adapt. Test them in your own environment before you rely on them.

What this page says about the products it names was checked against their public documentation on 4 October 2026. Product and company names are trademarks of their owners. Fokals is not affiliated with them or endorsed by them.