How to Choose the Best Way to Access TheirStack Data
Guidance to help you choose the most suitable way to access TheirStack data (App, API, Webhooks, or Datasets) for your use case, scale, and freshness needs—without prescribing a single path.
TL;DR
- If you're not a developer, use the App.
- If your code would run the same search on a schedule (every hour, every day, every week), use Webhooks. See API or Webhooks?.
- If you need the data once or on demand (a one-off pull, or a user who clicks a button, opens a company page or types a search), use the API.
- From roughly 2M to 5M records per month, Datasets may become more cost-effective, so contact us to compare the options.
- If you want your AI assistant to query TheirStack data directly, use the MCP server.
Basics
- Same data for all methods: Every delivery method accesses the same underlying dataset and quality.
- Filtering: All filters are available in the App, API, and Webhooks. Datasets are exported as flat files (CSV/Parquet) without server-side filtering.
Decide where the data and filtering should live
- Datasets: We deliver complete historical and ongoing files through S3. You ingest and store them, then build the filtering, enrichment, and query experience your users need. This provides the most control and is a strong fit for broad coverage, analytics, or model training, but requires the most data engineering.
- Webhooks: You define filters in TheirStack and we push matching records into your database. You store the records and build any additional filtering or customer-facing query experience. Webhooks consume API credits for each record delivered.
- API: Your product queries TheirStack on demand instead of ingesting the full database. This requires less data infrastructure, but every record returned consumes API credits, including records requested again.
API or Webhooks?
Both give you the same records, the same filters and the same price per record. What tells them apart is whether the same request repeats on a schedule.
- The same search, on a schedule → Webhooks. If you're about to write a cron job that calls the API every few hours or days with the same filters to pick up what's new, create a webhook with those filters instead. We send each new matching record to your endpoint as soon as we find it.
- A one-off or on-demand request → API. You pull a list of jobs once for an analysis or an initial load, a user opens a company profile in your product, or someone enriches a CRM record or runs a search from your UI. The request either runs once or depends on a moment nobody can predict, so there is no recurring search for a webhook to run.
Why a scheduled job is better as a webhook:
- You get the data sooner. A webhook fires when we discover the job. A cron job only sees it on its next run.
- You don't have to handle duplicates. Polling the same search returns records you already have, unless you track
discovered_ator pass the IDs you've stored withjob_id_not. A webhook sends each matching record once. - You don't pay twice for the same record. Every record the API returns costs credits, including the ones you already fetched in the previous run. A webhook only charges for records it delivers.
- Less code to maintain. You don't need a scheduler, a pagination loop or logic to resume after a failed run, and we retry failed deliveries every hour for 48 hours.
A quick test: if the request would look the same every time it runs, and only the results change, it should be a webhook. If you still need to poll, follow How to fetch jobs periodically to avoid duplicates and extra charges.
Estimate your monthly volume first
Most of the decisions below — Webhooks vs Datasets, which plan, what it costs — depend on a single number: how many jobs per month your use case actually needs. The fastest way to find out is to let the data tell you, before you talk to us or pick a plan:
- Open Job Search.
- Add the exact filters you'll use in production — job titles, countries / markets, and any industry, technology, or seniority filter that defines your niche. The more precise your filters, the more accurate (and cheaper) your estimate.
- Set Date posted to last 30 days.
- Read the total results count at the top of the results. That number is a close approximation of the jobs you'd ingest per month with those filters.
Repeat with a few filter variations to see how volume changes — for example, adding a second country or a broader title list. This gives you a realistic min/max range to plan around.
Use that number to choose your path:
- Under ~2M records/month → API or Webhooks. These are usually the simplest options for job boards, niche feeds, lead generation, and on-demand product features. You pay only for the filtered slice you need.
- From ~2M to 5M records/month → compare API/Webhooks with Datasets. API and Webhooks remain available through self-serve plans, while Datasets may start to make economic and operational sense depending on how broad your coverage is and where you want filtering to happen. Contact us with your monthly volume, markets, filters, and historical-data needs, and we will help you compare both routes.
- Above ~5M records/month → Datasets usually become more cost-effective. The exact crossover still depends on whether you need the complete database or only a filtered slice.
Delivery methods
App (UI)
Use the App when you want to explore, iterate, and export data without writing code.
- What it’s best for
- Target companies with active job openings
- Find companies by tech stack or discover a company’s tech stack
- Source candidates with niche technology experience
- Ad-hoc searches, validating data quality, and exporting to CSV/Excel
- Pricing
- Uses company credits
- Learn more
Datasets (bulk + historical)
Use Datasets when you need large-scale access for warehouses, analytics, or ML.
- What it’s best for
- Expand your job board with fresh listings (backfilling)
- Power sales intelligence platforms at scale
- Convert competitors’ paying advertisers into your customers
- Loading into a data warehouse or data lake (S3, BigQuery, Snowflake, etc.)
- Historical analysis, model training, and high-volume consumption
- Pricing
- Flat subscription; may become more cost-effective between roughly 2M and 5M records/month, depending on scope and architecture
- Learn more
- Overview: Datasets
- Jobs schema and coverage: Job data
- Access steps: Obtaining access to your datasets
Webhooks (real-time)
Use Webhooks when you need to ingest fresh data continuously without polling.
- What it’s best for
- Monitor past customers that are hiring again
- Spot your competitors’ next moves
- Expand your job board with fresh listings (backfilling)
- Match job seekers with relevant opportunities (academies, bootcamps, career services)
- Triggering workflows in Zapier, Make, n8n, or your own services via saved searches
- Pricing
- Consumes API credits; pay per record returned
- Learn more
API (on-demand)
Use the API for one-off pulls and for requests that need data right away: enrichment, lookups, or searches triggered from your product.
- What it’s best for
- Enrich companies with technographic data on demand (CRM, product records)
- Power sales intelligence platforms with server-to-server lookups
- Identify companies with problems your software solves
- Synchronous product features that fetch data on demand
- Pricing
- Consumes API credits; pay per record returned
- When not to use
- If you'd call it on a schedule with the same filters to pick up new records, use Webhooks instead. You get the records as soon as we find them and never pay twice for the same one
- Learn more
MCP (AI assistants)
Use the MCP server when you want your AI assistant to search TheirStack data conversationally.
- What it's best for
- Asking natural-language questions about jobs, companies, and tech stacks from Claude, Cursor, ChatGPT, or any MCP client
- Quick ad-hoc research without writing code or learning the API
- Create niche job newsletters with AI-assisted curation
- Pricing
- Consumes API credits at the same rate as the REST API
- Learn more
Examples
- Sales or RevOps team wants to test a new ICP and export a CSV for outreach → App (UI)
- Product team needs new US tech jobs streamed into an internal DB every hour → Webhooks
- Developer plans a nightly cron job that searches for new Data Engineer jobs in Germany → Webhooks
- SaaS product shows a company's open jobs when a user opens that company's profile → API
- Data team wants multi-year job postings for trend analysis and modeling → Datasets
How is this guide?
Last updated on
How to Check if a Technology or Keyword Is Available
Learn how to find out whether TheirStack tracks a specific technology or buying intent keyword, download the full catalogs, use job description search as a workaround, and request new keywords through the Support Center.
How to create a niche job newsletter with TheirStack MCP
Learn how to create a weekly curated job newsletter using Claude.ai and TheirStack MCP. Automate job newsletter creation with a reusable template — no code needed.