Guide
Where to sell data
Where to sell data depends on the dataset: compare cloud marketplaces, listing sites, brokers, direct deals and asset sales on fees and fit.
Updated
The short answer
Where to sell data depends on what you hold and who would buy it. Recurring, well-structured feeds suit cloud marketplaces such as AWS Data Exchange or Snowflake Marketplace. Listing sites like Datarade help buyers discover you. One-off or unusual datasets, such as codebases or company archives, usually sell through direct licensing, a broker, or an asset sale.
Five places data gets sold
Data changes hands through five broad routes. They are not interchangeable. A marketplace is a shop window and a billing system; it does little to find buyers for something they didn't know to search for. A broker does the opposite: no shop window, but active matchmaking. Picking the wrong route usually shows up as silence, not rejection.
| Channel | Who can sell | What it costs the seller | Best for |
|---|---|---|---|
| AWS Data Exchange | Registered AWS Marketplace sellers in eligible jurisdictions, after qualification by the AWS Data Exchange team | 3% listing fee on public offers; private offers 3% under $1M contract value, 2% from $1M to under $10M, 1.5% at $10M or more; data grants billed hourly | Structured files, APIs or tables that buyers already run on AWS |
| Snowflake Marketplace | Full Snowflake accounts with an approved provider profile and accepted provider terms | Paid listings use usage-based or subscription pricing, paid out through Stripe; the pages we checked don't state Snowflake's share | Tables buyers want to query inside their own Snowflake account |
| Datarade | Data providers who apply and are accepted | Commission-only plan with no subscription and a 30% commission; paid tiers lower the commission (Bronze 20%, Silver 15%, Gold custom) for an annual fee | Being found by buyers who search a catalogue of data products |
| A broker | Any owner with rights to license | A commission or fee agreed in writing, deal by deal | Unusual or one-off data where buyers must be found and convinced |
| Direct licensing | Any owner with rights to license | Your own time, legal fees and sales effort | Owners who already know their buyers |
| Asset or wind-down sale | Companies closing, in bankruptcy, or selling assets | Set by the sale process, often with court or creditor oversight | Codebases and workspace archives of closed companies |
How to choose a channel
- 1
Describe the dataset in one line
Is it a refreshed feed, a one-time snapshot, or an archive? Feeds suit marketplaces with subscription billing. Snapshots and archives tend to sell through negotiation.
- 2
Name five likely buyers
If you can name them and they already buy similar data, go direct or through a broker. If you can't, a catalogue listing can help them find you, but only if they search for that kind of data.
- 3
Check where your buyers work
A buyer whose data team lives in Snowflake will find a Snowflake listing easier to trial and buy. A buyer on AWS can subscribe and pay through their existing AWS bill.
- 4
Check eligibility before you plan around a platform
AWS Data Exchange limits sellers to listed jurisdictions and requires qualification; Snowflake requires a full account and an approved profile for paid listings. Read the requirements first.
- 5
Compare the real cost
A low percentage fee still needs you to build the product page, support subscribers and keep the data updated. A broker's higher cut can still be cheaper than months of your own sales time. Write both down.
- 6
Keep the licence yours
Whichever channel you use, make sure the licence terms (permitted use, resale limits, exclusivity) are the ones you intend. Marketplaces offer standard terms; read them and add your own where allowed.
Which channel for which data
| What you have | Usually start with | Why |
|---|---|---|
| A refreshed feed of market or company data | A cloud marketplace, plus a catalogue listing | Buyers expect subscription billing and delivery into their tools |
| A large one-off archive (documents, emails, chats) | A broker or direct approach to AI labs | Few buyers, high scrutiny of rights and personal data |
| A codebase from a closed or pivoted product | A broker, an AI lab directly, or an asset sale | Value depends on history and rights, judged case by case |
| Customer records with personal data | A privacy lawyer before any channel | Sale may trigger state data broker laws and consent rules |
| A niche dataset with unknown demand | A catalogue listing and a few direct conversations | Cheap way to test whether anyone asks |
Where marketplaces are the better choice
Marketplaces win when the data is a product rather than a deal: a feed with a schema that rarely changes, many potential buyers and a price that can be published. They handle billing, entitlements and delivery, and buyers can sign up without a long negotiation. If that describes your data, a broker adds cost for little benefit.
Brokers and direct deals win when each sale is different. Recent reporting shows AI labs and data companies buying code and workspace archives from shut-down startups, often through wind-down services, and a bankruptcy auction in which Google agreed to pay $10 million for Spirit Airlines' internal business data. Those are negotiated sales with legal review, not marketplace checkouts.
What this guide can't tell you
Sources
| Source | What it supports | Link | Checked |
|---|---|---|---|
| AWS Marketplace listing fees | AWS Data Exchange 3% public fee; private offer tiers | https://docs.aws.amazon.com/marketplace/latest/userguide/listing-fees.html | 2026-11-07 |
| AWS Data Exchange pricing | Hourly charge for data grants; fees on revenue AWS collects | https://aws.amazon.com/data-exchange/pricing/ | 2026-11-07 |
| AWS Data Exchange provider guide | Seller eligibility, jurisdictions, qualification | https://docs.aws.amazon.com/data-exchange/latest/userguide/provider-getting-started.html | 2026-11-07 |
| Snowflake: becoming a provider | Full account, approved profile, Stripe payouts | https://docs.snowflake.com/collaboration/provider-becoming | 2026-11-07 |
| Snowflake: paid listing pricing models | Usage-based and subscription pricing | https://docs.snowflake.com/en/collaboration/provider-listings-pricing-model | 2026-11-07 |
| Datarade provider plans | Commission-only plan at 30%; tier commissions | https://providers.datarade.ai/apply | 2026-11-07 |
| MIT Technology Review, Oct 2026 | AI labs buying data from shut-down startups | https://www.technologyreview.com/2026/10/13/1145755/ai-labs-buying-data-from-dead-startups-to-train-models/ | 2026-11-07 |
| iTnews (Reuters byline), Aug 2026 | Google's $10 million Spirit Airlines data purchase | https://www.itnews.com.au/news/google-to-buy-spirit-airlines-business-data-for-us10-million-628253 | 2026-11-07 |
Questions people ask
Is there one best data marketplace?
No. The right one is where your buyers already work. A buyer on AWS will find AWS Data Exchange easiest; a Snowflake shop will prefer a Snowflake listing. For data that needs explaining, no marketplace replaces a direct conversation.
Can individuals sell data on AWS Data Exchange?
AWS says providers must be permanent residents or citizens of an eligible jurisdiction, or a business organised there, be registered AWS Marketplace sellers and be qualified by the AWS Data Exchange team. In practice it is built for businesses with data products and customer support.
Does Datarade charge buyers?
Datarade's published provider plans charge sellers, through a commission and optional annual subscription. Check Datarade's own pages for current buyer terms before you rely on them.
Can I list the same data in several places?
Usually yes, unless you have granted someone exclusivity. Keep licence terms consistent across channels so one buyer's rights don't conflict with another's.
Where do closed companies sell their data?
Often through wind-down services, brokers or, in bankruptcy, a court-supervised asset sale. Rights and personal data get extra scrutiny because the people who created the data are no longer around to consent.
Related
- How to sell data your company holdsHow to sell data your business already holds, from listing what you have to signing the licence and handing it over safely.
- How to sell data to AI companiesSell data to AI companies only after you can prove where it came from, that you may license it, and why it isn't online already.
Become a data broker
The DataBroker course and tools: find companies with data to sell, price it, license it and close the deal.