A referral program for data and analytics consultants who know where the history lives

Why analytics consultants spot strong candidates early

You already know which source systems feed a client's warehouse, how far back each one goes and which ones are trustworthy. That puts you in a rare position to tell an owner that the records underneath their reporting may have value to AI developers, and to introduce them to SourceX, which runs the licensing from rights review to delivery and payment.

Your daily work makes this concrete. Source-to-target mappings, connector setups, backfill decisions, lineage diagrams and data dictionaries tell you, for example, whether a client's helpdesk history runs back a decade while its CRM is only a few years old. Nobody else in the client's orbit has that map.

The demand side explains why the raw layer matters. Researchers at Epoch AI estimate the effective stock of public human-written text at roughly 300 trillion tokens and project that, if current trends continue, language models will fully use it sometime between 2026 and 2032; it is a forecast with wide uncertainty. Meanwhile AI developers are building agents that carry out tasks, and those agents learn from records of real work: multi-step workflows, decisions and outcomes that exist inside companies and barely at all on the public web.

Raw records versus analytics outputs: what is actually licensable

The short answer: the records your pipelines read from, not the things your pipelines produce.

The distinction is worth explaining to clients, because many assume that a polished BI estate is what has value. A messy, decade-old ticket history with resolutions often matters more than the cleanest star schema.

The source depth score

Score each source system you have connected for a client, one point per yes. This is a partner rule of thumb, not SourceX's qualification criteria.

• History: the system holds several years of records, not just data since the warehouse went live.
• Free text: records include descriptions, comments, emails, chat or documents, not only codes and amounts.
• Outcomes: fields show how things ended, such as resolved, escalated, won, lost, approved or rejected.
• Joins: records link to other systems, for example ticket to account to invoice to engineering issue.
• Ownership: the client generated the records about its own operations rather than processing them for its customers.

A client with three or more systems scoring four or five is worth a conversation, provided the company baseline also holds: 50+ full-time employees at peak (contractors excluded), several years of documented operations, rights to license the data and an owner, CEO, CFO or other authorized sponsor who can sign. The who qualifies page has the full criteria.

Which clients in an analytics book tend to fit

Fit tracks how work is recorded more than the industry label.

Clients exploring AI agents sometimes conflate giving an agent live access to their systems with licensing data. The comparison of MCP data access and data licensing separates the two, and the guide to data monetization consulting covers where licensing sits next to internal analytics and data products.

When to raise it in an analytics engagement

Retirements and migrations overlap with the work of data migration specialists, who often sit on the same projects.

How the introduction works while you stay out of the data path

Your role ends at the introduction. No extracts, no schema files, no samples, no row counts sent to anyone.

1. Mention the idea to the executive sponsor and get their agreement to talk.
2. Register, then either send your referral link or fill in the referral form using basic fit details: company name, rough peak headcount, years in operation.
3. SourceX assesses size, history, breadth of records and rights with the sponsor.
4. The company's own team lists its systems and record types in a data inventory; the data inventory builder helps them think it through.
5. Pricing comes as one all-in figure that already includes SourceX's fee, with no separate charges, alongside the license terms.
6. AI labs and data buyers review, then the company signs or walks away; nothing binds it before signature.
7. Before any records are prepared, the company's agreed redaction and de-identification requirements take effect. Nothing is released until the agreement is executed and the company signs off, after which it receives a one-time payment.

What to say to the sponsor

How rewards work for analytics consultants

Partners earn 25% of the eligible platform fees SourceX actually collects from the referred company's licensing deals, capped at $100,000 per referred company. The reward becomes payable only after the buyer pays and SourceX receives its fee; an introduction, meeting or signed agreement alone does not trigger payment, and no reward is guaranteed.

The reward is a share of SourceX's fee, so the client's proceeds are untouched. If you work inside a consultancy or as a subcontractor, check your agreement on outside compensation before registering; if you are independent, mention the referral relationship to the client as part of the introduction. You can be based anywhere, but the company you introduce must be in the US. Colleagues on the strategy side can use the management consultants page.

When not to bother

• The client never reached 50+ full-time employees at peak, however large its database.
• The warehouse holds aggregates only and the source systems have been purged.
• The data is mainly consumer personal information or patient records.
• The client processes data on behalf of its own customers.
• An earlier AI-training license already covers these records.
• Nobody at the company can authorize a license or run exports.

Next step

Pull up the source inventory for your three longest-running clients and score each system. Where the depth is there, register as a partner; the sponsor can then apply at sourcex.si/apply using the link you send.