Referral reporting usually leaks by accident. A partnerships director builds a tidy slide for the quarterly business review, a sales leader forwards it, and a customer's licensing decision is now in a deck it never agreed to.
Why does a referral report leak deals?
Four leaks account for nearly every case, and the first is arithmetic.
| Leak | How it happens | Example (Illustrative) |
|---|---|---|
| Reward arithmetic | The reward is a set formula on what SourceX pays the customer, so one figure reveals the other | A reward amount beside a customer's name, worked through below |
| Status | A status of Paid means the customer signed, delivered and was paid | "Customer X: Paid" announces a completed license |
| Small counts | One company in a segment, region or month is that company | "One healthcare customer licensed data this quarter" names it for anyone who knows your healthcare book |
| Timing | A report sent the week after a single deal closes dates that deal | A new line appearing in the October report shows when it happened |
Illustrative: the default reward on a company's first qualifying transaction is the greater of $50,000 plus 5% and 15% of Actual Supplier Payout, with no cap, so a USD 55,000 reward beside a customer's name tells every reader that SourceX paid that customer USD 100,000, and a USD 70,000 reward says USD 400,000. There is no cap to muddy the arithmetic: the reward rises with every dollar of payout, so any reward figure can be worked back to the payout, and a repeat transaction within 12 months pays a flat 15%, which is simpler still to invert.
What can you see, and whose information is it?
Less than you might think, and most of it is about the customer. The program gives partners reward information, not deal files: the terms say partners never receive access to other partners' records, applicant documents or confidential buyer information. What you hold falls into three groups.
| Item | Where it comes from | Whose information it is | Default handling |
|---|---|---|---|
| Invitations, opt-ins, opt-outs | Your own outreach records | Yours, though each name is the customer's | Report as counts |
| Introductions submitted | Your partner records | Yours | Report as counts |
| Reward status for each referral | The program's statuses: Estimated, Eligible, Approved, Processing, Paid | The customer's progress | Aggregate only, minimum five |
| Reward amounts | The program | Reveals the customer's payout | Banded totals across five or more paid companies |
| Scope, buyer, terms | The customer and SourceX | The customer's, and partners do not receive buyer information | Never |
| Customer feedback on the program | Your own survey | The customer's words | Themes without names; quotes only with permission |
Two more rules come straight from the program terms. Before SourceX pays the customer, any reward shown is an estimate based on proposed compensation, and estimates are projections, not promises; and if a supplier payout is later reduced or reversed, the reward is reduced in proportion. A report that books estimates as earnings will be wrong in both directions.
The permission-based partner report
Four tiers, a minimum count and a lag. Every line in the report belongs to exactly one tier, and the tier decides who may see it.
| Tier | What it holds | Audience | Rule |
|---|---|---|---|
| A. Program activity | Invitations, opt-ins, opt-outs, introductions, median days from opt-in to introduction | Anyone inside your company; partner newsletters | Always reportable as counts |
| B. Program outcomes | Referred companies by reward status, total rewards paid | Executives and board | Only when five or more companies sit behind each figure; totals in bands; one quarter's lag |
| C. Named outcomes | A customer's name, that it licensed, a quote | The audience the customer approved | Only with the customer CEO's written permission, for the stated use and period |
| D. Never reported | Any one customer's payout, reward, dataset scope or buyer; record contents | Nobody | Not collected for reporting, so it cannot leak |
The minimum count of five is ours, borrowed in spirit from public-health statistics. The CMS data use agreement terms bar displaying any cell of 10 or fewer, and they also bar percentages or other formulas that would reveal such a cell. A referral cohort is small, so 10 would suppress almost every outcome line; five keeps a few lines alive while making it hard to single anyone out. Keep the second half of the CMS rule intact: no percentage, ratio or subtotal that lets a reader reconstruct a hidden figure, and if one segment is suppressed, suppress enough of its neighbors that subtraction does not reveal it.
A quarterly report built this way looks like the one below.
| Report line (Illustrative) | Q3 value | Tier and rule applied |
|---|---|---|
| Invitations sent | 46 | A: own activity |
| Customers who opted in | 14 | A: own activity |
| Introductions submitted after CEO confirmation | 9 | A: own activity |
| Referred companies in Estimated or Eligible status | Fewer than 5 | B: suppressed below the minimum count |
| Referred companies in Paid status, to date | Fewer than 5 | B: suppressed below the minimum count |
| Total rewards paid, to date | Not shown | B: fewer than five paid companies |
| Named outcome: a 300-person freight broker | Shown with an approved quote | C: written permission from its CEO, internal use, 12 months |
Illustrative. Invented figures that show the format only. Once five or more companies reach Paid, the total appears as a band, such as "USD 250,000 to USD 500,000", never as an exact sum. SourceX's rewards page explains the statuses, and the board reporting package template shows where a program summary fits in a wider board pack.
Who is the report for, and who can approve a name?
The customer's benefit from private reporting is simple: its licensing decision stays its own, and it never learns of its deal from your sales deck. Your benefit is that customers keep opting in. Two sponsors make it work:
- Inside your company, one executive owns the report and signs off each edition before it circulates, the same person who approved the referral pilot.
- Inside each customer, the CEO, who authorized the licensing discussion in the first place, is the right person to approve any named mention. A champion's enthusiasm is not permission.
Written permission should name five things: what may be said, to which audience, in which format, for how long, and how to withdraw it. Keep the approved wording with the permission, and if the customer withdraws, remove the line from the next edition and from shared folders. Being open about the reward itself, before any of this, is covered in how to discuss referral rewards transparently.
Keep product signals out of the referral report
Your product can often see things the referral program never told you: an unusual export, a new integration, a burst of admin activity. None of it belongs in a referral report. Using product signals to infer which customers licensed data would turn service data into deal intelligence the customer never agreed to share, and it would be guesswork presented as fact. Customer content generally belongs to the customer, as whether a SaaS company can license customer data explains, and the same goes for clues about what the customer did with it.
Build visibility into your own systems deal by deal. Microsoft's Partner Center, for example, lets partners create private deals that are not shared with Microsoft at all, alongside partner-led and co-sell deals. Give referral records the same choice: a referral flag visible to the partnerships team only, no reward fields in the CRM views sales can see, and no link from support tickets to referral records. If a customer asks you to move its records as part of a deal, the answer and the script are in when a customer asks the vendor to handle the data transfer.
Measure trust while you pilot introductions
The report itself can tell you whether customers trust the program. Add three measures and define them before the first edition:
| Measure | Definition | Denominator |
|---|---|---|
| Permission rate | Customers who agreed to a named mention | Customers asked |
| Withdrawal count | Permissions withdrawn during the period | Permissions in force |
| Report complaints | Customers or account owners who raised a concern about anything reported | Editions circulated |
SourceX has not published benchmarks for any of these, and this page gives none; the first two editions set your baseline. If permission rates are low, ask fewer customers and ask later, after they have seen how anonymous editions look. A small opt-in cohort, as described in the SaaS partner pilot with an opt-in customer cohort, keeps the first editions honest. Other partner playbooks are in the guides hub.
Next step
Refer a company. No cap on rewards. Default reward on a company's first qualifying transaction: the greater of $50,000 plus 5% and 15% of Actual Supplier Payout, with no cap; further qualifying transactions first agreed within 12 months of the first Closing pay 15%. Eligibility and payment conditions apply, including the USD 100,000 Minimum Transaction, and a signed Schedule A or addendum can differ. A lead, meeting or signature alone does not trigger payment.
Set up the four tiers before your first introduction, then register for a partner link, check customers against who qualifies and read the program terms, so the report you build later has nothing to take back.