R1 — Network membership precedes referral-program design; "we have a referral program" is not a strategy until "we are in the network" is true

Rule: Stop treating "we have a referral program" as a strategy. Start treating "we are members of the network whose information we want to participate in" as the strategy.

Why: Referral programs (in the Schmitt-Skiera-Van den Bulte sense — [[schmitt-skiera-van-den-bulte-2011-16pct-ltv-premium]]) are an incentive layer on top of an existing network. The Ontario trades network exists with or without Candid; the question is whether Candid is in it. Network membership precedes referral-mechanism design. [[expectation-setting-retention-by-tie-strength]] shows that referral depth > referral volume for lifetime value, which means cultivating fewer, deeper referrer relationships beats incentivizing broad weak-tie referrals.

How to apply:

• Audit current "referral program" infrastructure (incentive emails, "refer a friend" pages, partner programs). If the underlying referrer base is thin, the program is applying force to nothing. Pause or deprioritize.
• Identify the 20-50 highest-quality potential referrer relationships Candid should cultivate over the next 12 months (peer-coach implementers, well-connected supplier reps, BILD/OHBA committee members, alumni clients in BTA/EOS cohorts). Invest in those relationships explicitly.
• Once a network base exists, then layer a referral-incentive program on top to amplify what is already flowing.