Research brief: trust, referral networks, and in-group reputation in Ontario's trades economy (May 2026 — Foundation Brief #3)
Created 2026-05-25
Status: Third foundation brief in the psychology cluster, companion to [[research-brief-psychology-gc-marketing-aversion-may-2026]] (Brief #1) and [[research-brief-risk-aversion-post-failure-may-2026]] (Brief #2). May 2026, ~10,500 words. Internal reference; not customer-facing.
TL;DR
- Referred customers in service businesses have ~16% higher LTV than matched non-referred customers. Schmitt-Skiera-Van den Bulte 2011 (10,000 German retail-bank customers, 33 months). Retention advantage persists; margin advantage decays. See
[[schmitt-skiera-van-den-bulte-2011-16pct-ltv-premium]]. - WOM elasticity is ~0.53 — roughly 20× marketing-events elasticity, 30× media-appearances elasticity (Trusov-Bucklin-Pauwels 2009, online social network). See
[[trusov-bucklin-pauwels-2009-wom-elasticity-053]]. - Trust mediates the strong-tie advantage in useful knowledge transfer. Once competence- and benevolence-based trust are controlled for, the structural benefit of weak ties re-emerges — Levin & Cross 2004. See
[[levin-cross-2004-trust-mediates-tie-strength]]. - Closeness is the best tie-strength indicator, NOT frequency or duration of contact (Marsden-Campbell 1984). See
[[marsden-campbell-1984-closeness-best-tie-strength]]. - The Ontario residential-construction signal stack has three audit-verifiable tiers and two purchasable tiers. HCRA + Tarion + Gold Seal are audit-verifiable; RenoMark is membership-screened-with-attestation; Best of Houzz badges and pay-to-play "Top Contractor" lists are engagement- or fee-driven. See
[[signalling-theory-applied-ontario-credential-stack]]. - Guildmaster Awards are audit-verified survey signals, not paid badges; ≥90% Likely-to-Recommend rate against ~70% industry average, ≥20 responses, disqualification for omitting customers. See
[[guildmaster-guildquality-audit-verified-survey-signal]]. - In-group reputation and market reputation are not the same construct. Rindova et al. 2005: "being good" (perceived quality among informed stakeholders) vs "being known" (prominence). For high-ticket residential/ICI work, in-group dominates market. See
[[rindova-2005-being-good-vs-being-known]]. - The credence-good problem is solved by liability + verifiability, not by reputation alone. Dulleck-Kerschbamer-Sutter 2011 found experimentally that "liability has a crucial, but verifiability at best a minor, effect" on credence-good market efficiency, and "allowing sellers to build up reputation has little influence." See
[[dulleck-kerschbamer-sutter-2011-liability-verifiability-credence-goods]]. - Spence's signalling theory predicts purchased signals collapse when cost drops. Best of Houzz badges and pay-to-play lists are low-cost-to-acquire and carry low signalling value among informed GC buyers; HCRA licensure + Gold Seal certification are costly and retain value. See
[[spence-1973-job-market-signaling]]. - Peer-coached contractors (BTA / EOS / Vistage / TAB) are a meaningfully different persona. BTA reports 600+ contractor companies across North America; EOS reports 200K+ businesses, 800-1000 implementers; Vistage 45K members. Causal direction (coaching → marketing maturity) unproven; correlation strong enough to act on. See
[[peer-coached-contractor-persona-synthesis]]. - The strongest aversion countermeasure available to Candid is becoming a member of the network whose information flows it wants to participate in. Marketing assets without network membership are noise; network membership generates signal whether or not marketing assets exist. This brief's strongest take and a refinement of how to read Brief #1. See
[[research-brief-trust-networks-in-group-reputation-may-2026]]Section 6 below +[[rule-network-membership-precedes-referral-program-design]].
The eight strategic implications (R1–R8)
[[rule-network-membership-precedes-referral-program-design]][[rule-reweight-marketing-portfolio-toward-in-group-accrual]][[rule-target-peer-coached-contractor-persona-with-explicit-qualification]][[rule-earn-one-audit-verified-credential-per-year-refuse-purchased]][[rule-better-copy-is-wrong-strategy-credence-good-requires-liability]][[rule-build-candid-audited-reputation-infrastructure]][[rule-treat-events-as-reputation-witnessing-not-lead-generation]][[rule-name-credence-good-problem-explicitly-propose-liability-structures]]
Sections
- Social topology of the Ontario trades network —
[[granovetter-1973-strength-of-weak-ties]],[[granovetter-1983-weak-ties-revisited-walked-back]],[[burt-structural-holes-1992-2004-broker-advantage]],[[marsden-campbell-1984-closeness-best-tie-strength]],[[levin-cross-2004-trust-mediates-tie-strength]],[[ontario-trades-info-ecosystem-five-layer]],[[brokers-in-ontario-trades-three-roles]]. - Referral mechanics —
[[schmitt-skiera-van-den-bulte-2011-16pct-ltv-premium]],[[van-den-bulte-2018-better-matching-social-enrichment]],[[trusov-bucklin-pauwels-2009-wom-elasticity-053]],[[warm-intro-hierarchy-eight-levels-ontario-trades]],[[expectation-setting-retention-by-tie-strength]]. - In-group vs market reputation —
[[rindova-2005-being-good-vs-being-known]],[[in-group-vs-market-reputation-trades-distinction]],[[opacity-specificity-tradeoff-reputation-channels]],[[why-in-group-dominates-three-compounding-mechanisms]],[[dulleck-kerschbamer-sutter-2011-liability-verifiability-credence-goods]]. - Accreditation and peer-network infrastructure —
[[red-seal-endorsement-individual-journeyperson-credential]],[[bta-contractor-growth-method-600-contractors]],[[eos-worldwide-scale-200k-businesses-800-1000-implementers]],[[vistage-tab-generalist-peer-advisory-not-trades-specific]],[[peer-coached-contractor-persona-synthesis]],[[events-as-trust-witnessing-not-lead-generation-infrastructure]]; existing entries referenced: HCRA launched February 1, 2021 as Ontario's new-home builder regulator; 7,232 licensees at March 31, 2025, HCRA Ontario Builder Directory is the load-bearing public discipline regime for Ontario new home builders — covers 6,500+ licensed builders, publishes conduct findings / charges / convictions, Tarion established 1976 under the Ontario New Home Warranties Plan Act; not-for-profit administrative authority, not a Crown agency, Gold Seal origin: launched 1991 by the Canadian Construction Association to create a national, portable construction-management credential, RenoMark origin: launched 2001 by BILD (then GTHBA); trademark transferred from BILD-GTA to CHBA in June 2024, BILD GTA: formed 2008 from GTHBA + UDI/Ontario merger, "more than 1,000 member companies," threatened in 2023 to leave OHBA over a "one-voice" operating model dispute,[[breakthrough-academy-coaching-self-reported-outcomes]]. - Earned vs purchased reputation —
[[spence-1973-job-market-signaling]],[[signalling-theory-applied-ontario-credential-stack]],[[best-of-houzz-low-cost-engagement-driven-signal]],[[guildmaster-guildquality-audit-verified-survey-signal]],[[audit-gap-problem-pooling-equilibrium-degradation]],[[sophisticated-gc-five-question-credential-heuristic]]. - Research gaps —
[[trust-networks-brief-research-gaps-may-2026]].
Relation to Briefs #1 and #2
- Brief #1 (
[[research-brief-psychology-gc-marketing-aversion-may-2026]]): This brief refines the operational interpretation. If Brief #1 is read as "build better marketing assets to overcome aversion," that read is wrong. The aversion is downstream of the credence-good problem ([[dulleck-kerschbamer-sutter-2011-liability-verifiability-credence-goods]],[[marketing-services-as-credence-good-for-gc]]) — which is not solved by asset quality but by liability + verifiability + in-group reputation. Brief #1's empirical claim (48.7% of builders source >50% of sales from referrals —[[apb-sorci-2024-48-7pct-referrals-half-of-sales]]) points the same direction. - Brief #2 (
[[research-brief-risk-aversion-post-failure-may-2026]]): Amplifies, not contradicts. Brief #2's "social proof works only when the reference group is recognizably proximate" ([[goldstein-cialdini-griskevicius-2008-provincial-norms-hotel-towel]],[[rule-radius-matched-peer-social-proof-only]]) is operationalized here as: proximity in Levin-Cross terms is trust, not closeness. Proximity-by-shared-cohort (BTA, EOS, RenoMark, BILD) is stronger than proximity-by-geography. A Toronto GC referred by a BTA cohort peer in Calgary is a closer tie than a Toronto GC referred by a geographically-adjacent acquaintance.
Related
Referenced by (38)
- reference Granovetter 1973 (AJS) — strength of weak ties; surveyed Newton MA men found jobs more through acquaintances than close friends relates-to
- reference Granovetter 1983 (Sociological Theory) — weak-ties claim walked back; strong ties carry relational guarantees weak ties cannot, esp. for trust under uncertainty relates-to
- reference Burt 1992 (Structural Holes) / 2004 (AJS) — brokers spanning holes between groups capture information-arbitrage advantage relates-to
- reference Marsden & Campbell 1984 (Social Forces) — closeness is the best tie-strength indicator, NOT frequency or duration relates-to
- reference Levin & Cross 2004 (Mgmt Sci) — competence- and benevolence-based trust mediate tie-strength → useful knowledge transfer; once trust is controlled for, weak ties re-emerge as valuable relates-to
- reference Ontario residential-construction information ecosystem — five descriptive layers (regulators / trade assocs / suppliers / events / digital) relates-to
- reference Brokers in Ontario trades — three roles span structural holes: supplier outside sales reps, specialist subs, peer-coach implementers relates-to
- reference Schmitt, Skiera, Van den Bulte 2011 (JoM) — referred customers ≥16% higher LTV than matched non-referred; retention persists, margin decays relates-to
- reference Van den Bulte, Bayer, Skiera, Schmitt 2018 (JMR) — referred-customer LTV decomposes into "better matching" (temporary) + "social enrichment" (persistent) relates-to
- reference Trusov, Bucklin, Pauwels 2009 (JoM) — WOM elasticity on signups ≈ 0.53, roughly 20× marketing-events elasticity, 30× media-appearance elasticity relates-to
- reference Warm-intro hierarchy — 8 levels from cold inbound to family-equivalent referral; closest tie ≠ best referrer unless it carries domain-competent trust relates-to
- reference Expectation transfer + retention by tie strength — strong-tie referrals carry expectation specificity + active social-enrichment retention; weak-tie referrals churn at near-baseline rates relates-to
- reference Rindova, Williamson, Petkova, Sever 2005 (AMJ) — reputation is bidimensional: perceived quality ("being good") vs prominence ("being known"); prominence drove price premium relates-to
- reference In-group vs market reputation — for trades, in-group ("being good") dominates market ("being known"); applied translation of Rindova et al. relates-to
- reference Opacity-specificity tradeoff — in-group reputation is information-rich + opaque to outsiders; market reputation is transparent + information-poor relates-to
- reference Why in-group reputation dominates market reputation for trades — credence-good + loss-aversion + low-frequency selection compound relates-to
- reference Dulleck, Kerschbamer, Sutter 2011 (AER) — credence-goods lab experiment; "liability has crucial effect, verifiability minor, reputation little influence" relates-to
- reference Red Seal endorsement — national journeyperson credential; >80% of Canadian apprentices in Red Seal trades; 4-hour multi-choice exam, 70% pass, individual not firm relates-to
- reference Breakthrough Academy (BTA) — Vancouver-HQ, founded 2015; "Contractor Growth Method"; 600+ contractor companies (~30% Canadian / 70% US); Contractor Evolution podcast relates-to
- reference EOS Worldwide — "Over 200,000 businesses around the world"; 800-1000+ implementers; Traction (Wickman 2007/2011); heavy Ontario contractor adoption >$3M revenue relates-to
- reference Vistage (45K members in 35 countries) + The Alternative Board (franchised local-board model) — generalist CEO peer-advisory, not trades-specific; both serve Ontario GCs relates-to
- reference Peer-coached contractor persona — quarterly planning cadence, dashboards, cohort-mediated vendor decisions; highest-value AND highest-risk Candid prospect relates-to
- reference Events as trust-witnessing infrastructure — HBA dinners / awards nights are public ratification venues; measure share-of-voice, not leads-attributed relates-to
- reference Spence 1973 (QJE) — job-market signaling; signals work only when costlier for low-quality types to acquire than high-quality types; cost-drop collapses pooling equilibrium relates-to
- reference Signalling theory applied to Ontario credentials — HCRA / Tarion / Gold Seal / Guildmaster are costly; RenoMark moderate; Houzz / pay-to-play approach zero relates-to
- reference Best of Houzz badges — ~3% of 2.5M pros win annually; gating is user engagement on Houzz platform, not third-party audit; low-cost signal relates-to
- reference Guildmaster Awards (GuildQuality) — audit-verified survey signal; ≥90% LTR vs ~70% industry average; 20 responses min; disqualification for omitting customers relates-to
- reference Audit-gap problem — visual badge similarity collapses Spence pooling equilibrium; sophisticated buyers discount badge category and ask for underlying audit relates-to
- reference Sophisticated GC's 5-question credential-evaluation heuristic — revocability / written threshold / liability / individual vs firm / in-group survivability relates-to
- reference Research gaps and source caveats — trust / referral networks / in-group reputation brief (May 2026 Brief #3) relates-to
- rule R1 — Network membership precedes referral-program design; "we have a referral program" is not a strategy until "we are in the network" is true relates-to
- rule R2 — Reweight marketing portfolio toward in-group reputation accrual; reduce reliance on homeowner-targeted lead-gen and agency-directory listings relates-to
- rule R3 — Target the peer-coached contractor as highest-value persona; explicitly qualify BTA/EOS/Vistage/TAB membership in discovery; staff for the persona's information environment relates-to
- rule R4 — Earn one audit-verified in-group credential per year; refuse all purchased credentials; sparse + verifiable beats dense + decorative relates-to
- rule R5 — Recognize "better marketing copy" is the wrong strategy for a credence-good market; solution is liability + verifiability + in-group reputation, not copy relates-to
- rule R6 — Build Candid's own audited-reputation infrastructure (GuildQuality-style for marketing-agency-for-trades); the category lacks one — structural-hole opportunity relates-to
- rule R7 — Treat events as reputation-witnessing infrastructure, not lead-generation; measure share-of-voice within defined in-group, not leads-attributed-to-event relates-to
- rule R8 — Name the credence-good problem explicitly with prospects; propose liability + verifiability structures rather than projecting unwarranted confidence relates-to