On this page
HDYHAU (How Did You Hear About Us?)
HDYHAU (How Did You Hear About Us?)
A zero-party data survey deployed on the order confirmation / thank-you page immediately after purchase, asking customers to self-report how they discovered the brand. HDYHAU is the lowest-cost component of the 2026 attribution stack and fills the structural gap left by deterministic tracking: word-of-mouth, podcasts, AI chat recommendations, and private messaging ("Dark Social") produce no trackable click, so pixels and UTM Parameters cannot attribute them. The survey captures what no tag can.
What it is
HDYHAU is classified as Zero-Party Data — voluntarily shared by the customer with explicit intent — making it privacy-compliant by design, unlike first-party cookies or pixels that face increasing regulatory pressure (Fairing, 2024-03-21). The canonical placement is the order status / thank-you page at the moment of highest customer engagement, before purchase-satisfaction sentiment fades (Fairing, 2024-03-21; Byteandbuy, 2025-07-12).
Why it matters: the attribution gap
Refine Labs conducted a 12-month study of 620 converted customers and $21.5M ARR. Results (LeadGen Economy, 2026-06-11):
- Attribution software credited 78% of conversions to web search; customers self-reported only 12% from web search — a 90-percentage-point measurement error on a single channel.
- Podcast attribution showed 53% self-reported contribution against effectively zero software attribution, because a podcast listen produces no captured click.
- 85% of all converted customers self-reported Dark Social (WhatsApp groups, peer email, podcast in-show mentions, conference conversations, AI-search summaries) as a meaningful touchpoint on their path to purchase.
- A hybrid model combining deterministic attribution with HDYHAU survey data recovers 70%+ of the channel attribution signal that the deterministic-only model misses (LeadGen Economy, 2026-06-11).
The structural causes are persistent: iOS App Tracking Transparency global opt-in rates remain near 25% (as-of 2026-06-11), limiting deterministic cross-app attribution indefinitely (LeadGen Economy, 2026-06-11).
Fairing describes the same mechanism: Google last-click attribution systematically over-credits itself — customers who discover a brand via TikTok, a podcast, or word-of-mouth typically search for the brand before buying, handing Google the final attribution credit. Post-purchase surveys reveal this misattribution directly (Fairing blog, 2021-01-05).
Response rates
| Placement | Response rate | Source |
|---|---|---|
| Order confirmation / thank-you page | 50% average (vendor benchmark) (as-of 2024-03-21) | Fairing, 2024-03-21 |
| Order confirmation / thank-you page | "at least 60%" (as-of 2021-01-05) | Fairing blog, 2021 |
| Order confirmation / thank-you page | 40–80% (typical range for DTC brands) (as-of 2025-07-12) | Byteandbuy / 1-800-D2C, 2025-07-12 |
| Specific product preference question (Okendo, Soylent) | 20% (as-of 2024-10-29) | Okendo case study, 2024-10-29 |
| Email survey | 3.24% average (as-of 2025) | Retently, 2025 |
Pre-conversion HDYHAU form fields hurt conversion rate by 8–15%; post-conversion HDYHAU fields capture ~35–55% of converted users with no measurable upstream conversion impact (LeadGen Economy, 2026-06-11).
Question wording and framing
Implementation guidance from sources (Fairing, 2024-03-21; Byteandbuy, 2025-07-12):
- Randomise answer choices to avoid order bias.
- Target first-time customers only — returning customers do not need to re-answer the attribution question.
- Add conditional follow-up logic — if the customer answers "Podcast," prompt for the show name; if "Creator," prompt for the handle; if "AI assistant" or "ChatGPT," prompt for the query topic. This is the mechanic behind Fairing's Advanced Attribution product.
- Include a free-text "Other" field — free text yields richer insights, particularly for capturing emergent channels that were not on the original answer list (Byteandbuy / Recast, 2025-07-12).
- Fairing's proprietary question bank was reviewed by a survey methodologist for accuracy and bias reduction (Fairing, 2024-03-21).
Software and tooling
Fairing is the dominant dedicated HDYHAU tool for Shopify. It is a Shopify Certified Technology Partner, installs on the order status page, integrates natively with Klaviyo and Google Sheets, and provides LTV analysis segmented by survey response (Fairing, 2024-03-21). Fairing launched:
- Advanced Attribution (announced 10 Sep 2026, GA planned Jan 2027): resolves vague answers like "Instagram" to the specific creator or podcast by surfacing an auto-suggest list when customers type "Other" free text. Early rollouts showed up to 50% more usable attribution signal with no change in survey placement or response rate (SalesTechStar, 2026-09-10). (as-of 2026-09-10; beta, GA timeline could shift)
- Fairing MCP: a Model Context Protocol product connecting attribution survey data to Claude and ChatGPT, allowing brands to query HDYHAU data in plain English (e.g. "rank discovery channels by average order value", "count how many customers named a specific podcast this month") (Fairing product page, 2026; YouTube bxQiS95P_YA, Fairing channel, ~Aug 2026).
Other tools in the space:
| Tool | Notes | Source |
|---|---|---|
| KnoCommerce | Flexible survey templates for CX, CRO, and attribution | Fairing blog, 2025 |
| Okendo Surveys | Shopify Plus Certified App; 9 question types; conditional logic; Soylent case study (20% response rate on product preference question) | Okendo, 2024-10-29 |
| Triple Whale | Post-purchase survey bundled with pixel attribution dashboard; pixel takes priority when data conflicts with survey; Shopify, BigCommerce, WooCommerce | Triple Whale product page |
| Northbeam | ML / first-party multi-touch attribution platform; recommends HDYHAU as a validation layer, not as its primary product feature; ~$1,500/month starting (as-of 2025-07-12) | Northbeam blog; Byteandbuy, 2025-07-12 |
How retailers analyse results
The recommended weekly decision loop (Byteandbuy, 2025-07-12):
- Ingest post-purchase survey (PPS) responses to a BI tool, joined to order ID, product purchased, and customer cohort LTV.
- Group free-text responses into channel / creator / content-angle tags.
- Stack survey share vs Meta / Google / GA4 and flag channels where survey share exceeds platform attribution by 2×.
- Shift 10–20% of spend toward those channels and re-read after two weeks.
Fairing provides LTV analysis segmented by HDYHAU response, enabling budget allocation toward channels that recruit higher-value buyers even when last-click attribution is muted (Byteandbuy, 2025-07-12).
ADOPTER Media (podcast advertising) uses a survey-informed multiplier — survey-attributed conversions ÷ direct-code-attributed conversions — to reconcile podcast listener attribution with trackable redemptions at the show level (Byteandbuy, 2025-07-12).
Position in the 2026 measurement stack
HDYHAU is the cheapest component of the four-part 2026 attribution stack: HDYHAU + incrementality + Media Mix Modeling (MMM) + signal capture (LeadGen Economy, 2026-06-11). The recommended model is triangulation — self-reported survey data paired with platform attribution and sanity-checked with Incrementality Testing or MMM. HDYHAU supplements rather than replaces other methods (Byteandbuy, 2025-07-12).
Adoption context (as-of 2026):
- 52% of US marketers use incrementality testing as of July 2025; 60% describe it as most-trusted measurement method as of May 2026 (Haus, per LeadGen Economy, 2026-06-11).
- 46.9% of US marketers planned to invest in MMM in the next 12 months (eMarketer, Oct 2025, per LeadGen Economy, 2026-06-11).
- Google open-sourced its Meridian Bayesian MMM framework 29 Jan 2025; on 20 May 2026 Google announced Meridian integration into GA360, making Bayesian MMM accessible at mid-market at no incremental cost above the GA360 subscription (LeadGen Economy, 2026-06-11).
LLMs as emerging discovery channel
Fairing's benchmark data (as-of Q1 2025) from 2,500+ brands (Fairing, 2025-04):
| LLM | Share of LLM-attributed transactions (as-of Q1 2025) |
|---|---|
| ChatGPT | 64.6% |
| Grok | 12.6% |
| Perplexity | 6.1% |
| Gemini | 3.1% |
| Claude | 1.8% |
By April 2025, 11.8% of Fairing customers had at least one customer cite an LLM in HDYHAU survey responses — up from near-zero in May 2024. LLM mentions grew 5.4× from January to April 2025. Most brands see 0–1% of survey responses mentioning LLMs; the average per-brand LLM share is ~0.2–0.3% with a high-LLM-skewed tail at some consumer-facing niches (Fairing, 2025-04).
Fairing Q4 2025 benchmark (email-gated) covers ChatGPT, Gemini, Grok, and others with industry-specific breakdowns. Discovery signal from HDYHAU surveys appears to be the only real-time window into LLM-driven commerce discovery that brands can currently access at the product/transaction level.
Case studies
- Weezie (luxury towels): Fairing surveys revealed ~35% of business attributed to word-of-mouth referrals, confirming referral programme ROI and informing a 7-figure advertising budget (Fairing, 2024-03-21).
- Eastern Standard Provisions (food): HDYHAU revealed 38% of customers came from paid social; product taste surveys informed near-real-time product development (Okendo, 2024-10-29).
- Soylent (nutrition DTC): post-purchase attribution revealed 30–40% of customers find products in retail stores; brand used data to build a retail trial incentive with Aisle. Achieved 20% survey response rate (Okendo, 2024-10-29).
Gaps and open questions
- Fashion/apparel-specific data absent. All published case studies are food, nutrition, luxury home goods, or broad DTC. No fashion-specific response rates, channel distribution benchmarks, or implementation patterns are publicly documented.
- Non-Shopify / headless implementation undocumented. Fairing and KnoCommerce are Shopify-native. Large retailers on composable / headless stacks (e.g. custom checkout extensions) have no public implementation guides.
- GA4 / data warehouse integration specifics. How HDYHAU response data is joined to GA4 event data or a data warehouse in practice is mentioned but not documented with specifics.
- GDPR / EU implications. Zero-party data is privacy-compliant by design (URL-only storage; customer-volunteered), but no source addresses specific EU/EEA GDPR obligations for post-purchase survey data storage, retention limits, or consent flow requirements.
Key terms
| Term | Meaning |
|---|---|
| Zero-party data | Data voluntarily shared by the customer — the legal and methodological basis for HDYHAU |
| HDYHAU | "How Did You Hear About Us?" — the canonical question, placed post-purchase |
| Question Stream | Fairing's product name for the multi-question survey flow beyond the single attribution question |
| Advanced Attribution | Fairing product (Sep 2026 beta) that resolves vague answers to specific creator/podcast |
| Triangulation | Measurement philosophy combining HDYHAU + platform attribution + incrementality/MMM |
| Dark Social | Traffic/influence arriving without a referrer header — the main gap HDYHAU fills |
Frontier links
Post-Purchase Survey · LLM Commerce Discovery · Fairing · Podcast Advertising Attribution · Incrementality Testing · Marketing Mix Modelling · RLSA · Media Mix Modeling (MMM)