VTEX Post-Purchase Surveys: Fix Your Attribution Gaps
Every order placed on a VTEX store ends on a confirmation page. The customer just paid. They're reading it. They're not distracted by social feeds or inbox noise. For the enterprise brands powering $18.2 billion in annual GMV through VTEX, that moment is the highest-intent, most attentive instant in the entire customer journey — and most are leaving it blank.
Meanwhile, marketing teams argue over attribution: did this sale come from Meta, from Google, from the podcast sponsorship, or from a customer's friend who mentioned the brand? The pixels and UTM tags give one answer. The actual buyer often tells a completely different story.
Post-purchase surveys, embedded directly on the VTEX order confirmation page, close that gap. They capture zero-party attribution data — information customers volunteer themselves — at the exact moment it's freshest and most reliable.
Why Attribution Is Especially Broken on VTEX Stores
VTEX operates 3,400 active online stores across 43 countries, with particular density in Latin America. According to storeleads.app, Brazil accounts for nearly half of all active VTEX storefronts, with Colombia, Argentina, and Chile rounding out the top regions. These are mobile-heavy markets, and mobile shopping behavior creates attribution gaps that desktop-first setups don't face.
Here's the structural problem: approximately half of all VTEX stores run Facebook Pixel, according to storeleads.app technology tracking data. That's a significant chunk of attribution infrastructure resting on a foundation that privacy policy is actively eroding. Apple's App Tracking Transparency (ATT) framework — introduced with iOS 14.5 in April 2021 — requires explicit user consent before any advertising network can track behavior across apps and websites. Flurry, which tracks ATT consent across billions of iOS devices monthly, has documented persistently low opt-in rates since the rollout. Most iOS users are invisible to the Facebook Pixel by default.
Safari independently blocks third-party cookies. Ad blockers are standard features on mobile browsers. Server-side tagging helps, but doesn't fully recover the lost signal — and in Latin American markets where mobile-first shopping is the norm, the problem compounds.
The most accurate attribution signal in your stack isn't inside any ad platform dashboard. It's in your customer's own memory — captured right after they buy.
The gap between what your ad platforms claim and what actually drove those sales grows wider every year.
What the VTEX Confirmation Page Can Actually Do
The VTEX developer documentation makes this clear: the order confirmation page is fully customizable with HTML, CSS, and JavaScript. VTEX supports two technical paths depending on how a store is built.
Legacy Checkout (Classic stores) — Via the Admin panel under Checkout > Websites, you can edit the HTML templates directly: checkout-confirmation-top, checkout-confirmation-header, checkout-confirmation-bottom, and checkout-confirmation-footer. Custom JavaScript goes into checkout-confirmation-custom4.js, accessible under the Code tab. These files are hosted at https://{accountName}.myvtex.com/files/.
VTEX IO Stores — The order-placed app exports two React hooks — useOrderGroup and useOrder — that give full access to current order data. A custom component can be injected into the page's component tree, rendering a survey widget directly beneath the order confirmation message.
Both paths let you place a survey widget at peak post-purchase engagement, before the customer navigates away. Full reference: Customize checkout confirmation pages — VTEX Developers.
Which Survey Questions Deliver Real Attribution Signal
The confirmation page is not the place for a multi-question form. A single well-chosen question — answered in the seconds after checkout — outperforms a five-question email survey sent days later on both response rate and answer quality. For attribution, that question is the HDYHAU: "How did you hear about us?"
Present clean, channel-specific answer options that reflect your actual marketing mix: paid social, Google/search, a specific podcast, influencer organic, friend referral, email, and "Other (please tell us)." Customers know where they came from — they just need you to ask. The write-in option captures channels you haven't thought of yet.
A second question worth rotating in: "Was there anything that almost stopped you from buying?" This isn't attribution — it's conversion optimization. The answers feed A/B test hypotheses that checkout analytics and heatmaps can't surface: payment method anxiety, shipping timeline uncertainty, price hesitation at specific SKUs.
Keep the survey to one question by default. Two maximum. Completion drops sharply beyond that.
For a deeper guide on designing HDYHAU surveys that deliver reliable signal, see How to Run a How Did You Hear About Us Survey.
Setting Up Rauxdata on VTEX: A Practical Walkthrough
Rauxdata generates an embed snippet that works on any storefront that allows custom HTML and JavaScript — including both Classic and IO VTEX stores. Setup takes under ten minutes:
- Create your survey in the Rauxdata dashboard. Set your question, define answer options, and copy the generated embed code.
- Open VTEX Admin and navigate to Checkout > Websites > [your website] > Code.
- For Classic stores: Paste the snippet into
checkout-confirmation-custom4.js(or into thecheckout-confirmation-topHTML template for inline placement). - For IO stores: Add the snippet via a pixel/script-runner app from the VTEX App Store, or inject it through the
order-placedapp customization layer. - Test on a real order — use a test order in your sandbox environment — to confirm the widget renders correctly on mobile and desktop.
No server-side code is required. Survey question changes are reflected live in the embed; no VTEX deployment cycle is needed for content updates.
From Survey Answers to Budget Decisions
HDYHAU data becomes powerful when you compare it against what your ad platforms report.
Picture a brand investing in paid Meta campaigns, Google Performance Max, and a podcast sponsorship. The ad platforms collectively claim attribution for 85% of sales. But over 200 HDYHAU responses, a different story emerges: 30% of buyers say they heard through a friend, 18% cite a specific podcast, and only 25% attribute themselves to a paid Meta ad.
That's a hypothetical scenario — but the structure of the gap it describes is real for most growing brands. The channels that pixels can't see — word of mouth, podcast, organic earned media — consistently outperform what the dashboards credit them for. They're not invisible because they don't work. They're invisible because they don't leave trackable crumbs.
The practical move: sort your HDYHAU responses by channel and compare the distribution against your ad platform attribution. Channels with high survey share but low pixel credit are undervalued and underfunded. Channels with low survey share but high pixel credit are often benefiting from last-touch attribution that doesn't reflect real influence.
For a complete framework on reconciling survey responses with pixel and UTM signals, see Multi-Signal Attribution: Survey, Pixel, and UTM Together.
The LatAm Advantage VTEX Brands Are Leaving Unused
Brazil's LGPD, Mexico's privacy frameworks, and the broader regional trend toward user consent enforcement are all accelerating first-party data strategies across Latin America. VTEX brands operating across these markets already navigate multi-country tax rules, local payment methods, and multi-language storefronts — the platform handles the operational complexity natively.
What the platform can't do is decide what data to collect from buyers. That's a brand-level decision. And the brands that build a zero-party data layer into their confirmation page now will have a compounding advantage as tracking restrictions tighten — the survey answers collected today inform budget decisions in Q3, Q4, and through every algorithm change that scrambles pixel-based attribution in the quarters ahead.
One question. Asked consistently after every order. In every country your VTEX store serves. That's the baseline. Start there.
Stop guessing where your VTEX customers come from.