and runs during HTML parse, but hubspotutk is set by js.hs-analytics.net, measured loading at ~834ms. On a first visit the cookie does not exist yet, so a single read returns nothing and the visitor gets no external ID at all — and a first visit is exactly what an ad click is. So poll briefly until HubSpot sets it. The SDK loads lazily (~4.5s) and replays a queue, so an init queued a second or two in still lands before any events are transmitted. */ (function attachExternalId() { var attempts = 0; var MAX_ATTEMPTS = 40; /* 40 x 200ms = 8s, then give up quietly */ var timer = null; function tryAttach() { var raw = getExternalId(); if (!raw) { if (++attempts >= MAX_ATTEMPTS && timer) clearInterval(timer); return; } if (timer) clearInterval(timer); sha256Hex(raw).then(function (hash) { oaiq("init", { user: { external_id_sha256: hash } }); }).catch(function () {}); } tryAttach(); /* returning visitor: immediate */ if (!getExternalId()) timer = setInterval(tryAttach, 200); })(); /* --------------------------------------------------------------------------- 2. Load the SDK off the critical path Whichever comes first: the main thread going idle after load, or the first sign of real user intent. No preconnect on purpose — a TLS handshake during LCP for a script we deliberately deferred is a net loss. Measured on clickpost.ai: load event 3614ms, SDK fetched 4508ms. ------------------------------------------------------------------------ */ var loaded = false; var INTENT = ["pointerdown", "keydown", "scroll"]; function loadSdk() { if (loaded) return; loaded = true; INTENT.forEach(function (t) { removeEventListener(t, loadSdk, { passive: true }); }); var s = document.createElement("script"); s.src = SDK; s.async = true; s.fetchPriority = "low"; document.head.appendChild(s); } INTENT.forEach(function (t) { addEventListener(t, loadSdk, { passive: true, once: true }); }); addEventListener("load", function () { window.requestIdleCallback ? requestIdleCallback(loadSdk, { timeout: 2500 }) : setTimeout(loadSdk, 1500); }, { once: true }); /* --------------------------------------------------------------------------- 3. Conversion id sha256(lowercase(email) + "|" + formId + "|" + ISO-hour), first 20 hex, "lead_" prefix. - email lowercased/trimmed so casing does not fork the id - formId keeps two different forms in the same hour from colliding - the hour bucket is only the server's fallback; the id computed here is sent to the relay verbatim, so the two paths cannot disagree even when a submit straddles an hour boundary Ads Manager's own sample omits event_id. We send it so the pixel event and the relay event collapse into one conversion instead of counting twice. ------------------------------------------------------------------------ */ function buildEventId(email, formId) { var seed = [ String(email || "").trim().toLowerCase(), String(formId || ""), new Date().toISOString().slice(0, 13) ].join("|"); return crypto.subtle.digest("SHA-256", new TextEncoder().encode(seed)).then(function (buf) { var hex = Array.prototype.map.call(new Uint8Array(buf), function (b) { return b.toString(16).padStart(2, "0"); }).join(""); return "lead_" + hex.slice(0, 20); }); } /* Guards a re-fire from the history.pushState that book-a-demo-core.js runs after submit, and from back-navigation onto the same state. */ function firedOnce(id) { try { if (sessionStorage.getItem("oaiq_" + id)) return true; sessionStorage.setItem("oaiq_" + id, "1"); return false; } catch (e) { return false; /* private mode / blocked storage: fire rather than lose it */ } } /* Path 2. keepalive lets the request finish even if the page navigates away. Failures are swallowed deliberately: a reporting call must never surface an error to someone who just submitted a form. The relay hashes the email server-side into user.emails_sha256 and adds the caller's IP and user agent, so this event carries the same match signals the pixel gets for free. */ function sendToRelay(email, formId, eventType, eventId) { try { var payload = { token: RELAY_TOKEN, email: email, form_id: formId || "", event_type: eventType, event_id: eventId, /* verbatim, so both paths always match */ source_url: location.href }; var oppref = getOppref(); if (oppref) payload.oppref = oppref; var externalId = getExternalId(); if (externalId) payload.external_id = externalId; fetch(RELAY_URL, { method: "POST", keepalive: true, headers: { "Content-Type": "application/json" }, body: JSON.stringify(payload) }).catch(function () {}); } catch (e) {} } /* Capture the email while the form is still on the page. onFormSubmitted does not reliably carry submissionValues, and by then the form may be removed. */ var lastEmail = ""; document.addEventListener("submit", function (e) { var f = e.target; if (!f || !f.querySelector) return; var i = f.querySelector('input[name="email"], input[type="email"]'); if (i && i.value) lastEmail = i.value; }, true); /* Every HubSpot form on the site reports through this one message event, so no module or theme JS needs editing to add coverage. */ window.addEventListener("message", function (e) { var d = e.data; if (!d || d.type !== "hsFormCallback" || d.eventName !== "onFormSubmitted") return; var email = (d.data && d.data.submissionValues && d.data.submissionValues.email) || lastEmail; if (!email) return; /* without an email there is no stable id to dedupe on */ var eventType = EVENT_BY_FORM[d.id] || EVENT_DEFAULT; buildEventId(email, d.id).then(function (id) { if (firedOnce(id)) return; oaiq("measure", eventType, { type: DATA_TYPE }, { event_id: id }); /* path 1 */ sendToRelay(email, d.id, eventType, id); /* path 2 */ }); }); })();

91ÊÓƵ

Assess Your Post-Purchase Brand Readiness Start Assessment

ROI Calculator · Returns

Turn returns into
retained revenue

Estimate the revenue you recover with automated returns and exchanges. Calculate savings on return handling costs with ClickPost's free returns calculator.

1. Tell us about your business

Defaults are typical benchmarks. Change any value to fit your business.

2. Fine-tune the impact assumptions

Defaults are ClickPost platform benchmarks. Leave them for a conservative estimate, or match your own data.

How it works

Reduce returns in four steps

ClickPost intercepts returns at every stage — from delivery confirmation to exchange completion — before the refund leaves your business.

div.cp-roi__hvis (12)

Catch the reason at the source

Record why each item comes back, size, fit, defect or a change of mind, so you act on data instead of a hunch.

div.cp-roi__hvis (13)

Offer an exchange first

Put the right size or a close match in front of the shopper at the moment of return. Many take it, and the sale holds.

div.cp-roi__hvis (14)

Deflect the refunds you can

For low-value or hard-to-resell items, a keep-the-item or store-credit offer beats paying reverse postage twice.

div.cp-roi__hvis (15)

Win the next order

A return handled well is a reason to come back. Trigger the follow-up that turns a resolved return into a repeat sale.

Cost anatomy

Eight factors that determine e-commerce return costs

Understanding what drives return costs helps you prioritize where to intervene.
Reverse shipping

Reverse shipping

You pay to bring the item back, often at a worse rate than the outbound leg cost you.

Processing and QC

Processing and QC

Every returned unit has to be received, inspected, graded and re-shelved. That labor scales with volume.

Refund leakage

Refund leakage

A cash refund ends the relationship and takes the revenue with it. An exchange or credit does not.

Resale write-down

Resale write-down

Opened or worn items rarely sell again at full price. The gap between original and recovered value is a real loss.

Orders sent back before delivery

Orders sent back before delivery

An order that bounces before it is delivered pays both legs and earns nothing. It is often the single biggest line.

Fit and sizing churn

Fit and sizing churn

Apparel and footwear come back for fit more than anything else. Small catalog and PDP fixes move this number.

Support handling

Support handling

Return tickets pull agents off higher-value work. A self-serve return flow takes most of that load away.

Lost lifetime value

Lost lifetime value

A shopper burned by a clumsy return may never order again. That churn dwarfs the cost of the one return.

The math

ROI calculation formulas

Transparent methodology. These are the formulas the calculator runs. The benchmark constants inside them come from ClickPost platform data, and you can change every rate in the advanced panel.
Step 1
Eligible returns
Nelig = N × rf × (1 − m) + N × ro

Fit returns after multi-size buyers are removed, plus returns for every other reason.

Step 2
Shoppers reached
Neng = Nelig × S × Ep + (N − Nelig) × S × Eh

Returners you contact, plus happy customers you re-engage through post-purchase outreach.

Step 3
Exchange value
Vexch = Nelig × S × Ep × Lexch × (AUR − Cs − AUC)

Margin kept each time a return converts into an exchange instead of a refund.

Step 4
Keep-item value
Vkeep = Nelig × S × Ep × Lkeep × (AUR + Cr − AUC × p)

Value saved when a refund is deflected to a keep-the-item or credit outcome.

Step 5
Retained-shopper value
Vret = Neng × Lret × (U × (AUR − AUC − Cs) − Mr)
Profit from the next order a retained shopper places.
Step 6
Return-handling saved
Nelig × S × Ep × Lkeep × Cr
Reverse-logistics cost avoided on the refunds you deflect.
Total ROI
Vexch + Vkeep + Vret + cost saved
N = annual orders · AUR = price per item · AUC = cost per item · rf = fit return rate · ro = other return rate · m = multi-size share · S = collection rate · Ep/Eh = outreach rates · Lexch/Lkeep/Lret = conversion lifts · Cs = shipping per unit · Cr = return cost · U = units per order · p = resale probability · Mr = repeat acquisition cost

Powering post-purchase for 600+ brands worldwide

  • GOPURE

    ClickPost helped us create a return experience that fit our business instead of forcing us to fit the platform. They customized our workflows without additional costs or developer support. The automation they built for our returns, freed up 5 hours a week.

    Sandy Vangyi
    Sandy Vangyi
    Director of CX, GOPURE
  • BG

    We’ve been using Clickpost and it’s genuinely been super helpful. The post-purchase upsells have been a solid bonus for incremental revenue.

    Samarth Sindhi
    Samarth Sindhi
    Chief Executive Officer, MarsGHC
  • VanillaPura - BG

    Great experience and fantastic support from ClickPost. This immediately cut our top post-purchase support tickets, making ROI easy to quantify. Plus, the customization features and upsell tools help increase AOV. Highly recommend.

    Paul Fulton
    Paul Fulton
    VanillaPura
High Performer  Mid Market - Winter 2026
Highest User Adoption - Mid Market - Winter 2026
Momentum Leader - Winter 2026
User Loves Us
High Performer  Mid Market - Winter 2026
Highest User Adoption - Mid Market - Winter 2026
Momentum Leader - Winter 2026
User Loves Us

4.8 ★

out of 5 stars on G2.com

Users Love Us - G2 Winter 2026
Momentum Leader - Winter 2026
Highest User Adoption - Winter 2026
High Performer Mid-Market - Winter 2026
Users Love Us - G2 Winter 2026
Momentum Leader - Winter 2026
Highest User Adoption - Winter 2026
High Performer Mid-Market - Winter 2026
Users Love Us - G2 Winter 2026
Momentum Leader - Winter 2026
Highest User Adoption - Winter 2026
High Performer Mid-Market - Winter 2026
Users Love Us - G2 Winter 2026
Momentum Leader - Winter 2026
Highest User Adoption - Winter 2026
High Performer Mid-Market - Winter 2026

4.8 ★

out of 5 stars on G2.com

5B+

shipments automated per year

Post-purchase that plugs into
your existing stack

  • 600+ Carrier Partners
  • 75+ Communication Gateways
  • 50+ Storefronts, OMS, WMS
  • Shopify
  • Magento
  • Klaviyo
  • Attentive
  • Shipbob
  • Shiphero
  • UPS
  • USPS

It estimates the revenue you keep by converting refunds into exchanges and store credit, plus the processing cost you save per return. Instead of measuring return rate, it measures retained revenue, which is the number that actually protects margin.

Your return rate, average order value, the split between refunds and exchanges today, and your cost to process one return. Defaults are provided, but this tab is most accurate when you enter your own return and processing numbers.

They are category-aware, since a healthy return rate for apparel looks nothing like one for electronics or beauty. The calculator starts from US category averages and lets you set your own rate, so the estimate reflects what you actually see, not a blended number.

The return flow leads with a size swap, color change, or store credit before it offers a refund, so the customer stays and the order value stays with you. A large share of would-be refunds convert to exchanges when the flow is built this way.

Yes, you set the rules. Policies decide which outcomes each customer sees, by product and segment, and you can offer bonus credit for exchanges or apply a fee on refund-heavy behavior. The flow only presents what your policy allows.

Payback comes from two places at once: revenue retained on each exchange and lower processing cost per return. The lift scales with your return volume and how refund-heavy you are today, so higher-return categories like apparel usually see the biggest gain.

. window.disableHubSpotCookieBanner = true; function killNative() { var n = document.getElementById("hs-eu-cookie-confirmation"); if (n && n.parentNode) { n.parentNode.removeChild(n); return true; } return false; } killNative(); if (window.MutationObserver) { var mo = new MutationObserver(function () { killNative(); }); mo.observe(document.documentElement, { childList: true, subtree: true }); setTimeout(function () { mo.disconnect(); }, 15000); } var COOKIE = "cp_cookie_consent"; var EU = ["at","be","bg","hr","cy","cz","dk","ee","fi","fr","de","gr","hu","ie","it","lv","lt","lu","mt","nl","pl","pt","ro","sk","si","es","se","gb","uk","is","li","no","ch"]; function readCookie(nm) { var m = document.cookie.match(new RegExp("(?:^|; )" + nm + "=([^;]*)")); return m ? decodeURIComponent(m[1]) : null; } function writeCookie(nm, val, days) { var d = new Date(); d.setTime(d.getTime() + (days * 864e5)); document.cookie = nm + "=" + encodeURIComponent(val) + ";expires=" + d.toUTCString() + ";path=/;SameSite=Lax" + (location.protocol === "https:" ? ";Secure" : ""); } function stored() { var raw = readCookie(COOKIE); if (!raw) return null; try { var o = JSON.parse(raw); if (typeof o.analytics === "boolean") return o; } catch (e) {} return null; } function apply(state, revoke) { window._hsp = window._hsp || []; window._hsp.push(["setHubSpotConsent", { analytics: !!state.analytics, advertisement: !!state.advertisement, functionality: !!state.functionality }]); if (revoke && !state.analytics && !state.advertisement && !state.functionality) { window._hsp.push(["revokeCookieConsent"]); } window.dataLayer = window.dataLayer || []; window.dataLayer.push({ event: "cp_cookie_consent", cp_consent_analytics: !!state.analytics, cp_consent_advertisement: !!state.advertisement, cp_consent_functionality: !!state.functionality }); document.documentElement.setAttribute("data-cp-consent", "set"); } function inScope(banner) { if (banner.getAttribute("data-scope") !== "eu_only") return true; var geo = (document.body.getAttribute("data-geo_location") || "").toLowerCase(); if (!geo) return true; return EU.indexOf(geo) !== -1; } function init(banner) { var days = parseInt(banner.getAttribute("data-days"), 10) || 180; var prefs = banner.querySelector("[data-cpcb-prefs]"); var compact = banner.querySelector('[data-cpcb-actions="compact"]'); var expanded = banner.querySelector('[data-cpcb-actions="expanded"]'); var toggles = banner.querySelectorAll("[data-cpcb-toggle]"); function show() { banner.hidden = false; requestAnimationFrame(function () { banner.classList.add("is-visible"); }); } function hide() { banner.classList.remove("is-visible"); setTimeout(function () { banner.hidden = true; }, 300); } function collapse() { if (!prefs) return; prefs.hidden = true; if (expanded) expanded.hidden = true; if (compact) compact.hidden = false; } function expand() { if (!prefs) return; prefs.hidden = false; if (compact) compact.hidden = true; if (expanded) expanded.hidden = false; } function paint(state) { Array.prototype.forEach.call(toggles, function (t) { var key = t.getAttribute("data-cpcb-toggle"); if (key === "necessary") { t.setAttribute("aria-checked", "true"); return; } t.setAttribute("aria-checked", state[key] ? "true" : "false"); }); } function fromToggles() { var state = { analytics: false, advertisement: false, functionality: false }; Array.prototype.forEach.call(toggles, function (t) { var key = t.getAttribute("data-cpcb-toggle"); if (key in state) state[key] = t.getAttribute("aria-checked") === "true"; }); return state; } function commit(state, revoke) { writeCookie(COOKIE, JSON.stringify({ analytics: !!state.analytics, advertisement: !!state.advertisement, functionality: !!state.functionality, v: 1, ts: Date.now() }), days); apply(state, revoke); hide(); } Array.prototype.forEach.call(toggles, function (t) { if (t.disabled) return; t.addEventListener("click", function () { t.setAttribute("aria-checked", t.getAttribute("aria-checked") === "true" ? "false" : "true"); }); }); banner.addEventListener("click", function (e) { if (!e.target || !e.target.closest) return; var btn = e.target.closest("[data-cpcb-action]"); if (!btn) return; var action = btn.getAttribute("data-cpcb-action"); if (action === "customise") { expand(); return; } if (action === "accept") { commit({ analytics: true, advertisement: true, functionality: true }, false); return; } if (action === "reject") { commit({ analytics: false, advertisement: false, functionality: false }, true); return; } if (action === "save") { commit(fromToggles(), true); return; } }); var existing = stored(); if (existing) { apply(existing, false); paint(existing); collapse(); } else if (!inScope(banner)) { // Out of the consent-required region: implied consent, banner never shown. commitSilently(); } else { collapse(); show(); } function commitSilently() { var all = { analytics: true, advertisement: true, functionality: true }; writeCookie(COOKIE, JSON.stringify({ analytics: true, advertisement: true, functionality: true, v: 1, ts: Date.now(), implied: true }), days); apply(all, false); } // Public re-open hook for a footer "Cookie settings" link. window.cpCookieBanner = window.cpCookieBanner || {}; window.cpCookieBanner.open = function () { var s = stored() || { analytics: false, advertisement: false, functionality: false }; paint(s); expand(); show(); }; } function boot() { var banners = document.querySelectorAll("[data-cpcb]"); if (!banners.length) return; init(banners[0]); document.addEventListener("click", function (e) { if (!e.target || !e.target.closest) return; var trigger = e.target.closest("[data-cp-cookie-settings]"); if (!trigger) return; e.preventDefault(); if (window.cpCookieBanner && window.cpCookieBanner.open) window.cpCookieBanner.open(); }); } if (document.readyState === "loading") { document.addEventListener("DOMContentLoaded", boot); } else { boot(); } })();