Tuesday’s revenue row shows $147 from Gumroad. Tuesday’s visitor count looks flat. You are not haunted, buyers often convert on a different day than their last tagged visit, or they checkout on Gumroad’s domain without a fresh pageview on yours.
Digital product analytics guides.
Email lag pattern
Reader clicks Monday, thinks, buys Wednesday from Gmail search. Webhook fires Wednesday; last tagged session was Monday, still attributable if UTMs persisted in session or you use newsletter UTMs on the buy link.
Buy link directly on Gumroad profile
Some traffic never touches your domain. Revenue exists without your visitor counter moving. Options: Gumroad profile as source=gumroad-profile via redirect through your domain, or accept blind spot and track profile clicks separately.
Mobile app handoff
Instagram in-app browser → Gumroad app → sale. Pageview may not fire on your site. Use redirect bridge page on your domain with 302 to Gumroad with UTMs.
Gift purchases
Buyer ≠ visitor intent path. Gift sales skew ratio; tag product_metadata=gift if platform exposes it.
Comparing daily revenue to daily visitors
Use 7d rolling overlays, not single day equality. Stripe webhook ordering has parallel lessons.
Pin the send, measure the week
Newsletter send pin captures traffic spike day 0; extend revenue pin +7 days for digital products with slow buys.
Worked timeline
Mon 400 visits (email), Tue 200, Wed 50 visits but 18 sales webhook, ratio on Wed alone looks impossible; week pin shows healthy conversion from Mon traffic.
Honest reporting
Tell stakeholders: “Revenue attributed to campaign X over 14d” not “same-day conversion 0.2%.”
Hub
Lag is normal. Pin windows wide enough for how humans actually buy templates and courses.
Attribution models for slow digital buys
| Model | When to use |
|---|---|
| First-touch UTM in 14d | Newsletter-heavy, trust builds slow |
| Last-touch click in 7d | Short promo windows |
| Same-session only | Strict, undercounts email |
Pick one for creator products and write it in your metrics doc. Changing models retroactively rewrites which “campaign won.”
Gumroad Discover and marketplace traffic
Sales from Gumroad’s own discovery may never touch your domain. Revenue exists; your visitor graph flatlines. Tag what you can with redirect bridges; otherwise bucket source=gumroad-marketplace manually from platform reports and exclude from site ratio numerators or accept two parallel truths.
Bookmark and direct return buyers
Second purchase from a customer who bookmarked checkout bypasses fresh UTMs. They are expansion or repeat, not acquisition, filter is_returning_customer if the platform exposes it, or exclude emails with prior sales from launch ratio math.
Timezone boundaries
Webhook at 11pm UTC Monday vs your analytics “day” in US Eastern makes “Tuesday sales, Monday traffic” worse. Align reporting timezone to Stripe/Gumroad dashboard or use date-only buckets in UTC everywhere.
Stripe lessons on the same chart
Pageview after webhook is about seconds of ordering; Gumroad lag is often days of consideration. Same fix family: metadata and tagged links, wide pins, distrust same-day equality.
Newsletter-specific timeline
Send Tuesday 9am: traffic spike Tuesday, sales Wed–Sun. Pin newsletter-2026-03-w12 for visitors and revenue through following Tuesday. Compare to pin date range for newsletter week if you use pulse pins.
Lemon Squeezy receipt emails
Buyers click receipt links days later for downloads, not new acquisition. Do not count receipt clicks as campaign traffic unless you tagged them medium=transactional and filter out.
Worked attribution sentence
“Campaign spring-templates-2026 drove $840 gross over 14d post-send, attributed by first-touch UTM on buy link, 62% of sales on non-send days.”
That sentence is what stakeholders need, not “Wednesday conversion 0%.”
Ratio denominator on lag weeks
When sales lag, do not compute ratio on the sale day alone. Use pin-wide visitors ÷ pin-wide sales. Daily charts are for monitoring spikes, not strategy.
Product Hunt hangover
PH visitors day 0; sales day 3–10 via email and direct. Extend pin per PH baseline guide. Otherwise PH looks like traffic without revenue and email looks like revenue without traffic, both wrong.
Two products, same lag
Template impulse buy same day; course purchase after three videos. Split products per two products one domain before interpreting lag, blended lag hides course consideration time.
Action checklist
- Tag every buy link in email and social.
- Pin send + 14d revenue.
- Report attributed revenue, not same-day CR.
- Bucket marketplace sales separately.
- Reconcile webhook count to platform dashboard weekly.
Hub links
Traffic-to-sales guide · Digital product guides
Affiliate and collaborator lag
Affiliate links may record click day 1 and conversion day 9. Affiliate dashboard and your analytics will disagree on “conversion day.” Pick affiliate’s definition for payouts; pick your attribution window for creative decisions.
Podcast show notes
Listeners buy days after episode airs without clicking tracked link, direct or branded search rises instead. Show notes UTMs help but undercount; ratio by utm_source=podcast is lower bound, not truth.
YouTube description links
Viewers search product name instead of clicking description. Same pattern as podcast, wide revenue pin, modest tagged traffic.
Cart abandonment emails
Platform abandonment emails recover sales without fresh site pageview. Revenue may attribute to email medium on Gumroad/Lemon while your site shows quiet day, expected.
Multi-device research
Research on desktop, buy on phone Gmail app, Monday pageview desktop, Wednesday sale mobile. Session join fails; first-touch UTM on buy link in email survives if link was tagged at send time.
Cohort of “warm list” buyers
Second purchase from same email should not enter acquisition ratio. Maintain simple “first sale date per email hash” exclusion in spreadsheet if analytics lacks it.
Communicating lag to partners
Tell collaborators: “We measure 14d attributed revenue, not same-day CR.” Prevents false arguments when their screenshot shows Tuesday traffic and your webhook shows Thursday sales from their link.
Dashboard visualization
Plot 7d rolling sales and 7d rolling visitors on same chart, lines converge visually when lag is healthy. Diverging lines late in pin suggest broken tracking, not lag.
Webhook retries and duplicate sales
Dedupe order ids before counting sales per day; duplicate webhook should not double-count Wednesday “spike.”
Seasonal gift lag
Holiday gifts purchase early, redeem late, unusual for digital but happens with coupon gifts. Exclude gift SKUs from tight lag analysis if metadata allows.
Closing principle
Digital products sell on trust timelines, not session timelines. Analytics should respect trust timelines or founders optimize for phantom same-day conversion that does not exist.
Quick reference card
Visitors: pin start → +14d. Sales: same window, webhook deduped. Ratio: pin-wide only. Same-day: ignore for strategy. Pin that card above your desk until the habit sticks.
Stripe SaaS on the same domain
If you also sell subscriptions via Stripe, lag patterns differ, trials convert weeks out. Read subscription vs one-time attribution for event filters; Gumroad lag is usually shorter but email-driven in both cases.
Saved pins for repeat launches
Reuse the same 14d visitor + revenue pin template every newsletter send, saved funnel UTMs name the campaign; the pin width stays constant so lag math stays comparable quarter to quarter.