Exit Intent Popup Audit: harrowgate-knits.example
Tool: Klaviyo popup. Period: 2026-10-01 to 2026-10-28. 32 result rows (16 page patterns, desktop and mobile). Offer: 10% off your first order for your email.
1. Settings review
| Setting | Yours | What the saved sources say | Note |
|---|---|---|---|
| Timing | Based on rules > When visitor is exiting the page | Klaviyo: on mobile this fires when a visitor "scrolls quickly upward on a mobile device" | Mobile shoppers scrolling back up a product page can trigger it |
| Close frequency | Show again after 0 days | Klaviyo: 0 days shows the form on "every page load after it is closed" | Flag F1 |
| Don't show again if submitted | Off | Klaviyo: with it on, the form will never appear again after someone fills it out | Flag F2 |
| Devices | All devices | ||
| Audience | Show to all audience members | Existing subscribers can see a first-order offer; a question below | |
| Pages | All pages | Includes /cart and /checkout; flag F4 | |
| Mobile layout | Full screen overlay | Google: "Don't obscure the entire page" | Flag F3 |
| Desktop layout | Centered box, about half the screen | Not the whole page; no flag |
Sources: https://help.klaviyo.com/hc/en-us/articles/4413544555547 and https://developers.google.com/search/docs/appearance/avoid-intrusive-interstitials
2. Results by row
Submit rate = submits / displays; close rate = closes / displays, per row, by script. Low data = under 500 displays (my working cut).
| Row | Page pattern | Device | Displays | Submits | Closes | Submit rate | Close rate | Low data |
|---|---|---|---|---|---|---|---|---|
| r1 | /products/* | desktop | 4,200 | 126 | 1,890 | 3.0% | 45.0% | no |
| r2 | /products/* | mobile | 6,900 | 97 | 4,140 | 1.4% | 60.0% | no |
| r3 | /collections/* | desktop | 2,600 | 91 | 1,170 | 3.5% | 45.0% | no |
| r4 | /collections/* | mobile | 4,100 | 57 | 2,460 | 1.4% | 60.0% | no |
| r5 | /cart | desktop | 900 | 38 | 410 | 4.2% | 45.6% | no |
| r6 | /cart | mobile | 1,300 | 19 | 820 | 1.5% | 63.1% | no |
| r7 | /blog/* | desktop | 1,500 | 30 | 690 | 2.0% | 46.0% | no |
| r8 | /blog/* | mobile | 2,300 | 16 | 1,450 | 0.7% | 63.0% | no |
| r9 | / | desktop | 3,100 | 109 | 1,360 | 3.5% | 43.9% | no |
| r10 | / | mobile | 5,200 | 73 | 3,180 | 1.4% | 61.2% | no |
| r11 | /pages/sizing | desktop | 700 | 12 | 330 | 1.7% | 47.1% | no |
| r12 | /pages/sizing | mobile | 1,100 | 6 | 690 | 0.5% | 62.7% | no |
| r13 | /pages/returns | desktop | 520 | 4 | 260 | 0.8% | 50.0% | no |
| r14 | /pages/returns | mobile | 880 | 3 | 560 | 0.3% | 63.6% | no |
| r15 | /account/login | desktop | 400 | 2 | 210 | 0.5% | 52.5% | yes |
| r16 | /account/login | mobile | 610 | 1 | 390 | 0.2% | 63.9% | no |
| r17 | /search | desktop | 1,200 | 41 | 520 | 3.4% | 43.3% | no |
| r18 | /search | mobile | 1,900 | 25 | 1,150 | 1.3% | 60.5% | no |
| r19 | /pages/about | desktop | 330 | 9 | 150 | 2.7% | 45.5% | yes |
| r20 | /pages/about | mobile | 450 | 4 | 280 | 0.9% | 62.2% | yes |
| r21 | /pages/shipping | desktop | 610 | 8 | 300 | 1.3% | 49.2% | no |
| r22 | /pages/shipping | mobile | 940 | 5 | 600 | 0.5% | 63.8% | no |
| r23 | /pages/gift-cards | desktop | 280 | 11 | 120 | 3.9% | 42.9% | yes |
| r24 | /pages/gift-cards | mobile | 390 | 6 | 240 | 1.5% | 61.5% | yes |
| r25 | /pages/contact | desktop | 250 | 3 | 120 | 1.2% | 48.0% | yes |
| r26 | /pages/contact | mobile | 360 | 2 | 230 | 0.6% | 63.9% | yes |
| r27 | /pages/stockists | desktop | 190 | 5 | 90 | 2.6% | 47.4% | yes |
| r28 | /pages/stockists | mobile | 240 | 2 | 150 | 0.8% | 62.5% | yes |
| r29 | /pages/care-guide | desktop | 420 | 14 | 190 | 3.3% | 45.2% | yes |
| r30 | /pages/care-guide | mobile | 610 | 7 | 380 | 1.1% | 62.3% | no |
| r31 | /checkout | desktop | 640 | 0 | 0 | 0.0% | 0.0% | no |
| r32 | /checkout | mobile | 980 | 0 | 0 | 0.0% | 0.0% | no |
Check by script: no row has submits plus closes above displays.
3. Totals by device
Totals are from summed counts, not averaged rates: device submit rate = sum of submits / sum of displays.
| Device | Displays | Submits | Closes | Submit rate | Close rate |
|---|---|---|---|---|---|
| desktop | 17,840 | 503 | 7,810 | 2.8% | 43.8% |
| mobile | 28,260 | 323 | 16,720 | 1.1% | 59.2% |
| all | 46,100 | 826 | 24,530 | 1.8% | 53.2% |
This rests on one period of 28 days and one popup version, so the desktop and mobile gap is a pattern to test, not proof that the mobile layout causes it. Mobile also differs in how exit intent fires, per the Timing row above.
4. Flags
| Flag | What | Applies to |
|---|---|---|
| F1 | Close frequency 0 days: the popup returns on every page load after a close | setting |
| F2 | Don't show again if submitted is off: people who already submitted can see it again | setting |
| F3 | Mobile full screen overlay covers the whole page | setting |
| F4 | Popup shows on /cart and /checkout (r5, r6, r31, r32): should it show there at all? | rows r5, r6, r31, r32 |
| F5 | Displays with 0 submits and 0 closes (r31, r32): is the close or submit tracked on these pages? | rows r31, r32 |
Counted by script: 5 flags; 4 distinct result rows flagged (r5, r6, r31, r32); F1 to F3 are settings that affect every row.
| Count | Value |
|---|---|
| Flags | 5 |
| Setting flags | 3 |
| Distinct rows flagged | 4 |
Google's interstitials page says intrusive dialogs "may lead to poor search performance". I treat F3 as a risk to fix, not a measured ranking effect.
5. Proposed settings (held)
Setting names are Klaviyo's, from the saved Targeting and behavior article. Nothing is changed until you approve each item.
| ID | Change | Fixes | Status |
|---|---|---|---|
| P1 | If visitor closes form, show again after: a longer gap than 0 days (you choose the number; the sources give no best value) | F1 | HELD |
| P2 | Turn on Don't show again if form was submitted or if go to URL button was clicked | F2 | HELD |
| P3 | Mobile: replace the full screen overlay with a smaller box or banner that leaves the page visible | F3 | HELD |
| P4 | Exclude /cart and /checkout from page targeting, if you agree with F4 | F4 | HELD |
| P5 | Check tracking on /checkout before any change there | F5 | HELD |
6. Test plan
- Hypothesis: a smaller mobile layout (P3) plus a longer close gap (P1) may raise mobile submits per display without raising closes per display. This is a hypothesis to test, not an expected result.
- Compare: the mobile totals in the Totals by device table as the baseline, against the same 28 day length after the change.
- Change one thing at a time if you want to know which one mattered; changing P1 and P3 together tests the pair only.
7. Questions for you
- Should existing subscribers see a first-order offer? Audience is Show to all audience members.
- How many days should pass before the popup returns after a close (P1)?
- Is the /checkout row count real, or is the popup blocked there by your checkout? Displays show but no submits or closes.
- Is the 10% offer also shown anywhere else on the site? I only know the popup text.
Needs your decision
- Approve changing the close frequency (P1, HELD).
- Approve turning on Don't show again if submitted (P2, HELD).
- Approve changing the mobile layout (P3, HELD).
- Approve excluding /cart and /checkout (P4, HELD).
- Approve the /checkout tracking check (P5, HELD).