Popup and Modal Optimizer with Holdout Testing
Popups that earn the interruption: device-specific triggers, suppression and caps, accessible dismiss, SEO and consent checks, holdout test.
Before & after
Same prompt, same agent — the only difference is the skill.
Exit-intent popups are effective at capturing users who are about to leave. The key is making the offer compelling without being too pushy. You'd want to test different copy variations, timing, and design approaches to see what resonates. Some people respond better to a discount, others to FOMO messaging or a free trial extension. You'll also need to decide on the modal size and where to position it on screen. What's your current trial signup rate, and what are you trying to improve?
EXIT-INTENT POPUP SPEC
TRIGGER
Activates when mouse moves toward back button or tab close (desktop only)
Delay: 2-second pause before showing (avoids accidental triggers)
Frequency: max 1 per session, suppress for 30 days after signup or dismiss
COPY AND MESSAGING
Headline: "Wait—try Agentvolt free for 7 days"
Subheadline: "No credit card needed. Full Pro access."
Body: "Most teams see results in their first 2 days. Join 1000+ companies building AI agents faster."
CTA BUTTONS
Primary: "Start Free Trial" (blue, bold)
Secondary: "Show me examples" (ghost, links to case studies)
Dismiss: "Not now" (subtle close)
DESIGN BRIEF
Size: 450px wide, 300px tall (centered on screen)
Background: white card with subtle shadow
Border: thin 1px #EEEEEE
Font: dark gray #1A1A1A for body
Spacing: 32px padding inside
CONVERSION TRACKING
• Click rate on "Start Free Trial" (target: 12%)
• Completion rate for trial (target: 35%)
• Cost per acquisition from popup (compare to other channels)
A/B TEST VARIANTS
Variant 1: Discount frame ("7 days 100% off")
Variant 2: Social proof frame ("1000+ teams trust Agentvolt")
Variant 3: Risk-removal frame ("No credit card required")
Run each for 2 weeks, measure signup rate and trial completionAbout this skill
Stop annoying buyers and find out whether your popup adds sales or just takes credit for them. The rules come first, then the copy, then a holdout test.
What it does
- Sets triggers, targeting, exclusions, suppression, and caps separately for desktop and mobile
- Writes a one-line value exchange, polite decline, and instant success state
- Checks dialog accessibility, mobile search interstitial risk, consent, and discount cost
- Plans a holdout test on revenue per visitor instead of trusting popup-tool attribution
What you get
A popup spec table, copy, a checks list, and a holdout test plan.
Best for
Ecommerce and content sites using email capture, discount, or promo overlays.
SKILL.md preview
---
name: popup-modal-optimizer
description: Plan, write, and fix on-site popups, modals, slide-ins, and banners - trigger and targeting rules per device, frequency caps, suppression, copy, accessible dismiss behavior, search-ranking and consent risks, and a holdout test that shows whether the popup adds conversions or just claims them. Use when the user says "set up an exit-intent popup", "our email popup converts badly", "should we show a discount popup?", "is this popup hurting SEO?", "popup for our sale", "spin to win wheel", or "review my popup settings". Covers overlays on an existing site; the page itself, signup forms, and in-app paywalls are separate jobs.
---
… (sign up to view the full skill)More marketing & growth skills
View all Marketing & Growth skills →Paid Ads Campaign Planning and ROAS/CPA Diagnosis
Plans paid campaigns on Google, Meta, LinkedIn and X from break-even math, writes ads, and finds the metric behind a ROAS or CPA drop.
Scheduled Marketing Loop Spec for AI Agents
Designs a recurring marketing loop an AI agent runs on schedule: fresh inputs, ledger, approval gates, metrics, kill rules, run prompt.
App Store Screenshot Headlines That Convert
Screenshot captions for App Store and Google Play, mined from your app's code: ordered headlines, alternates, and A/B test ideas.
Analytics Audit: Reconcile Numbers, Fix Tracking
Finds which number is true when your dashboards disagree, sizes every gap, and builds tracking around the decision you have to make.