The decision layer for web push
Know when to nudge.
Don't notify everyone. Nudge the right people.
Most people show up once and disappear. I spent a long time with a channel that can bring them back, and with tools that mostly knew how to shout.
Permission-based web push. Human control. Standards-based delivery on supported browsers and devices.
I could not buy the judgment layer.
Web push worked. The dashboards did not tell me who actually received a send, or why the list shrank most weeks. Teams wanted every campaign out. I wanted fewer, better ones, and I showed up to that meeting without numbers. EngageNudge is me building the tool I was looking for.
One engine. Shops, sites, and audiences.
The same Decision Engine adapts to your object, your audience, and your outcome. A product, a story, a listing, without blasting everyone who opted in.
Early access
Commerce
Product → shopper. Cart, restock, and price Nudges only when intent is real. Early access. We do not claim revenue lift yet.
Available
Publishers
Story → reader. Fewer notifications. More valuable return visits.
Early access
SaaS
Product → user. Re-engage people who already chose you, without treating every login the same.
Early access
Marketplaces
Listing → buyer. Saved searches that convert without alert fatigue.
Most sites never build their browser audience.
Most visitors browse and leave. A browser audience gives them a direct, permission-based way back, without requiring an email address or app install.
Newsletter
Owned and powerful, but it depends on an email address and inbox attention.
App
Deep engagement, but it depends on download, install, and continued app use.
Browser audience
A direct permission from your website to reach subscribers on supported browsers and devices.
Not every push is a Nudge.
A Nudge is a timely, permission-based browser notification designed to bring the right person back when something genuinely matters. EngageNudge delivers Nudges through standards-based web push.
Fewer interruptions, each one worth more.
Live demo
Feel a Nudge on this page.
A customer installs one script. Visitors see a prompt, allow notifications, and get a welcome Nudge. Start the demo to feel that path here.
What happens
- 01
Start the demo.
- 02
Allow notifications when the prompt appears.
- 03
A welcome Nudge arrives on this device.
Permission-based. No email. No app. You can turn notifications off in the browser at any time.
Welcome to EngageNudge
nowThis is what a first Nudge feels like. We only send when it matters.
Know when to nudge
Who, when, and whether to send at all.
EngageNudge evaluates the object, who it is relevant to, timing, recent exposure, and fatigue. It recommends targeted delivery, broad delivery, or do not send, and shows why.
Illustrative example
- Total browser audience
- 48,000
- Viewed this product
- 6,200
- Added to cart
- 1,140
- Dormant people protected
- −4,800
- High fatigue protected
- −2,100
- Already nudged this week
- −890
- Recommended audience
- 1,140
- Decision
- NUDGE TARGETED
Attention runs out before subscribers do.
Frequency caps treat every subscriber the same. EngageNudge also considers ignored streaks, recent clicks, topic cooldowns, quiet hours, and how relevant the current message is.
- Protect overexposed subscribers.
- Suppress weak-relevance sends.
- Respect quiet hours and cooldowns.
- Allow explicit urgent overrides.
- Keep your team in control.
How the intelligence works
Language models read the content. Your team decides.
EngageNudge uses language models to understand content topics, entities, geography, urgency, and shelf life. That understanding is combined with audience signals and transparent fatigue policies. Your team reviews, adjusts, or overrides the recommendation.
AI-assisted content understanding
Analyse each object once, not every subscriber individually.
Transparent decision policies
Relevance, exposure, and fatigue rules are versioned and explainable.
Human control
Recommendations can be reviewed, changed, or overridden.
From your object to measurable outcome.
- 01
Understand
Paste a URL, connect RSS, or trigger from the CMS.
- 02
Recommend
See relevance, audience, exclusions, and reasons.
- 03
Review
Edit the message, send a test, and approve.
- 04
Nudge
Deliver through standards-based web push.
- 05
Measure
Compare exposed subscribers with a holdout group.
- 06
Learn
Improve affinity, timing, and fatigue policies.
Know what the Nudge actually changed.
CTR shows who clicked. It does not show who would have visited anyway. EngageNudge can reserve a small holdout group and compare outcomes over the same measurement window.
Illustrative example
- Reached
- 32,941
- Clicks
- 1,714
- Qualified visits
- 1,283
- Holdout qualified-visit rate
- 0.7%
- Exposed qualified-visit rate
- 4.6%
- Estimated incremental qualified visits
- +1,065
Incrementality is an estimate. EngageNudge shows observations, assumptions, sample size, and measurement window rather than false precision.
Your audience stays yours.
EngageNudge is built around customer-specific VAPID keys, audience export, documented migration paths, and no advertising inserted into your notifications.
Know when to nudge.
Know when a notification is worth sending.
The useful question is who should not get this one.