Website notifications only help if they earn attention without stealing it, so measuring them is as much about restraint as it is about clicks.
Adding notifications to a website is easy. Knowing whether they're helping is harder. A notification can get plenty of clicks and still hurt your site if it irritates the people who didn't click. This article explains how to read the statistics in EasySocialCred and how to balance results against the visitor experience.
How many website notifications are too many?
Every extra script and popup competes for a visitor's attention and their device's resources. The HTTP Archive Web Almanac 2024 found that the median mobile page already weighs about 2.3 MB and that 92% of pages load third-party resources. A visitor on a phone, which, according to the Pew Research Center, 91% of U.S. adults own, has less screen space to spare than you think. As a rule of thumb, show one notification at a time, with a gap of several seconds or more.
The three numbers to watch
Impressions. How many times a notification was displayed. This tells you how much exposure it gets. If it's lower than expected, your display rules may be too narrow, or the delay may be longer than most visits last.
Hovers. How many times visitors moved their mouse over it. On desktop this is a rough signal of interest: people paused to look. Phones don't have hovers, so this number is mostly a desktop measure.
Clicks. How many times visitors clicked through. This is usually the main outcome for informational, coupon and review notifications.
From these you can work out the click-through rate: clicks divided by impressions. It's the fairest way to compare notifications that are shown different amounts.
For collector-type notifications (email, request, feedback), the important number is how many submissions you received, which you can see and export from the dashboard.
What counts as "good"?
We won't give you a target number. Click rates vary hugely by site, audience, page, offer and design, and any industry "average" you read online is unlikely to apply to your situation. The useful comparison is against yourself:
- Is this notification doing better or worse than last month?
- Is version A doing better than version B?
- Is this notification doing better on product pages than on the homepage?
Test one change at a time
When you want to improve a notification, change one thing, then wait long enough to see a difference. Things worth testing:
- The message. Specific beats vague almost every time.
- The delay. Showing it later can mean fewer impressions but a higher click rate, because only engaged visitors see it.
- The page. The same message can work well on one page and poorly on another.
- The position. Bottom-left vs. bottom-right vs. top bar.
- The image. A product photo vs. an icon vs. no image.
If you change the message, the delay and the position at the same time and results improve, you won't know which change did it.
A simple way to test is to duplicate a notification, change one setting in the copy, and run both on similar pages or alternate them over equal time periods. Then compare click-through rates.
Give it enough time
A day or two of data is usually too little to draw conclusions, especially on smaller sites. Wait until each version has a reasonable number of impressions before deciding. If your traffic is low, that might mean a few weeks. Weekday and weekend visitors also behave differently, so try to compare whole weeks.
Watch for signs of annoyance
The statistics show what the people who interacted did. They don't directly show the people who were annoyed and left. Watch for these indirect signs:
- Bounce rate or time on page gets worse in your website analytics after adding notifications.
- Conversions on the page drop, even though the notification gets clicks.
- Complaints or comments from visitors mentioning popups.
- High impressions with almost no clicks or hovers. People are seeing it and ignoring it, which often means it's in the way.
If you see these, scale back before trying anything else: fewer notifications, longer delays, fewer pages.
Good habits that protect the experience
- One at a time. Don't stack several notifications on screen.
- Respect mobile. Small screens make overlays feel bigger. Consider a longer delay, a smaller position or turning certain notifications off on mobile.
- Stay away from forms and buttons. Never cover a checkout button, a form field or navigation.
- Let people close it. Make sure the close option is easy to find.
- Limit repeats. If someone closed a notification, they shouldn't see it again on every page.
- Keep it true. Honest messages hold up over time; exaggerated ones erode trust. Our honesty guide explains why.
Using email reports
EasySocialCred can send periodic email reports summarizing your campaign statistics, so you don't have to remember to check. Use them as a prompt for a short review: what changed, what's working, what to try next.
Exporting your data
Statistics and collected data can be exported (for example as CSV), which is handy if you want to combine them with your own analytics or keep records.
A monthly review routine
- List your active notifications and their click-through rates.
- Turn off anything that hasn't earned its place.
- Pick one notification to improve and decide on a single change.
- Check your site analytics for any signs of annoyance.
- Note what you changed and when.
This takes less than half an hour a month and keeps your notifications working for you instead of against you.
If you're just getting started, our setup guide walks through creating your first campaign. Questions about your stats? Contact us. For broader marketing and automation help, see justinzwick.com.
Key takeaways
- Track impressions, hovers, clicks and conversions together
- Compare periods with and without the notification
- Change one setting at a time
- Limit frequency, delay first display and respect closes
- Watch for signs of annoyance, such as quick exits
Frequently asked questions
What's a good click rate for a notification?
It varies widely by notification type, audience and placement. Compare against your own past results instead of chasing a universal number.
How long should a notification stay on screen?
Long enough to read comfortably, usually a few seconds. Very long display times make notifications feel intrusive.
Should I show notifications on mobile?
Yes, but check them carefully. Make sure they don't cover buttons or navigation on small screens.
How do I test whether notifications help?
Run a period with the notification and a comparable period without it, then compare your main goal, such as signups or sales.
Can I stop notifications from showing to the same visitor repeatedly?
Yes. Use the display frequency settings so returning visitors aren't shown the same message over and over.
