The short answer
Measure reader engagement with three kinds of signal, per article: attention (engagement time and scroll depth), loyalty (return visits and newsletter sign-ups) and the reader's own verdict (one question at the end, such as "Was this article useful?"). Analytics show what readers did; only asking shows whether the article helped. Compare articles of the same type, and read a long engagement time next to the answers before you call it good.
"Engagement" has become the number every editor is asked about, and every tool defines it differently. Google Analytics counts a session as engaged once it lasts longer than ten seconds. A reader who finds the answer in the first paragraph and leaves looks disengaged. A reader stuck on a confusing guide for five minutes looks deeply engaged.
This guide sets out what the common reader engagement metrics measure, where each one misleads, and how to combine them so they tell you which articles to keep, fix or drop.
Reader engagement metrics compared
| Metric | What it measures | Where it misleads |
|---|---|---|
| Page views | How many times the article was opened | Reach, not reading: a headline that over-promises gets views and disappoints |
| Engaged sessions and engagement rate (GA4) | Sessions longer than 10 seconds, with a key event, or with 2 or more page views | Eleven seconds on an article counts the same as reading it to the end |
| Engagement time (GA4) | Time the page is in focus in the browser | Long is not always good: readers also spend time on pages they find confusing |
| Scroll depth | How far down the article readers get | An answer near the top lets satisfied readers leave early |
| Return visits and sign-ups | Whether readers come back or subscribe | Credits the publication as a whole, rarely a single article |
| Comments and shares | Readers who cared enough to respond in public | A small, loud group, rarely typical of everyone who read |
| Reader answers | Whether the article helped, in the reader's own words | Only readers who reach the question answer, so read counts, not just percentages |
What Google Analytics counts as engagement
In GA4, a session is engaged if it lasts longer than 10 seconds, has a key event, or has 2 or more page or screen views. The engagement rate is the share of sessions that are engaged, and the bounce rate is the share that are not.
Engagement time is "the amount of time someone spends with your web page in focus". Google says that for websites it reflects active interaction with the page. It is a fair measure of attention, but it cannot tell a reader who is absorbed from one who is lost.
For an article, the 10-second rule is a low bar: almost anyone who reads the first paragraph passes it. Use engagement rate to spot articles that people leave at once, and engagement time to compare articles of similar length, not as a score of quality.
Why attention alone cannot tell you an article worked
Readers spend most of their time near the top of a page. In a 2018 eye-tracking study of 120 people, Nielsen Norman Group found that 57% of page-viewing time was spent above the fold and 74% within the first two screenfuls.
That cuts both ways for articles. A news story that gives its answer in the first lines may leave readers satisfied after one screen, with a short engagement time and a shallow scroll. A long how-to guide may hold readers for minutes because a step is unclear. In both cases the attention numbers point the wrong way, and only the reader can say which it was.
Add the reader's verdict: one question at the end
Ask one short question at the end of every article, such as "Was this article useful?" with Yes and No, and an optional "What was missing?". The share of Yes answers gives each article a second number to set next to its engagement time.
| Most readers say Yes | Many readers say No | |
|---|---|---|
| Long engagement time | Readers stayed and it paid off: keep it and link to it | Readers are working hard and failing: read the comments and fix the unclear part |
| Short engagement time | Quick answer, job done: leave it alone | Readers gave up early: check the headline promises what the article delivers |
Read the number of answers as well as the percentage. Ten answers, eight of them Yes, is a hint; two hundred is a pattern.
Compare like with like
A breaking news story, a long explainer, a review and a step-by-step guide behave differently, so a site-wide average hides more than it shows. Group articles by type or section and compare each article with its neighbours.
Pick one time window, such as the last eight weeks, and keep it. When you rewrite an article, compare the weeks before and after the change, and keep the question the same so the answers stay comparable.
Example: two guides with the same engagement time
Example
Say you run a tech publication. Two how-to guides each show an average engagement time of about three minutes and similar scroll depth. On the analytics alone, they look equally good.
You add "Was this guide useful?" at the end of both. After four weeks, the first guide has 140 answers, 82% Yes. The second has 95 answers, 41% Yes, and the comments under its No answers keep naming the same step: a command that has changed since the guide was written.
The same three minutes meant two different things. You update the command in the second guide and watch whether its Yes share rises over the next four weeks.
Where Asklet fits
Asklet adds the reader's verdict to the numbers you already have. One snippet puts a short question at the end of every article, and you see the answers and comments per article, next to each other. Readers answer anonymously without signing in, and the widget sets no cookies.
Keep your analytics for reach and attention. Use the answers to tell which of your long-engagement articles are good and which are hard work.
Your next steps
-
Pick one attention metric and one loyalty metric
For example engagement time per article and return visits per section, and keep them for a full quarter.
-
Group articles by type before you compare
News, explainers, guides and reviews each get their own comparison.
-
Ask one question at the end of every article
"Was this article useful?" with Yes and No, plus an optional "What was missing?".
-
Read time and answers side by side
Long time with many No answers means readers are struggling; short time with many Yes answers means a quick win.
-
Re-check after every rewrite
Same question, same time window, before and after the change.
Try these steps on your own site. Asklet adds the questions to every page with one snippet. Compare the plans.
Start freeCommon questions
What is reader engagement? How readers interact with an article beyond opening it: how long they pay attention, how far they read, whether they come back, and whether the article helped them. Analytics measure the first three; asking readers measures the last.
What counts as an engaged session in Google Analytics 4? A session that lasts longer than 10 seconds, has a key event, or has 2 or more page or screen views. The engagement rate is the share of sessions that meet at least one of those.
Is time on page a good measure of engagement? It measures attention, not success. Readers spend long on great articles and on confusing ones, so read engagement time next to a direct question such as "Was this article useful?" before you judge an article by it.
What is a good engagement rate for a blog? There is no benchmark that fits every site, and GA4 counts any session over 10 seconds as engaged, so most article readers pass. Compare each article with others of the same type on your own site.