Medium Analytics Explained: Presentations, Views, Reads and Earnings

Medium analytics answers several different questions: was your story shown, did someone open it, did they stay long enough to count as a read, did they follow you, and did eligible member activity earn money? Start with the question that matters to your next decision. A presentation is an exposure inside Medium; a view is a landing on the story; a read requires at least 30 seconds. None of those alone proves a completed read or predicts earnings.

My view is that writers should diagnose the earliest weak stage before changing the whole article. If few people saw it, rewriting the ending is unlikely to resolve the immediate problem. If people opened it and left, chasing more exposure leaves the reading experience unexamined. Narrareach can keep available Medium results beside your next draft, while Medium remains the authority for its native metrics and payouts.

Read your Medium stats as a set of decisions

The useful progression is exposure, opening, reading, audience relationship and eligible earnings. It is a diagnostic framework, rather than a claim that every reader follows one trackable path.

Metric	Question it helps answer	What it does not establish
Presentations	Did Medium show the story on tracked surfaces?	Total exposure across the internet
Views	Did readers land on the story?	Unique people or finished reading
Feed clickthrough rate	Did feed exposure lead to opening?	Conversion from every traffic source
Reads and read ratio	Did visits reach the reading threshold?	Completion or satisfaction
Followers and subscribers	Did the story prompt an ongoing connection?	Future return visits or paid membership
Earnings	What did eligible activity accrue?	A fixed payment for each raw view

An article intended to introduce your work may deserve attention because readers follow you. A technical reference may attract useful search traffic slowly. A paywalled essay needs a separate earnings review. Decide which job you gave the story before deciding whether its results were disappointing.

What presentations mean on Medium

Presentations count tracked occasions when Medium shows a story inside its own ecosystem. They cover more than the home feed: search, topic pages, publications, profiles and other tracked surfaces can contribute. Google search impressions and exposure on outside social platforms are excluded. Medium provides this measurement for stories published from January 1, 2025 onward. Its Stats FAQ explains the presentation boundaries.

Consequently, an older story without presentation data is not necessarily receiving no exposure. Nor does low presentation volume mean the story has no audience outside Medium. Treat missing history as unavailable evidence, and inspect the views and traffic sources you do have.

When presentations are low, examine where the story could reach suitable readers. Does it belong in a relevant publication? Do its topics accurately describe the work? Is the subject intelligible to a person encountering the title without knowing you? Those questions address discoverability and fit. They do not establish that an algorithm has penalized your account.

A publication can introduce work to an existing readership, but placement is not a guaranteed volume of presentations. Choose one whose readers need the article, rather than treating the largest publication as automatically the best match. A story can reach fewer people and create more relevant conversations.

Why views can be higher than presentations

Views record landings on the story, whereas presentations describe tracked exposure within Medium. Someone following an outside link can create a view without a corresponding Medium presentation. Dividing all views by all presentations therefore produces a number with incompatible scopes.

A September 17, 2026 snapshot of Medium data held in Narrareach illustrates the practical consequence. Among 311 distinct stories published from January 1, 2025 onward, drawn from 16 connected accounts whose data had refreshed within the preceding 48 hours, 28 had more lifetime views than lifetime presentations. That is 9.0% of this particular set: 28 divided by 311. These are lifetime story totals observed in September, not traffic generated during those 48 hours. Different story ages and a small, selected account group make this unsuitable as a Medium benchmark. The counts also cannot identify which outside source caused the difference.

The point is useful even without a benchmark: higher views than presentations can be compatible with the measurement definitions. You should inspect traffic sources before declaring the dashboard broken or assuming your headline converted more than 100% of impressions.

For a constructed example, imagine a story with 600 Medium presentations and 900 total views, including visits from a newsletter and Google. The calculation 900 ÷ 600 gives 150%. It does not mean a 150% clickthrough rate. You have compared all entrances with only one category of exposure.

Keep outside distribution visible in your review. A search-led story may continue bringing readers after feed activity quiets down. A story with considerable internal exposure may have a different pattern. Neither should be assessed using the other's exposure denominator.

Feed clickthrough rate needs its own denominator

Medium's feed clickthrough rate uses feed presentations and the openings attributed to them. It is not total views divided by every presentation. The rate appears only once Medium has enough data; an absent percentage should not be entered into your report as zero.

Suppose a constructed story receives 2,000 feed presentations and 80 openings from that feed exposure. Its feed clickthrough is 4%. Another 120 visits arrive elsewhere. Using all 200 visits would incorrectly produce 10%. Preserve the reported feed rate when comparing packaging, and keep total views in a separate column.

A modest rate can also accompany broader exposure. Readers beyond your established niche may be less inclined to click. Before revising a title, compare the story with your own work on similar subjects, at similar ages, with reasonably comparable audiences. Medium does not publish one universal good feed clickthrough or read-ratio target.

A better title makes the specific value easier to recognize. “A Better Writing Routine” offers little information. “How to Protect Two Writing Sessions When Your Week Keeps Changing” tells a reader what problem the article addresses. Use that specificity only if the piece delivers it. A compelling promise followed by an unrelated opening can improve clicking while weakening the next stage.

A Medium read is not a completed article

Medium currently counts a read after at least 30 seconds. The threshold helps distinguish brief entrances from more sustained attention, but does not prove that someone reached the ending. The detailed story stats guide distinguishes the reach funnel from audience and earnings outcomes.

Consider a constructed comparison: Story A has 300 views and 150 reads; Story B has 1,200 views and 240 reads. Their read ratios are 50% and 20%. A retains a larger share of its entrances through the threshold, while B records more reads overall. If your objective is to improve the opening, A warrants examination. If your objective is total reading reach, B has produced more threshold reads. Neither percentage tells you which story earned more.

Inspect both the count and the ratio. With five views, one additional read changes the ratio by 20 percentage points. With 500 views, one additional read changes it by 0.2 points. That difference should affect your confidence, even when the dashboard presents both percentages with equal visual weight.

When views are reasonably substantial but the ratio is weak relative to comparable work, read the title and opening together. Does the introduction answer the question that earned the click? Does the reader have to pass through several paragraphs of background before finding the promised explanation? Is an unfamiliar term left unexplained? Correct the specific obstacle you can identify, rather than declaring all long articles ineffective.

Shortening everything to improve a ratio can remove the depth that made a story valuable. Keep necessary explanation, worked examples and exceptions. Cut repetition and delayed answers. The reader's task should determine the length.

Claps, followers and subscribers answer different questions

An enthusiastic reaction can coexist with little sustained reading. It is a signal of response, not a substitute for the reading measurement. Distinguish the number of people who clapped from a total clap count before comparing engagement across stories.

Followers and subscribers indicate an ongoing connection attributed to the story in Medium's impact reporting. Those actions are different from someone becoming a paying Medium member. Do not use a writer's subscriber total as a count of paying customers, or import Substack's free-versus-paid newsletter model into the Medium dashboard.

For an acquisition-focused story, record the audience outcomes beside views and reads. A piece with fewer reads may still encourage more people to follow your work. That is a reason to examine its relevance and invitation, not proof that each new follower will return.

Make the next reading opportunity clear. If a guide solves an introductory problem, link a genuinely relevant follow-up rather than adding an unrelated list of popular stories. Someone who found one explanation useful should be able to recognize what they can learn next. Assess later work separately; the initial follow is an intention, not a retention measurement.

Why reads do not convert into a predictable dollar amount

Partner Program earnings depend on eligible member activity on eligible paywalled stories. Raw views and total reads include activity that does not have identical earnings value. Medium's current earnings explanation describes member reading and listening time, engagement, bonuses and a member read-ratio adjustment.

The member read ratio uses member reads divided by member views. It differs from the total read ratio when members and non-members behave differently. For example, 120 total reads from 300 total views give 40%. If the member subset contains 60 reads from 100 member views, the member ratio is 60%. Both can be correct. You cannot derive the member figure from the total percentage alone.

Avoid adopting another writer's dollars-per-read result as your expected rate. Similar totals can contain different member activity, sources and eligible periods. A story's earnings per 1,000 views can describe its observed performance, but it remains an outcome calculation rather than a published tariff.

Before investigating a revenue decline, confirm that you are comparing the same period, that the story was eligible during it, and that the report has had time to settle. Medium says earnings update daily on a UTC day and can take up to 48 hours to finalize. A live morning traffic total and an unfinished earnings day do not form a fair comparison.

Separate a reporting change from a writing problem

Take a screenshot or record a dated observation if a recent number appears inconsistent. Then compare the same story and period after the normal update cycle. Medium's FAQ explains that recent real-time counts can change as deduplication and other checks turn them into stable counts.

That explanation does not prove every blank chart is normal. If a discrepancy persists, record the story URL, the metric, the selected dates and when you checked it, then contact support. A reproducible description is more useful than a claim that the platform erased your audience.

Do not rewrite a story in response to a missing percentage or one anomalous hour. You risk changing useful work before knowing whether the observation reflects reader behavior. Likewise, do not dismiss a sustained decline simply because reporting can fluctuate. Confirm the period and scope, then investigate the stage that changed.

Turn the report into your next draft

Narrareach's Medium stats workflow brings available story presentations, views, reads and earnings into the publishing workspace, with the connected account's audience context. Use it to keep the evidence close to the next action. Open Medium's native report whenever you need a detailed traffic breakdown, distribution status or definitive payout information; a synced overview is not a replacement for those surfaces.

Choose three to five comparable stories and record their publication dates, observation date, purpose, views, reads and reported feed rate where available. Identify one question the comparison raises. Perhaps titles on a specific subject attract readers, but their introductions take too long to deliver the promised answer.

Write a concrete hypothesis before drafting: “The next guide will put the decision framework in its first 150 words, then explain the exceptions.” Prepare the piece, review whether the title and opening agree, and check the destination, formatting, links and any paywall settings. Narrareach's Medium publishing integration keeps the reviewed article in the same distribution workflow as its supported adaptations and schedule.

After publishing, compare the new story at a similar age with those earlier pieces. Record the differences in audience and distribution, too. This is an editorial experiment, not a controlled proof that one paragraph caused the outcome. Continue the change when repeated observations and reader feedback support it; reconsider it when they do not.

For your next review, take one story and write two sentences beneath its stats: what the figures actually establish, and what remains unknown. Then choose one change that addresses the established weakness. That gives the dashboard a useful job without asking it to judge the whole value of your writing.