How to estimate a newsletter's subscriber count
TL;DR
You cannot get an exact subscriber count from the outside, but you can land inside a believable range. Three methods do most of the work: back-solving from a published sponsorship rate, dividing opens by open rate, and reading platform and ESP tells. The catch nobody mentions is Apple Mail Privacy Protection, which inflates reported open rates 40 to 44 percent on consumer lists, so the open-rate method overstates reach unless you correct for it. The sponsorship back-solve is the one I trust, because it never touches opens.
The honest version of how to estimate a newsletter's subscriber count starts with admitting you will never get the exact figure. No outside method does. What you can do is build a range you can defend in a meeting, with the assumptions written down next to it. We size competitor lists this way constantly when we audit a rival for a client, and the number we hand over is always a band, never a point. A point estimate is a guess wearing decimal places.
The short answer, and why "you can't know" is wrong
Search this question and most of the results say some version of "you can't, it's private." That answer is lazy. Subscriber count is not visible, but it leaves fingerprints all over the public web: sponsorship rate cards, media kits, like counts, the sending platform, the words "join 24,000 readers" sitting in the signup box. Each one narrows the range. Stack three of them and the band gets tight enough to act on.
Set expectations before you start. A good outside estimate for a mid-size newsletter lands within plus or minus 20 to 30 percent of the truth. That is enough to know whether a competitor has 8,000 subscribers or 80,000, which is the decision that matters. It is not enough to know whether they have 24,000 or 27,000, and you should stop pretending it is. State the range, state the confidence, move on.
Method 1: back-solve the subscriber count from sponsorship rates
This is the method I reach for first. Any newsletter that sells sponsor slots has to publish a price somewhere, on a media kit, a Passionfroot page, or a "sponsor us" link in the footer. That price is the most honest size signal a sender produces, because they have to justify it to people writing checks.
Newsletter sponsorship is usually priced as a CPM on the list, meaning a rate per 1,000 subscribers or per 1,000 sends. So the math runs backwards cleanly:
subscribers ≈ (flat per-issue price ÷ niche CPM) × 1,000
Worked example. A B2B SaaS newsletter charges $1,200 for the primary sponsor slot. B2B newsletter CPMs in our sponsorship rates by niche data cluster around $40 to $75, so call it $45. That gives ($1,200 / $45) x 1,000, or about 26,700 subscribers. Cross-check it against a consumer rate and you would use a lower CPM, roughly $20 to $30, which is why the same $1,200 slot on a consumer list implies a bigger audience near 48,000.
The reason I trust this over everything else: the rate is priced on list size, not opens. Apple MPP never enters the calculation. The named tradeoff is the pricing basis. Some newsletters, and most of the beehiiv ad network, price per click or per open instead of per send. The moment a rate card quotes CPC or cost-per-open, you are back in open-rate territory and this method loses its main advantage. Check the unit before you trust the output.
Method 2: estimate subscriber count from open rate (mind the MPP trap)
The textbook formula is the one everyone reaches for:
subscribers = opens ÷ open rate
It is correct and it is a minefield. Two things go wrong. You rarely have a clean open count from the outside, and the open rate you do find is inflated by Apple Mail Privacy Protection.
Here is the part the Quora threads miss. Across the consumer lists in the Newsletrix corpus, 55 to 70 percent of recorded opens now come from Apple proxies that pre-fetch images, not from humans reading anything. That pushes reported open rates 40 to 44 percent above genuine human opens. We pulled those figures from our own analyzer, and they line up with what we wrote in the Apple Mail Privacy Protection breakdown.
So the trap is mixing sources. Say a media kit brags about 18,000 opens at a 45 percent open rate. Divide those two directly and you get 40,000 subscribers, and the inflation cancels because both numbers came from the same inflated measurement. The error happens when you take the 18,000 opens and apply a "realistic" 28 percent benchmark you read elsewhere. Now you get 64,000, and you have overstated the list by 60 percent because you corrected the numerator and left the denominator alone.
If you only have one number and must assume the other, undo the inflation first. Divide a reported consumer open rate by 1.42 to recover the rough human open rate before you reason about engagement. For anchors when you have nothing else, reported open rates in our open rate benchmarks run near 44 percent for media, 38 percent for B2B SaaS, and 36 percent for ecommerce. Those are the inflated public numbers, which is exactly what you want when your open count is also inflated. Keep both on the same footing.
Method 3: platform and ESP tells
Before you trust any number, figure out who sends the mail, because the platform alone brackets the plausible range. Detect the ESP from the raw headers and footer links. A brand running Iterable, Braze, or Salesforce Marketing Cloud is almost never under 100,000 subscribers, because nobody pays those contracts for a small list. A sender on Mailchimp's free or Essentials tier is usually under a few thousand. ESP tier is a size signal on its own.
Substack hands you a second tell for free. A post's public like count is usually 1 to 3 percent of the people who opened it. Multiply likes by 40 to 100 and you get a rough open floor, then divide by an open rate to back into list size. It is crude, but it sanity-checks the other two methods, and for a hidden-count Substack it is often all you have. Beehiiv leaves the "powered by beehiiv" badge and a recommendation network you can read the same way. For a deeper read on platform analytics, our Substack analytics comparison covers what each surface exposes.
Find the sending platform first
Paste a competitor's email HTML into the Newsletrix ESP detector to identify the sending platform, tracking domains, and click-wrapping in about three seconds. The ESP tier alone narrows the list-size range before you run any math.
Try the ESP detector →How to find a subscriber count that's already public
Sometimes the estimate is unnecessary because the sender published the number. Check these before you start any math. Substack subscribe widgets often state it outright, like "Join 24,000 readers." LinkedIn newsletters show a public subscriber count on the masthead. Media kits and sponsor pages print "X,000 subscribers" because advertisers ask. Sparkloop and recommendation widgets sometimes expose counts too.
Treat every published number as a marketing figure, not a database query. Senders round up, quote stale highs, and rarely separate free from paid. A Substack that says 24,000 is usually counting free and paid together, with paid running 5 to 10 percent of the total. If you are sizing revenue, the paid slice is the only number that matters, and it is almost never the one they print. For more ways to recover this from the outside, see how to read competitor newsletters without subscribing.
Where every estimate breaks
Each method has a failure mode, and they do not all push the same direction. Knowing which way the error leans is what lets you correct it instead of guessing blind.
| Factor | Direction of error | Typical size |
|---|---|---|
| Apple MPP inflation | Overstates reach from opens | 40 to 44% on consumer lists |
| Ghost / inactive subscribers | Stated count too high | 10 to 30% of a list |
| Free vs paid split | Stated count hides revenue size | Paid often 5 to 10% |
| Segment-only sends | Opens understate full list | Varies, can be 50%+ |
| Rounded / stale claims | Either direction | Unknown, treat as soft |
The segment trap is the sneaky one. A sender with 100,000 subscribers might send a given campaign only to the 40,000 who opened in the last 90 days, to protect deliverability. If you size the list off that send's opens, you will report 40,000 and miss more than half the list. This is why a single send is a weak basis and why I cross-check against a sponsorship rate, which reflects the whole sellable audience.
Which method to trust
Rank them. The sponsorship back-solve goes first whenever a rate card exists, because it sidesteps the open-rate mess entirely. The open-rate formula goes second, and only with same-source discipline: opens and open rate from the same report, or no estimate at all. Platform and ESP tells go third, as the sanity band that tells you whether the first two numbers are even plausible for the contract that sender is paying.
If I had to put one figure in a client deck tomorrow, I would take the CPM back-solve, present it as a range of plus or minus 25 percent, and write the assumed CPM right next to it so anyone can challenge the input. That is the difference between an estimate and a number you made up. Start with the ESP, find the rate card, and let the open math be the tiebreaker, never the foundation.
Frequently asked questions
Can you tell how many subscribers a newsletter has?
Not exactly, but you can land inside a believable range from the outside. The three reliable inputs are a published sponsorship rate, a stated open count divided by its own open rate, and platform tells like Substack like counts or the sending ESP. Treat the result as a range with a stated confidence, not a single number.
How do you estimate subscribers from open rate?
Divide the open count by the open rate: subscribers equals opens divided by open rate. The catch is both figures must come from the same report so the Apple MPP inflation cancels out. If you take a media-kit open count and apply a lower benchmark open rate from somewhere else, you overstate the list, often by 40 to 60 percent.
How does Apple MPP affect subscriber estimates?
Apple Mail Privacy Protection pre-fetches images and registers opens that no human made. In the Newsletrix corpus, 55 to 70 percent of recorded opens on consumer lists now come from Apple proxies, which inflates reported open rates 40 to 44 percent. Any subscriber estimate built on opens overstates real reach unless the open count and open rate share the same inflated source.
How many subscribers does a Substack have?
Many Substack subscribe widgets state it directly, for example "Join 24,000 readers." Where the number is hidden, a post's like count is usually 1 to 3 percent of who opened it, so multiply likes by 40 to 100 for a rough open floor, then divide by an open rate to back into list size. The stated number usually counts free plus paid, with paid typically 5 to 10 percent of the total.
What is the most accurate way to size a competitor's list?
Back-solve from a published sponsorship rate. List size equals the flat per-issue price divided by the niche CPM, times 1,000. It is the most reliable method because the rate is priced on subscribers or sends, not opens, so Apple MPP inflation never enters the math. Check the pricing basis first, because per-click or per-open ad pricing reintroduces the open-rate problem.