Newsletter benchmarks: how to build your own

Newsletter benchmarks are worth reading only when you know what sits under them. Published industry numbers rarely disclose the list, the period or the methodology behind them, so the figures you can trust are the ones you compute from your own sends. What follows is how to define each metric, fix the window and the sample it rests on, build a baseline from your own periods, and record it — all of it sitting on the compliance floor every commercial send has to clear.

What to take away

A newsletter list is a compliance asset before it is a marketing one. Every metric you pull from it, open rate, click rate, churn, sits on top of a legal floor you have to clear first. This is not legal advice; confirm your own obligations with a licensed attorney.

State privacy law adds another layer. In the United States that means the California Consumer Privacy Act, as amended by the California Privacy Rights Act, and the Colorado Privacy Act, each carrying notice, access and deletion duties for the subscriber data a publisher holds. The California Privacy Protection Agency enforces the California law, and the Colorado Attorney General enforces the Colorado statute.

What CAN-SPAM actually requires

The FTC's CAN-SPAM Act compliance guide covers commercial email sent by a business, and it applies whether you send it yourself or a vendor sends it on your behalf. The statute is the CAN-SPAM Act, 15 U.S.C. § 7701 et seq., and the FTC's rules under it sit at 16 CFR Part 316. The vendor rule is the one owners miss. If you hire an agency or a platform to run your sends, the responsibility stays with you.

Once someone opts out, you have ten business days to stop sending. That window is the reason suppression lists belong in your process, not in someone's memory. A subscriber who unsubscribes and still gets the next issue is the failure case that turns into a complaint.

The rest of the list is short: header information that does not mislead about the sender, a subject line that does not mislead about the contents, a valid physical postal address in the message, and a clear opt-out that works without making the reader do anything extra.

When the offer is a trial

Free trials and discounted first months are negative-option offers under the Restore Online Shoppers' Confidence Act, 15 U.S.C. § 8401 et seq. The FTC has said plainly that a "risk-free" trial that bills without clear disclosure is not risk-free. Its ROSCA recap lays out what the offer has to do before it takes a card: the material terms disclosed before billing information is collected, and the buyer's consent before the first charge.

Pick the numbers that change a decision

Run the arithmetic in your own variables. Let S be subscribers at the start of the period, D the emails delivered in it, U the unsubscribes, C the spam complaints, and A the new signups.

Every rate is a division with a denominator you should be able to name. Open rate is opens divided by delivered emails. Click rate is clicks divided by delivered emails — the same rate the worked example below uses. Click-to-open rate is clicks divided by opens, and it answers a different question than click rate does. Unsubscribe rate is unsubscribes divided by subscribers at the start of the period, and spam complaint rate is complaints divided by subscribers at the start. Write the denominators down once and use the same ones every period, so a moving number means your list moved, not that your arithmetic did.

Your churn rate is (U + C) / S, your unsubscribe rate is U / S, and your net growth is (A - U - C) / S. Churn is the wider number because it folds spam complaints in with the unsubscribes; the unsubscribe rate is the narrower one you quote to a sponsor.

Neither number means anything without a comparison window. Fix the window before you compute anything: open rate, click rate and revenue per subscriber belong to a single send, while churn, unsubscribe rate and net growth belong to a fixed calendar period, with month to month the working default. Keep the length the same once you start comparing — a new send measured against a mature back catalog reads as a collapse that never happened.

Track four things and stop. List size, churn rate, click rate on the main link, and revenue per subscriber. A high list count is not proof of economic value; a list of 20,000 that never clicks is worth less than 2,000 that do. If a metric has never changed a decision, retire it.

Revenue per subscriber is the revenue a single send brings in divided by the number of subscribers it reached. The worked example below does that arithmetic for a sponsor slot priced two different ways.

Then watch the sample. A rate built on one send is an anecdote, and a rate built on a handful of sends will swing on a single unusual issue. Record how many sends and how many subscribers sit behind each figure, and do not treat it as a benchmark until one bad issue can no longer move it. Comparing the middle of your own recorded range against the middle of the previous period is the honest version of an industry average.

Build the baseline table

One row per period — dates, subscribers at the start, delivered, opens, clicks, unsubscribes, complaints, new signups and revenue — and one column per rate, computed the same way every time. Your benchmark is then your own recorded range for that rate, not the best issue you ever had.

Keep the table in a spreadsheet you own and can export, stored next to your consent records. A platform migration should not take your history with it, and a figure you cannot re-check is not a benchmark.

Records that back the numbers

The IRS recordkeeping guidance says you can choose any system that clearly shows your income and expenses, as long as it supports purchases, sales, payroll, assets and other transactions. For a newsletter business that means sponsor invoices, platform fee statements, contractor payments and the 1099-NEC forms you file for anyone paid $600 or more.

Keep the consent record too. The timestamp and source of every opt-in is what you produce if a subscriber complains. If you cannot show when someone signed up and from where, you cannot show they agreed.

Worked example: a sponsor slot priced on the wrong number

Say you sell a sponsor slot at a flat rate R per send and your list is S subscribers. Cost per thousand impressions is R / (S / 1000).

If R is $400 and S is 4,000, that is $100 CPM. That figure prices a glance rather than a response, which is why the same slot looks different when you price it against clicks.

Now price it against clicks instead. If the click rate is c and the sponsor pays P per click, your revenue is P * c * S.

At a 2% click rate and $2 per click on 4,000 subscribers, that is $160, well under the flat rate. The flat rate only holds if you can show the sponsor a conversion number, and that is a conversation about their data, not yours.

Common questions

Does CAN-SPAM apply if a vendor sends the email for me?

Yes. The FTC guide states that the business on whose behalf the message is sent is responsible, and vendors that send on your behalf can also be liable. Put the compliance terms in the vendor contract and check suppression handling before the first send.

How fast do I have to honor an unsubscribe?

Ten business days under CAN-SPAM. Build the suppression list into the send process so the next issue cannot go to someone who already opted out.

Can I benchmark my churn rate against published industry numbers?

Only if you can verify the source and the list it came from. Most published newsletter benchmarks do not disclose their methodology, so use your own prior periods as the comparison and treat outside numbers as directional at best.

What records should I keep for a newsletter business?

Whatever system clearly shows income and expenses, plus the documents behind them: sponsor invoices, platform statements, contractor payments and opt-in records. The IRS recordkeeping page lists the categories; your CPA can tell you how long to hold each one.