Don't take the open rate's word for it

A media kit is a sales document. That is not an accusation; it is what the format is for. Nobody builds one to talk you out of a booking.

So the numbers in it are true, selected, and undated in whatever combination flatters most. Here is how to read past that in about twenty minutes.

Check one: does the subscriber number have a date on it?

"48,000 subscribers" is not a fact. "48,000 subscribers as of 1 February" is.

Lists decay. People change jobs, abandon inboxes, and get filtered into a promotions tab nobody opens. A number captured at a growth peak eighteen months ago can overstate a live audience by a third.

Ask for the figure with a date. Then ask a better question: how many people did the last four issues actually deliver to? Subscribers is a database count. Delivered is an audience.

Check two: open rate, and why it stopped meaning what it meant

Since Apple's Mail Privacy Protection began pre-loading tracking pixels, a meaningful share of recorded opens are a machine, not a reader. The effect varies by audience — consumer lists skew far more Apple-heavy than developer lists — which means open rates are no longer comparable between newsletters, and often not comparable to the same newsletter's own numbers from three years ago.

Treat a quoted open rate as evidence the sender knows their tooling, not as evidence anyone read anything.

The number that survived is clicks. Ask what a typical sponsored link gets, in absolute numbers rather than percentages, across the last several placements. An operator who has that to hand is an operator who has run sponsorships that worked.

Check three: how was the list built?

Ask directly. The answers sort into three groups.

• Earned — search, word of mouth, a referral programme, someone recommending the thing. These readers chose the newsletter specifically.
• Bought — paid acquisition, newsletter swap networks, co-registration. Cheaper to grow and much weaker: many of these readers do not remember signing up.
• Bundled — subscribers who arrived via a giveaway, a lead magnet, or a checkbox on an unrelated form. Treat as noise until proven otherwise.

A list can be honestly built and still be mostly bought. It just means the engagement numbers matter far more than the size does.

Check four: who has sponsored before, and who came back?

This is the check most buyers skip, and it is the one that predicts your result best.

A newsletter's sponsor history tells you two things a media kit never will. First, whether brands like yours have found the audience worth reaching. Second — and this is the real signal — whether any of them came back.

Repeat bookings are the honest metric in this market. A sponsor who runs once was persuaded by a media kit. A sponsor who runs four times has seen their own numbers and decided to keep paying. You cannot fake a renewal.

You can usually assemble this yourself by reading the last few months of the archive and noting who appears more than once. If you would rather not do it by hand, that is exactly what Bellwether Intel is for, and it is free.

Check five: how many ads run in a single issue?

Ask how many sponsored placements appear in a typical issue, and where yours would sit.

An issue carrying one placement gives it the reader's whole attention. An issue carrying six has turned into a classifieds page, and your primary slot is competing with five other logos for a reader who has learned to scroll past that whole region.

Ask for two or three recent issues as they were sent — not screenshots of your placement, the whole issue.

Check six: what happens if the send slips?

Newsletters are made by people, usually one person. Issues slip.

Agree in advance what happens: does the placement move to the next issue, is it credited, is it refunded? For a date-sensitive campaign — a launch, an event, a sale ending — a slipped send is not an inconvenience, it is the campaign not running.

Nobody minds being asked. An operator who has no answer has not run enough sponsorships to have needed one.

What to get in writing before you pay

Short list, and none of it is unusual to ask for:

• The issue date, or the specific issues for a multi-issue run
• The placement position, named the same way both sides name it
• The creative spec — dimensions, copy length, file limits — before you brief a designer
• Expected delivered volume, and what happens if it lands materially under
• The reporting you will get afterwards, and when
• What happens to a slipped send

If that list feels heavy for a £600 placement, it is because it is usually spread across nine emails instead of one document. Getting it in one place at the start is the cheapest part of the whole booking.

The short version

Subscriber count tells you the size of a database. Open rate tells you about a tracking pixel. A well-designed PDF tells you someone had a good designer.

Delivered volume, click numbers, how the list was built, and who came back — those four tell you whether to spend the money. If you only have time for one, make it the last.