📺 Dropping and hoping

I watch all eight episodes of The Diplomat in about two days (which probably tells you more than what you need to know about my self-control).

It’s a great show. A strong female lead with a point of view, a marriage that doubles as a political alliance, about four moral compromises an episode.

Season two turned up unexpectedly eighteen months later. By then I’d more or less forgotten the show existed, and had to remind myself who half the characters even were.

Compare that to how I watched Game of Thrones. Ten episodes a season, one a week, Monday nights reserved. My calendar was marked with a new season every spring. I’d have binged that too, given the chance. HBO didn’t give it, and that turned out to be the point. They knew a loyal following isn’t built in a weekend binge.

I was reminded of both examples while reading a Slate piece about Netflix’s season-two problem. Put everything out at once, promote hard for two weeks, go quiet, come back whenever. Dropping and hoping.

Which, incidentally, is how a lot of us launch new content.

What the off-weeks were doing for you

The numbers Slate pulls together are brutal. Avatar: The Last Airbender took more than 21 million views for its first season and under 9 million for its second. The Four Seasons and A Man on the Inside each landed less than half their first season’s views. All of these were shows people liked, that lost most of their audience in the gap.

Slate puts it this way: “The time when a show’s not on the air is as important to building its audience as the time when viewers are actually watching.”

In marketing, we treat the gaps between publishing as dead air. The argument here is that they’re where the audience gets made.

I’ve had that work in my favor, which is how I know the mechanism is real.

In 2024 I stopped sending this newsletter for almost six weeks. No announcement, just silence as life started life-ing a bit too hard. In week three, a reader emailed to ask whether something had broken on their end, or whether I was okay.

If I’m honest, I found that more moving than any engagement metric I’ve ever looked at.

When I came back, open rates settled about where they’d been. Replies went up for the first couple of editions, mostly variations on “good to see you back”.

Turning up every week had booked me a slot in people’s inbox. A rhythm buys you that, and it’s the thing those Netflix shows never build.

The word that does the damage

There’s an upstream issue here, and it’s valid in both media and content marketing.

Netflix mostly commissions one season at a time, and showrunners write with that constraint. Audiences pick up on that. People get careful about how much emotion they invest in something when they can’t tell whether it’ll stay or get dropped.

We have a word for the same posture in marketing. We call it “an experiment.”

The podcast catalog shows what this leads to. Amplifi Media ran the count and found that out of two million shows, just one-third have ten or more episodes. The rest are dead on arrival, because the consistency was never agreed on.

The issue, I think, is that we’ve imported the evaluation rhythm of performance marketing into work that doesn’t run on it. A paid campaign can tell you something real in three weeks. A podcast cannot.

When you last proposed a new format, did anyone in that room say out loud when you’d all sit down and look at its performance?

There’s a second engine, and I wrote about it back in June: we get bored of our own work long before our audience does. Put the two together and a content program has almost no chance, especially as you factor in leadership pressure and new projects.

Showrunning, on a marketing budget

The fix is unglamorous, and it starts in the pitch.

When you propose a new content program, propose a season. Three things go in with it:

• How many. A defined number of episodes or editions. Ten, if you want somewhere to start arguing from, or 12, as this is roughly a three-month weekly schedule.
• How often. A cadence you can genuinely sustain (which is usually slower than the one you want to promise).
• When you’ll look. A named date when you’ll sit down with the people endorsing or approving the pilot and go through what happened, stats and qualitative data both.

That third one gets skipped almost every time, and it’s the one I’d highlight the most. A named review date gives the kill decision somewhere to live. Without it, that decision lands on a random Tuesday in week six, when your manager asks a passing question in a meeting about something else.

It also needs preparing. What would you accept as a “sign of life”?

Pick two or three signals and write them into the pitch next to the date. Repeat listeners rather than raw downloads. Whether reply rates move. Whether your sales team has started forwarding episodes to prospects.

One boundary, because this doesn’t apply to everything you publish. Cadence matters for the relationship work: the newsletter, the podcast, the point-of-view blog series you’re slowly building over a couple of quarters. Social media is much less dependent on this logic; an SEO post works on completely different rules; nobody is sitting at home waiting for your pricing page.

***

The Diplomat is on its fourth season now, which I’m glad about, but we know nothing more than “hopefully some time in late 2026”. By the time it turns up, I can’t promise my excitement for the characters will still be there.

That’s what dropping and hoping buys you. Attention in a burst, then a slow decay into “oh right, that thing.”

So before your next launch, ask which mode you’re in. The Netflix one, where everything is an experiment. Or the HBO one, where you build the rhythm in and let it do the work.

The second takes longer to prove. It’s also the only version where you find out the genuinely good ideas.

Rhythm matters,

👌 Handpicked stories for you

Here are some interesting things I stumbled upon this week while reading, writing, researching, and being a generally curious marketing person on the internet:

💼 Three for work

Marketing teams are stuck in single-player Claude mode. Here’s how to go multiplayer. | Most marketing teams run Claude in single-player mode: one person builds something genuinely good, and nobody else ever sees it. Emily Kramer’s fix is the 4 Cs, with a shared repo doing the version control so one person’s improvement lands for everyone. Relevant if your AI setup currently lives in three people’s chat histories.

The HRTech Content Benchmark: What 90,000 pages say about HRTech Content in the AI Era | This research on 90,000 HRTech pages found content that’s competently optimized and barely differentiated: only 5.7% show high information gain, 4.7% have strong narrative ownership. The trend line travels well beyond the HR category.

How to turn internal docs into content that converts | What’s already sitting inside your company makes for great content: objection-handling guides, battlecards, onboarding decks, support tickets. The practical recommendation from the post is the monthly sourcing meeting with sales and success, which turns this into a supply line rather than a one-off audit. Worth a look if your ideation currently starts from a blank page.

👀 Two for fun

This tour of the hand-made sign business | Neon, paint, and a great deal of patience. I loved watching an old craft get pulled into modern restaurant design.

This virtual drive through a random city | Pick a place, watch the road roll by, listen to whatever the local radio is playing. Weirdly calming.

🐈 Weekly cat

A dashboard-friendly accessory.