🗳️ Every question needs a job

I spent the better part of yesterday planning my less-than-annual reader survey for this newsletter and deciding which questions to fit in.

Which is why this edition will feel a bit meta.

It’s what I’ve been working on this week, so it’s what’s on my mind, and I think it’s a good topic to help your next content planning cycle. But that also means you’re getting a newsletter edition about reader surveys, with one attached. Please don’t read it just as a promo for my survey.

(But also, please take the survey. It’s right below this editorial.)

Why asking helps and why we rarely do it

When I suggest you need a reader survey, I imagine you’re thinking, “Ugh, I know, I know.” You probably don’t need convincing that asking your audience beats guessing what they want. What’s missing is the time and a reason to do it this month instead of someday. (oh, that mythical “someday”!)

Q4 is a darn good reason. You’re about to decide what next year’s content looks like and which new initiatives make the cut. A survey that closes before that planning meeting feeds the plan. But to work, it needs to be done right.

A good survey is a small exchange.

People give you a few minutes, and in return you do something with what they tell you. So it has to work for both sides: every answer has to feed a decision you’ll make, and every question has to be quick and easy to answer.

Balancing the two is a series of judgment calls. So here are the ones I made this week and the bigger reasons behind them that could apply to you.

For you: would the answer change anything?

Back in 2021, I wrote a blog post with a test for survey questions: ask yourself how your work will change once you know the answer. If you can’t name something specific, cut the question.

I still use it. What I’ve learned since is that for recurring surveys, you need to rerun the test every time, because a question can help one year and be fully useless the next.

Take the send day. Asking about it in 2024 made sense: I had no baseline and no idea what you preferred. Almost half of you said you didn’t mind the day. My send tests this summer backed that up, too. Asking again would be theater, because the decision is settled.

Other questions stay, rebuilt so they can drive a decision.

Last year I asked which topics you wanted, and nearly everyone ticked “marketing-related.” Of course you would, it’s a marketing newsletter, duh! I couldn’t plan a single edition from that answer, though.

This year, I changed the question to include the eight topics I actually write about, and a mark telling me if you want less or more of each. Paired with the questions about your biggest challenges, that tells me what new editions should cover.

One more version of the test: don’t ask what you can already look up. In a customer survey, “Which plan are you on?” and “How long have you been with us?” are already sitting in your CRM, and each one still costs the person answering a click.

For them: can they actually answer it?

The person at the other end of your survey is usually filling it out in the few minutes between two meetings. They need questions they can answer without guessing, and memory is the first trap.

I’d planned to ask how you first found this newsletter. Then I thought about how many of you have been reading for years. Would you really remember? I’m not sure I would.

So I rebuilt the question: “When you find a new resource worth following, how does it usually happen?” That’s a habit you can describe today, and it tells me more about where to show up next.

Grounding a question in real behavior helps with more than memory. It ties the answer to something the person really did. Ask “Would you use this feature?” and you’ll collect a pile of polite yeses from people who never will. Ask “How have you solved this problem before?” and you get a behavior you can build on.

Then there’s effort.

My survey takes about three minutes, because most questions are a single click. How much work you ask of people comes down to a choice you make for every question, and here it is:

Open or closed? (Depends on what you don’t know yet)

Last year’s best answers came from the open-ended questions. They gave me content ideas, the testimonials on my newsletter landing page, and a list of the other newsletters and creators you read.

Open questions are worth it when you’re exploring: when you don’t know the answer yet, when you want people’s own words, or when the details around the answer matter. Keep it to a couple, though. Every open box is real work for the person answering.

When you can, make open-ended questions as specific as possible so they’re easier to answer.

My favorite this year is “What’s the marketing question you’d most like answered right now?” It asks about the problem on your desk this week. “What topic should I cover?” asks you to do my planning for me, and it’s too abstract for most of us to answer.

Close-ended questions are for counting, so their options have to split people apart. Three principles guided mine:

• Give people options they can sort themselves into. Free text is hard to count, and rarely adds much value: last survey’s industry answers ranged from “IT” to “Coffee.” Four offer types (B2B tech, B2B services, B2C, or a mix) give me enough detail about who’s reading and let a person answer with one click.
• Force a choice when everything sounds relevant. My last survey let people pick three ways they find new resources, but the votes spread so evenly that nothing stood out. So I lowered the limit to two picks, forcing people to prioritize and showing me the channels that matter most.
• Label every point on a scale. On an unlabeled 0 to 10 slider, a 6 means nothing in particular. Five labeled points, each with a one-line definition, tell me which way you lean between strategic and practical topics, and how far.

***

A lot of my client work starts in the same place: a team guessing what its audience needs, because nobody has had the time to ask. A survey is the cheapest fix I know.

If you’re heading into planning season, I’d start with the three or four decisions next year’s content strategy depends on: which topics get more room, which formats get the budget, which angles people see as uniquely yours. One question per decision, a couple of open ones for what you don’t know yet, and you have a survey you can send before the plan is locked. But make it good for both you and your participants!

Good questions matter,

👌 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

Have brands lost the plot in comment sections? | A super helpful piece, breaking down community management into four modes: table stakes, banter, good times, and hero. The point that stood out to me: “one hero moment is worth 50 banter moments.” It is a great bridge between customer experience and community work, with the focus on customers over virality and eyeballs.

How Vanta’s $517K/month SEO Moat Translates Into AI Visibility | Foundation’s breakdown shows how Vanta organizes its content into deep “Collections,” one per compliance topic, and why that structure keeps it cited in AI answers. It can look out of reach for a small content team, but I think it’s the case of “eating the elephant one bite at a time.” Start with one deep collection, then build the next.

Evergreen Content Is Over – The Individual Is The Only Strategy Left | A good tie-in between three things often discussed separately: zero-click search, AI visibility, and thought leadership. Its most interesting idea to me is the “reverse halo effect”: the brand used to make the journalist, and now the journalist makes the brand. The same goes for subject matter experts and is exactly why I’m so bullish on thought leadership.

👀 Two for fun

This Emmy-nominated subway show | I hadn’t seen a full episode of Subway Takes until I heard it was nominated. One guest, one hot take, one subway car, and proof that interesting content doesn’t need polish.

This very good Venn diagram | Having fun with data viz is a nerdy type of fun, but I love it.

🐈 Weekly cat

My kind of grocery shopping!