What Is a Marketing Signal?

You have more data than you’ve ever had. PostHog, Stripe, GitHub, Google Analytics. Dashboards full of numbers that update in real time. And most weeks, you stare at all of it and still don’t know what to post about.

That’s not a you problem. That’s a signal problem.

Data and signals are not the same thing. Data is everything your product generates. Signals are the small fraction of that data that’s actually worth talking about. The gap between the two is where most founder marketing dies.

Understanding what a signal is, and what it isn’t, is the first step to knowing what to say.

The Lemonade Stand Version

Imagine you run a lemonade stand. Every day you track how many cups you sell, what time of day, what the weather is. Numbers go up, numbers go down. Normal variation.

Then one Tuesday in March, you sell four times your usual volume. It’s not that warm out. You didn’t change the recipe. But you noticed that a youth soccer tournament was held two blocks away, and a kid who plays on one of the teams had posted a photo of your stand on Instagram the night before.

That’s a signal. The data told you something happened. The context told you why. The combination told you a story worth sharing.

A friend who’s been watching your stand all week would tap you on the shoulder and say: “Hey, you should probably post about this.” Not because they’re a marketing expert. Because they noticed something specific, surprising, and timely.

Glorbie is that friend. For your product.

What Makes Something a Signal

Not all unusual data points are worth talking about. A signal has four properties.

Anomalous. It’s statistically surprising relative to your baseline. “More users than usual” doesn’t count unless it’s significantly more. The threshold that matters is roughly 2x your normal baseline or better. A 10% uptick in signups on a Tuesday is noise. A 200% spike is worth asking about.

Specific. It points to something concrete. Not “engagement was up this week” but “users who completed the onboarding flow in under 3 minutes had a 40% higher 30-day retention rate.” The more specific, the more you can actually say something interesting about it.

Surprising. It tells you something you didn’t already know. If your best customers are power users who log in every day, and you see that your best customers log in every day, that’s not a signal. A signal challenges an assumption or reveals something you didn’t expect.

Timely. It’s happening now, or just happened. A signal that’s three months old is a historical footnote. Signals have a shelf life. The best posts are written close to the event they describe.

Data that’s anomalous, specific, surprising, and timely is worth paying attention to. Data that fails on any one of those dimensions is probably noise.

Internal Signals vs External Signals

Signals come from two places: inside your product, and out on the internet.

Internal signals are what your product data tells you. A cohort that retained at 3x your average. A feature that went unused for six months and suddenly lit up. A pricing tier where customers are upgrading faster than your model predicted. These signals come from PostHog, Stripe, GitHub, whatever you’re using to instrument your product.

Internal signals are usually the strongest. They’re specific to you and your users. No competitor has this data. The story is yours.

External signals are what’s happening in the world that’s relevant to your product. A thread on Hacker News where people are describing the exact pain your product solves. A Reddit post going viral in a community you’ve been trying to reach. A competitor announcing something that changes how people think about the problem space you’re in.

External signals matter because they tell you when the audience is already paying attention. Timing your story to an existing conversation is significantly more effective than trying to start one from scratch.

The best moments combine both. A Hacker News thread about a problem you just shipped a solution for, at the same time your activation rate doubled. That’s a post that writes itself.

Weak Signals vs Strong Signals

Not every signal demands immediate action. There’s a difference between a signal worth watching and a signal worth acting on today.

A weak signal is a hint. One data point that’s slightly outside normal variation. A small spike with unclear cause. A single mention in a niche community. Weak signals are worth logging. They’re not worth publishing about yet.

A strong signal is a pattern. Multiple data points pointing in the same direction. A spike you can explain. A conversation with real volume behind it. Strong signals demand a response quickly, because they’re time-sensitive by definition.

The practical question is: if I posted about this today, would it feel true and specific and interesting? Or would it feel forced? Weak signals feel forced. Strong signals feel obvious in retrospect, like: “Of course I should have talked about this.”

When a signal is strong, the cost of not acting is real. The moment passes. The conversation moves on. The post you would have written becomes irrelevant.

What This Means for You

If you’re a founder trying to build an audience or stay visible with customers, the most common mistake isn’t posting too little. It’s not knowing what to post about, guessing wrong, and putting out generic content that doesn’t connect.

The fix isn’t a content calendar. It’s signal awareness. Paying attention to what your data is actually telling you, what the internet is saying about your problem space, and what’s surprising enough to be worth sharing.

That takes time and a certain kind of attention that’s hard to sustain when you’re also building a product.

Glorbie automates the detection part. It watches your product data and the internet for you, filters out the noise, and surfaces the signals that are actually worth talking about. When something worth saying happens, it tells you. Then it helps you say it.

Knowing what to post is the hard part. The post itself is easy.