I recently got to work on a campaign that lived entirely online — no print, no in-store, no physical anything, just a message and the internet to carry it. For a brand used to having shop windows and glossy pages to lean on, that's a genuinely different exercise, and I learned more from it than from anything else I've done this year. So here are the notes, while they're fresh.
Everything is measurable — and that's the trap
The obvious appeal of an online-only campaign is that you can measure everything. Every click, every share, every second of attention, every step towards the thing you wanted people to do. Compared to a print ad — where you launch it into the world and mostly pray — it feels like turning the lights on.
But that visibility is also the trap, and it caught us more than once. When you can measure everything, you're tempted to optimise everything, and to mistake the easy-to-measure for the important. The number that's simplest to move — clicks, say — is rarely the number that matters, and a campaign can look like it's "working" on the dashboard while doing nothing for the brand or the sales. The discipline an online campaign demands is deciding, up front and honestly, what actually counts as success, and then refusing to be seduced by the prettier numbers that don't.
The message has to do all the work
Here's what caught me off guard. When you strip away the physical — the tactile quality of a printed piece, the atmosphere of a shop, the presence of a place — the message itself has to carry the entire load. There's nowhere to hide. Online, a person is one unimpressed half-second from gone, and no production values or store ambience are there to buy you goodwill. If the message isn't immediately clear and immediately compelling, you've lost them, and the analytics will show you exactly how fast.
This is where I found my communications instincts earning their keep hardest. An online-only campaign is a pure test of whether you actually understand your audience and can say something that lands with them, fast, in their terms. All the measurement in the world can't rescue a message that doesn't connect — it can only tell you, precisely and a little brutally, that it didn't.
Data and creative have to be in the same conversation
The failure mode I watched for was the two halves of the campaign talking past each other — the creative people making something they loved, the data people measuring something unrelated, and neither changing what the other did.
What worked was keeping them in one conversation. The creative shapes what you make; the data tells you, quickly, whether it's landing — and crucially, you can act on that within the campaign, not in a post-mortem afterwards. That's the real gift of online: the feedback loop is fast enough to change the thing while it's still running. You try a message, you watch how people respond, you adjust. But that only works if the people making the thing and the people reading the numbers are genuinely listening to each other, rather than defending their own half. When they do, the campaign gets smarter as it goes. When they don't, you just get a very well-documented failure.
The moment the data changed the plan
Let me make the feedback loop concrete, because it's the moment the whole thing clicked for me. We had two versions of the central message: one clever, layered, the kind of thing the creative side (me included) was quietly proud of; and one much plainer, almost blunt, that we'd nearly dismissed as too obvious. We ran both.
The plain one won, and not by a little. The clever version got admired and scrolled past; the blunt one got acted on. The numbers were saying something I didn't want to hear but needed to: this audience wants clarity, not cleverness — get out of your own way. So mid-campaign, we shifted weight to the plainer message, and the results lifted.
That's the whole gift of an online campaign in one story. Our taste — what we, the people making it, found satisfying — was not the same as what the audience responded to, and without the fast feedback we'd never have known. The data was humbling precisely where it was most useful: at the point where it contradicted us. A campaign you can't be wrong in front of teaches you nothing. This one could, and did.
What I took from it
Two things stuck.
First, measurability is a tool, not a virtue. The point of measuring everything isn't to optimise everything — it's to know, faster than you otherwise could, whether the thing that actually matters is happening. Choose that thing deliberately, and let most of the other numbers be context.
Second — and this is the one I keep coming back to — an online-only campaign is, at bottom, a communication problem wearing a marketing costume. The technology gives you reach and a fast feedback loop. It does not give you anything to say. The saying is still the whole job, and the data's real value is that it tells you, quickly and honestly, whether you said it well.
Online, you can measure everything. That's precisely why the hardest discipline is deciding what's actually worth measuring — and remembering the message still has to be worth reading.
I came away more convinced, not less, that the numbers are the easy part. The hard part is having something worth saying and saying it so it lands. Everything else is just how fast you find out whether you managed it.