Viral Surface

Getting users to bring users has long meant asking. Send an invite link. Earn a reward for bringing a friend. Please tap the share button. Every one of those carries the same cost: the user has to perform one extra action that has nothing to do with the work they came to do. Which is why, in most products, only one or two people in ten ever send an invite.

But there are places where exposure happens without anyone asking. Create a document, share it as a link, and whoever receives that link looks at the product's screen. They weren't invited and they haven't signed up, yet they already know what the product is and what it looks like.

Pagedeck found the gap while splitting acquisition sources apart. Last quarter 90 people signed up by clicking an invite link, while 340 people viewed a shared document, left, and came back later to sign up on their own. The first was a managed feature; the second was a path nobody had designed. The undesigned one was four times larger.

Three conditions have to hold for something to count as a surface.

1. It has to appear on its own while the product is used for its actual purpose. The moment a user does something extra for the sake of spreading it, that's a request, not a surface.
2. It has to be visible to someone without an account. A screen that circulates only inside the team is not exposure.
3. Whoever sees it has to be able to recognize the product and get in from there. Seeing a screen without learning the product's name produces nothing.

In practice the forms cluster into five.

1. Shared output. Someone receiving a document or report link views it inside the product's own screens.
2. Collaboration entry. A viewer wanting to leave a comment or an opinion needs an account to do it.
3. Marks left on artifacts. The logo in a downloaded file's corner, the made-with line under a shared view.
4. Byproducts of work. Links in meeting invitations, email signatures, automated notification mail — all of them carry the product's name.
5. Public pages. A document left open to everyone gets indexed in search results and brings in people with no connection to anyone.

What you manage is three steps per surface: how many people were exposed, what share crossed over to the product, and what share of those signed up. Pagedeck's shared links drew 12,000 viewers last quarter and 340 signups, or 2.8 percent. The reason each surface keeps its own numbers is that the levers differ. Repositioning the signup prompt on a viewer screen and rewriting an email signature are not the same job.

There are three levers. Where the mark sits and how large it is, what action the viewer is offered, and whether the mark can be removed at all. The last is the awkward one. Selling logo removal as a paid feature makes revenue, but paying users tend to be the ones who share most widely, so the mark disappears first from exactly the places doing the most spreading. Most products split the difference: the mark stays for free accounts, shrinks for paid ones, and is only fully removable on the plans bought by customers who would never have shared publicly anyway.

The term appears most in design meetings about the set of tactics that produce adoption through product experience instead of sales. No ad budget and no sales headcount makes it the cheapest acquisition path available. Self-serve signup — where an exposed person creates an account without talking to anybody — sits directly behind the surface. It isn't free, though: making viewer screens usable for people without accounts takes engineering and design time.

It also comes up when designing a free plan. In a product where free users' output travels outside, the free plan isn't a cost line — it's a distribution channel. Putting sharing behind a paywall becomes a trade: gain one revenue line, close one acquisition path.

On the enterprise side it's the channel into an organization. When a viewer is a colleague at the same company, that person creating an account is the start of bottom-up adoption, and the path from one practitioner's tool to a whole department usually opens here. Separating internal viewers from external ones in the counts pays for itself.

Some products don't suit it. An internal data analytics tool or an accounting system, where nothing produced ever leaves the building, has almost nowhere to put a surface. Forcing adoption there sends you back to asking users for favors, which is worth knowing before a quarter gets spent looking for a surface that the product's shape does not allow.

"Twelve thousand people viewed shared links and 340 signed up. It's our biggest surface and we've never touched it."

The exposure already exists, so the work is the conversion at that point.

"Charge for logo removal and the logo disappears first from the teams that share the most."

Revenue and distribution are pulling the same lever in opposite directions.

"Split the viewers on our own customers' email domains out from the rest."

Spreading inside an organization and spreading outside it are different paths.

What Pagedeck did about that 2.8 percent was rework the viewer screen. The signup button in the top right came out, and instead, a viewer who clicks anywhere in the document to edit it gets asked whether they want a copy of their own to work in. The account request moved to the moment that copy gets saved.

Two months later signups from the same surface stood at 6.1 percent. Viewer counts were flat; what changed was what those people were able to do. Letting someone edit a copy immediately also shortened time to wow, the gap between signing up and the moment the product impresses you.

The most common reading is that this means building a share button. The word viral brings up a picture of a user spreading something. But a button is a request for one more action, while a surface is exposure riding on an action that already happens. Left uncorrected, a team keeps adjusting the placement and wording of a share button while the viewer screen most people actually pass through stays unusable for anyone without an account.

Assuming a bigger, louder mark spreads further is the second mistake. It follows from treating exposed area as proportional to traffic. In practice, the more intrusive the mark, the less willing users are to send that output anywhere. Blow the logo up to half the screen, gain a little on signups per view, and lose 30 percent of the sharing, and total traffic falls. What's worth tuning isn't the size of the mark but what the viewer can do on the spot.