Social and Community

There was a time when what a company said about itself and what a customer said about it carried the same weight. Advertising copy was the only description available, so when a vendor wrote "number one in the category," buyers had nothing else to weigh it against.

That order is now reversed. Someone choosing a tool for their team takes a lap somewhere else before they ever open the product page. They ask "has anyone here used this?" in a group full of people doing the same job, they read reviews that turned up in search, they look at a post from someone they know. The company's prepared message gets read afterward, and if it contradicts what they heard first, everything that follows loses its footing.

The difference comes down to who has something to lose. A company has no reason to speak badly of its own product. A person who keeps showing up in the same group does have a reason, because anyone caught overselling gets ignored on their next post. So the same sentence reads as an advertisement from a brand account and as evidence from a user. This is also why work in this branch moves slowly. Trust cannot be bought forward; it accumulates only as a record of showing up in one place and answering questions over time.

At Pagedeck the point landed as a lost deal. Sofia had spent two months on a sixty-person company, and at the last step the buyer closed the contract folder — not over features, not over price. In a group chat of product managers she belonged to, two people who had tried Pagedeck answered that they had switched away because it was slow. The performance problem had been fixed six months earlier. The fix had never reached that chat. No amount of explaining from Sofia outweighed two people from the buyer's own field. What made it unrecoverable was that the conversation had happened where nobody at Pagedeck could see it, and by the time Sofia heard about it the buyer had already made up their mind.

Social and community work exists so that space is not left empty. If advertising means renting a placement next to where people gather, this is closer to going inside: running an account, answering questions, and giving the people who use the product something worth mentioning. The price is control. An ad can be written down to the last word. What spreads here was written by somebody else.

The work underneath splits six ways: the channels you run yourself, the content users make for you, the borrowed reach of collaborations, the mechanisms that turn users into more users, the listening that tells you what is being said elsewhere, and the metric that says whether any of it worked. Six is worth separating because when results stall, the fix lives in a completely different branch each time. Nobody seeing you is an operations problem. People seeing you without reacting is a content problem. Reactions that never turn into users is a distribution problem.

Operations covers the channels you run directly, and it splits in two.

1. Organic social — running accounts and earning reach without paying for it. Reach means how many screens a post actually appeared on, and the number is set by the platform's distribution rules rather than by you. An account with 3,000 followers routinely reaches a couple hundred people per post, and that ratio shifts every time the platform changes how it ranks things.
2. Community building — creating and growing a space where users talk to each other. What separates it from the above is that it is not a place where the brand talks; it is a place where users ask and answer among themselves. Early on a moderator answers nearly everything, and once the space settles, longtime users start answering newcomers. Until that handoff happens, someone has to be present every day, which makes this the most labor-intensive branch of the six.

Content is the part users produce instead of you. User generated content (UGC) covers reviews, how-to posts, and screenshots people publish on their own. It reads better than a company-made video for the reason described above: it comes from a position where the author had no obligation to be complimentary. It also cannot be commissioned into existence. What a company can do here is make it easy to produce and move the good pieces somewhere visible.

Collaboration borrows reach from individuals who have already assembled an audience, and it splits three ways. What separates the three is how long the relationship lasts and what exactly is being paid for.

1. Influencer marketing — paying a person with an audience to feature the product, usually one piece at a time. The deal typically ends when the post goes out, and the fee is agreed per post in advance.
2. Ambassador — one person representing the brand over a long stretch. Because they appear repeatedly rather than once, the audience forms the impression that this is simply someone who uses the product. The trade-off is that choosing the right person takes months, and if that person draws criticism for something unrelated, the brand is tied to it.
3. Affiliate marketing — paying a commission only when a referral turns into an actual purchase. Nothing is paid up front, so a failed partnership costs nothing, but the person doing the describing is the one being paid on results, which is how exaggerated claims work their way in.

Distribution is the set of mechanisms that make users bring in more users.

1. Referral program — a structure that rewards an existing user for bringing in a new one. Rewarding only one side makes the referrer look like they are collecting a fee, so most programs pay both. The assumption people skip is that this only works on users who are already satisfied. Attach a reward to a product that is not good yet and you collect accounts that take the reward and leave.
2. Viral coefficient — the average number of new users one existing user brings in. It is the number of invitations a user sends multiplied by the share of those invitations that convert. If 100 users each send four invitations and ten percent sign up, the coefficient is 0.4. Above 1, the user base grows on its own with marketing switched off. Below 1, referrals merely add a bit on top of whatever else brings people in. Products that clear 1 are rare, so moving 0.4 to 0.6 counts as a real result.

Analysis is where you watch what is said outside your own channels. Social listening means collecting the places your name comes up and reading the context around it. As the lost deal above shows, decisions get reversed in rooms you are not in, so with this branch empty you do not even learn that wrong information is circulating. Some teams set up alerts and tooling to gather mentions; others assign a person to walk a handful of important communities on a schedule. Either way the output is a short weekly read rather than a dashboard, because the useful part is the wording people choose, not the count of mentions.

The metric branch is how you tell whether the activity is landing. Engagement rate is the share of people a post reached who then reacted to it — likes, comments, shares. Reach 1,000 people, get 30 reactions, and the rate is three percent. Teams watch this instead of follower count because followers only ever go up, while reactions are an answer about this specific post. An account with 3,000 followers reacting at six percent moves more people than one with 50,000 followers reacting at 0.2 percent.

The six run in sequence. Operations and content create something worth mentioning, collaboration carries it further, distribution mechanisms connect each arriving person to the next one, analysis feeds back what is being said outside, and the metric points at which box to fix. Teams that start from the metric tend to drift toward whatever is easiest to raise and end up with plenty of reactions and flat signups, because the posts that draw a crowd and the posts that get the product used are not always the same posts.

"Do we run this as an ambassador deal or a one-off with an influencer?"

Even with the same person being paid, the question is whether this ends after one post or runs for half a year.

"Someone brought up the speed issue in the community again. We fixed that six months ago."

Outdated information is still out there blocking deals, and somebody needs to go answer it.

"Engagement rate is up but signups are flat."

Getting reactions worked; the reactions are not carrying over to the product, so look at what sits between them.

At Pagedeck this work belongs to Ravi, but he does not do it alone. About half the questions in the user chat are about how a feature behaves, so Theo goes in once a week to answer those, and Noa takes the ones about why a screen is laid out the way it is. They decided to answer under their own names rather than from a single brand account, because the same answer reads differently when it comes from the person who built the thing.

After the deal fell through, one step got added to the release process. When something ships, a community note goes on the release checklist, and Ravi digs up the old posts where people complained about that exact problem and replies to them. Old posts cannot be deleted or buried, so the approach is to attach the fix underneath them. Six months later, when Pagedeck came up in that same chat, the line about it being slow was read alongside a reply saying it had gotten better. The person who wrote that reply was not Ravi but another user who had tried the product again in the meantime.

The most common misreading is that this work means posting regularly from a brand account. The visible output is posts and post counts are easy to tally, so the assumption comes easily. In practice the job is joining conversations that are already happening, and publishing announcements into a room where nobody invited you leaves reach flat no matter how the count climbs. Left uncorrected, a target like "three posts a week" gets set first, and a team out of material fills the slots with office news and holiday greetings. That is how a company posts 150 times in a year without adding a single person who talks about the product.

The other frequent mistake is expecting that building a community produces word of mouth by itself. Gather users and conversation looks like it should follow. But people gather around their own problem, not around your product. If there is nothing to be gained by showing up, you get a room where everyone joined and nobody speaks, and six months of that tells every new arrival that the product is quiet too. The thing to settle before launching one is not the name but what a person takes away from being there. Fast answers to hard questions, introductions to peers at other companies — something other than product announcements has to be on offer.