← All case studies
Positioning · Voice of customer · Messaging

I was hired to moderate a Discord. I left having built the education layer I pointed every new member to.

A category problem mistaken for a product problem. Two years of support-queue research, three segments, and a messaging ladder that became my default answer for anyone new.

Client

Distributed Crafts · BOB (Build on Bitcoin)

Hired as

Community Moderator

Reported to

Head of Marketing

Timeframe

Feb 2024 – Jan 2026

20K+Discord members moderated
12K+Telegram members moderated
25K+Views on content I wrote
10+Pieces published across four formats

What I was hired to do

Distributed Crafts hired me as a Community Moderator for BOB, a hybrid Bitcoin and Ethereum Layer-2 building toward becoming the "Bank of Bitcoin". The job was Discord and Telegram: welcome people, enforce the guidelines, handle escalations, keep the rooms healthy.

I did that job for two years, reporting to the Head of Marketing. This case study is about the part I added to it.

What moderating showed me that analytics could not

Moderating means reading every question, every day, in the order real people think of them. After a few months the pattern was impossible to miss.

The same handful of questions kept arriving. Not variations. The same ones. What is BTCFi? Is my Bitcoin still mine if I bridge it? Why would I want my BTC to do anything?

People were not confused by BOB's interface. They were confused about the category BOB existed in. Every product the team shipped assumed a mental model most Bitcoin holders had never been given: the one where Bitcoin does something rather than sits still.

A question asked fifty times is not a moderation problem. It is a missing piece of content with a timestamp on it.

Technical documentation existed, written for developers. Between "read the docs" and "ask in Discord" there was nothing, so Discord became the documentation, delivered by hand, one person at a time, by me.

Nobody asked me to

None of this was in my brief. I worked directly with the Head of Marketing day to day, and the mod queue was already telling us exactly what to publish. Answering the same question live, one person at a time, was the most expensive possible way to deliver an explanation.

So I built the education layer myself, alongside the moderation I was actually hired for.

The strategy: a ladder, not a masterpiece

The instinct with confused users is to write one great explainer. It fails, because "confused" is not one audience.

The queue held at least three different people: the Bitcoin holder who had never touched DeFi and half suspected it was a scam, the ETH-native user who understood DeFi but not BOB's bridge, and the newcomer who mainly wanted to know whether their money was safe. A single piece pitched at the middle talks over the first and bores the second.

So I built a content ladder, a structure running from "what even is BTCFi" up to restaking mechanics, where every rung stood on its own and pointed at the next. Readers entered at their own level instead of being talked over or talked down to.

The tiering was deliberate and visible in how I filed everything: beginner, advanced, and category-level pieces, each allowed to assume a different amount.

What I built

WTF is BTCFi: the series

The flagship explainer, then Why BOB Matters in BTCFi, Wrapped BTC vs Real BTC, Bitcoin vs Ethereum DeFi, Lending & Borrowing on BOB, and BTC Restaking.

BOB Gateway walkthroughs

Step-by-step guides to bridging BTC across 11 chains, including a full V4 walkthrough of native BTC ↔ WBTC transfers, written for the user who called Bitcoin "internet money", because that user was real and in the Discord.

Weekly community content

Ecosystem developments, launches and governance updates, so the channels had a reason to be checked between announcements.

Each topic ran across formats with distinct jobs: X threads for people already in the ecosystem, Substack for readers who wanted the full argument, short video for a single idea, and long video for anyone setting a product up for the first time.

The judgement call: translate, don't simplify

The temptation with a technical product is to strip detail until it reads easily. That loses the people who matter most, because the thing they need explained is the mechanism.

"Making your BTC do more than just sit there" is not a dumbed-down version of restaking. It is an accurate description that starts from what the reader already believes, that Bitcoin sits still, and moves exactly one step from there.

My test: would someone who understands this deeply wince at the sentence? If yes, it is wrong, no matter how readable it is.

Clarity bought with accuracy is not clarity. It is a support ticket with a delay on it.

Results

What the content did

25K+Views and engagements
10+Pieces across four formats
ReusedBecame the thing I pointed every new member to

Instead of answering the same question by hand, I had something to point at. That is the outcome I cared about: the explanation stopped living in my head and started living somewhere people could find it.

What I ran as a moderator

20,000+Discord members
12,000+Telegram members
2 yrsDaily moderation and escalation handling

Context: what BOB achieved while I was there

$250MTVL milestone
$100MBTC staked by 2,800 users during Fusion
10M+Mainnet transactions

These are protocol-level milestones, not content attribution. I include them to describe the environment I was working in. Not to claim I caused them.

"You are producing really top quality content." Caroline · COO, Distributed Crafts
"…consistently demonstrated diligence and a strong work ethic while moderating the BOB community on Discord, with a clear talent and enthusiasm for video content creation." Pankaj Dalmia · Community Lead, Distributed Crafts

What this work taught me

I went in as a moderator and came out understanding that the support queue is the cheapest audience research a company owns, and almost nobody reads it as research.

Every repeated question is a piece of content that does not exist yet, pre-validated by demand, with the exact wording your audience uses already attached to it. I did not need a survey or a persona workshop. I needed to notice what I was already answering by hand.

The other lesson: content that gets adopted beats content that gets published. The series mattered because the team started pointing people at it. That is the difference between making something and changing how a company works.