Blog / Case study
A prediction market platform turns posts on X into first trades
We built an outreach system that automatically scans for relevant users on X, qualifies them with AI, and sends personalized invitations to try a prediction market platform.
The team reported an approximately 5% response rate, with around 20% of respondents completing a first trade. Together, those figures imply an approximately 1% conversion rate from outreach recipient to first trade.
- 5% response rate
- Outreach recipients who replied
- 20% respondent conversion
- Respondents who made a first trade
- Approximately 1% overall conversion
- Outreach recipients who made a first trade
Overview
People discussing crypto and prediction markets on X were a potential audience for the platform. Finding them required someone to search relevant topics, read posts and replies, review profiles, and decide who to contact.
Then came the outreach itself: writing messages, addressing each person, sending invitations, and keeping track of the relationship.
We connected those steps into a repeatable process. A bot searches X for relevant conversations, AI evaluates potential leads, and personalized outreach invites qualified users to try the platform.
The system also feeds a CRM and a mobile outreach app, giving the team a way to review prospects and manage messages from a phone.
The problem
Finding prospects required repeated searches
Relevant users appeared across posts and comment threads about crypto, trading, and prediction markets.
Someone had to find those conversations, identify the participants, and review whether their activity suggested a fit for the platform. New conversations appeared throughout the day, so the search needed to run repeatedly. Searching for new leads at such a constant cadence required heavy man-hours.
Qualification and messaging competed for the same time
Finding an account was only the first step.
The team still needed to assess its relevance, prepare an invitation, and send it. Handling each prospect manually limited how many people the team could reach.
Personalizing those messages added more work. Even using someone’s first name required checking their profile and editing the message before sending.
Outreach needed a record
Prospects discovered through X needed to carry into the team’s CRM so the relationship could continue beyond the initial message.
Without a connected process, discovery, outreach, and contact management each created a separate task, leading to extra man-hours and an inefficient process.
The objective
Build a recurring outreach process that finds relevant users, qualifies them, and invites them to try the platform with less manual work.
The team also needed a way to review selected leads, edit messages, and send DMs from a phone.
What we built
We built an X discovery and qualification system, personalized messaging workflows, an Attio CRM connection, and a mobile outreach app.
| Component | What it does |
|---|---|
| X discovery | Searches posts and replies related to crypto and prediction markets |
| AI qualification | Evaluates discovered users for relevance to the platform |
| Personalized outreach | Prepares messages that address recipients by first name |
| Automated messaging | Sends invitations without requiring each DM to be written and sent manually |
| CRM connection | Sends selected prospects into Attio |
| Mobile outreach app | Lets the team review leads, prepare drafts, edit messages, and send DMs |
Recurring discovery on X
The discovery system scans X every 15 minutes for relevant activity.
It captures posts and identifies potential prospects among the people posting and participating in the conversation. That gives the team a recurring source of leads without repeating the same searches by hand.
Within a few weeks, 2,199 posts were logged in the discovery sheet.
AI lead qualification
AI evaluates discovered users for relevance to the platform based on their activity around crypto and prediction markets. It also determines likelihood of the account being bot or human-operated.
This reduces the profile review required before outreach. The team receives prospects drawn from conversations related to the product.
Personalized invitations
The outreach workflow prepares invitations asking qualified prospects to try the platform.
Messages can address recipients by small things such as first name, allowing the team to personalize outreach at scale. Automated sending removes the need to manually compose and send every invitation.
CRM integration
The discovery sheet connects to Attio. Selecting a prospect sends that person into the CRM.
The company assembled 1,215 CRM contacts. That figure represents contacts stored in the CRM; it does not establish how many received an outreach message.
Mobile review and messaging
We also built a mobile outreach app for the co-founder.
Leads arrive from the X discovery system. The app lets him swipe through prospects, select people to contact, and prepare AI-drafted messages in batches.
He can then review, edit, and send each DM from his phone.
“This increased reach-out speed ~10x.”
How it works
| Stage | Input | Output |
|---|---|---|
| Discover | Relevant X posts and replies | Captured activity and potential prospects |
| Qualify | User activity and available profile information | Leads assessed for relevance |
| Personalize | Qualified lead and recipient details | Invitation addressed to the prospect |
| Send | Prepared message | Automated DM or reviewed message sent through the mobile app |
| Track | Selected prospect | Contact stored in Attio |
| Convert | Prospect engages with the invitation | First trade on the platform |
The impact
The team reported an approximately 5% response rate. Around 20% of the people who replied went on to make a first trade.
Assuming both percentages describe the same outreach group, that implies an approximately 1% conversion rate from message recipient to first trade.
| Funnel stage | Reported rate | Illustration per 1,000 recipients |
|---|---|---|
| Outreach delivered | Starting group | 1,000 recipients |
| Response | Approximately 5% of recipients | 50 respondents |
| First trade | Approximately 20% of respondents | 10 first traders |
| Overall conversion | Approximately 1% of recipients | 10 first traders |
The figures per 1,000 recipients illustrate the reported rates. They are not the campaign’s confirmed totals.
| Activity | Before | With the system |
|---|---|---|
| Prospect discovery | Repeated manual searches | X scans every 15 minutes |
| Lead review | Read posts, replies, and profiles individually | AI qualifies prospects |
| Message preparation | Write and personalize each invitation | Generate personalized messages |
| Sending | Send each DM manually | Automated sending, with a separate mobile review workflow |
| CRM entry | Transfer prospect information separately | Send selected leads to Attio |
| Mobile outreach speed | Previous manual process | Approximately 10 times faster, according to the co-founder |
The outcome
The platform gained a recurring process for finding and contacting people already discussing relevant topics on X.
Discovery runs four times an hour, which is much more frequent than a single human could reasonably handle. AI deals with qualification and message preparation. Personalized outreach reaches prospects with less manual effort, while the mobile app gives the team a way to review and send selected messages.
The reported funnel connects that activity to product use: approximately one first trader for every 100 outreach recipients.
The team was able to spend more time on product development as well as curation of higher-quality clients.