HotMentionHotMention

Core Concepts

Understanding how HotMention finds, scores, and delivers leads.

πŸ”‘

Keywords

Keywords are the phrases HotMention monitors across social platforms.

Think of keywords as buyer-intent signals. Choose phrases your ideal customer would use when they have a problem you solve.

Good keyword examples:

  • "referral program software"
  • "best CRM for startups"
  • "affiliate tracking alternative"
  • "project management tool for remote teams"

How matching works: a keyword matches when the phrase appears in the post, case-insensitively. There is no stemming or synonym expansion, so keep keywords short (one to three words): "affiliate tracking" matches far more real posts than"affiliate tracking software for SaaS". The AI intent score, not the keyword, decides whether the match is a lead.

πŸ’‘ Tip: Use AI-powered keyword discovery in the dashboard to find related phrases your customers use.

🎯

Intent Scoring

Every mention gets an AI-generated intent score from 0-100.

HotMention's AI reads each conversation and scores it based on buying intent. High scores mean the person is actively looking for a solution.

HOT
70-100
Actively shopping, ready to buy
WARM
40-69
Interested, needs more info
LOW
0-39
Casual mention, low urgency

What affects the score:

  • Urgency ("I need this ASAP" = higher score)
  • Specificity ("looking for a tool with X feature" = higher)
  • Budget signals ("willing to pay" = higher)
  • Competitor mentions (indicates active research)
  • Question format ("what's the best..." = higher)

πŸ’‘ Tip: Focus on HOT (70+) leads first. These convert at 5-10x higher rates.

🌐

Platforms

HotMention monitors 6 social platforms for buyer-intent conversations.

Reddit

(All plans (including Free))

High-quality discussions in niche subreddits. Great for B2B SaaS.

X (Twitter)

(All plans (including Free))

Real-time conversations, product launches, and feature requests.

LinkedIn

(Paid plan)

Professional networks, B2B decision-makers, and industry discussions.

Threads

(Paid plan)

Fast-growing community discussions, alternative to X/Twitter.

Quora

(Paid plan)

In-depth Q&A, long-tail searches, high Google visibility.

YouTube

(Paid plan)

Comment threads on channels you add β€” buyer intent shows up as replies asking what tool a creator used.

πŸ’‘ Tip: Start with Reddit (great signal-to-noise ratio), then expand to X and LinkedIn as you scale.

πŸ“Š

Pipeline Stages

Track your lead progress from discovery to conversion.

HotMention provides a kanban-style pipeline to manage your leads:

πŸ†•

New

Fresh leads matching your keywords. Default stage for all new mentions.

πŸ’¬

Replied

You've posted your reply. Paste the comment link and HotMention verifies it on day 1, 3 and 7.

🀝

Engaged

The author responded or the conversation moved forward.

βœ…

Converted

They signed up, bought, or became a customer. Track your wins!

πŸ“¦

Skipped

Not a fit, spam, or low quality. Keep your pipeline clean.

Why it matters: Moving leads through stages helps you measure conversion rates, optimize your approach, and focus on the right conversations.

✍️

Draft Replies

AI-generated replies customized for each mention.

For every lead with high enough intent, HotMention's AI writes a draft reply. It:

  • Reads the full conversation context
  • Matches your product to their specific need
  • Suggests helpful, non-spammy language
  • Keeps replies short and conversational

Example Draft

"Hey! You might want to check out [Your Product]. It's built specifically for [their use case]. There is a 3-day free trial to get started β€” no credit card needed. Happy to answer any questions!"

πŸ’‘ Tip: Always customize the draft! Add personal touches, reference their specific pain point, and be genuinely helpful.

πŸ›‘οΈ

Reply Survival

A reply only counts when it is still there a week later.

Reddit removes replies from accounts it does not trust, however good the text. In our previous product real people wrote the replies from a pool of accounts, and most were still deleted, while the same kind of reply from the customer's own account lived far longer. So the only account that posts through HotMention is the customer's (or their own agent's, from that account), and the product makes that reply survive, in three parts.

πŸ“ˆ

Subreddit survival score

Every subreddit carries a 0-100 share of verified replies that are still live. Scores start from a year of removal data and are recomputed daily from HotMention's own checks; once a subreddit has 5 verified replies, only our own data counts.

🧭

Posting guidance

Derived from the score: ok (at least half of replies survive), careful (fewer than half β€” answer first, one product mention, no link in the first comment), avoid (blacklisted, or fewer than one in five survive on our own data). The draft is written to these rules and every API response carries them.

πŸ”

Verification on day 1, 3 and 7

Paste the comment link when you mark a lead replied. HotMention reads the comment through the Reddit API three times and records Live or Removed. Three failed checks in a row give Unknown; replies on X and other platforms are Not verified.

⏱️

Reply budget

A per-project cap you set: 3 Reddit replies a day and 10 a week by default, 60 minutes apart. The dashboard, the notifications and the API all refuse to record a reply past the cap unless you force it.

πŸ’‘ The Overview shows Reddit replies live / removed / verifying and a 30-day survival rate. The weekly report repeats it, and the first reply that is still live on day 7 is worth celebrating.

πŸ“‘

Citation Share

Where AI engines look when they answer your buyers, and whether you are there.

Every project gets up to 30 buying prompts: category-level questions a buyer would ask (β€œbest email warmup tool for cold outreach”), never naming your brand or a competitor. Each week HotMention asks ChatGPT (with web search), Google AI Overviews, Perplexity and Claude every prompt several times, because answers vary from run to run, and records which brands and which sources each answer cites.

Cited in N of M
the scoreboard
M = prompts Γ— engines Γ— samples in the run. N = checks where your brand was cited. The share is N / M.
Delta
vs the previous run
Shown only when both runs asked the same number of checks. Week-to-week noise of Β±1 prompt is normal.
Prompts you lose
who is cited instead
For every prompt where you are absent: the brands cited, the engines, and the top source feeding the answer.

AI-cited threads. When a cited source is a Reddit thread worth replying in, it joins your lead feed with a draft and the badge AI-cited. These leads are capped at WARM, never HOT, never trigger notifications, and cost nothing until you act on one: they are strategic, not urgent. Filter for them with view = AI-cited in the dashboard or include_ai_cited in the API.

Competitors. Up to 5 per project, with domains, so the scoreboard can name who is cited instead of you and whether their own site or a Reddit thread is the source.

πŸ’‘ We do not claim that a reply moved the scoreboard. Engines answer probabilistically and change with every model update; the scoreboard shows where they look and whether you are cited, nothing more.

Next: Master the Dashboard

Now that you understand the concepts, learn how to use the dashboard to filter leads, manage your pipeline, and track analytics.