Overview
Social listening here means unbranded discovery. You are not searching for your name or a competitor's — you are searching for the language people use to describe a problem before they know that a product category exists to solve it.
That distinction is the whole workflow. Someone writing 'we're drowning in manual data entry' has never heard of you, is not evaluating vendors, and is at the earliest and most winnable moment of a buying process.
The output is a prospect list with context attached: who has the problem, in their own words, with a timestamp. It is prospective and offensive, and it feeds pipeline rather than positioning.
Who this is for
SDRs and founders doing outbound who want warm context rather than a cold list.
Growth teams seeking net-new demand outside the accounts already in the CRM.
Anyone selling into a category where buyers describe the symptom fluently but do not know the solution has a name.
The problem
The most qualified prospects aren't filling out forms — they're posting on Reddit and Twitter about exactly the problems your product solves. You're not seeing those posts.
The VoxScrape solution
VoxScrape monitors Reddit and Twitter for keywords, pain points, and buying signals. AI Intent Detection scores every post. High-intent posts get surfaced first so you can engage at exactly the right moment.
Step-by-step workflow
- 1
Set up keyword monitoring
Enter the pain-point keywords your buyers use — not your product name. Think 'best way to find leads' not 'VoxScrape'.
- 2
Pull Reddit and Twitter in parallel
Run Reddit Pro and Twitter searches simultaneously. Pull posts from the past 7 or 30 days across your target communities.
- 3
Score for buying intent
AI Intent Detection scores each post 1–10. Posts mentioning evaluation criteria, comparisons, or budget score highest.
- 4
Engage and convert
Filter to 7+ scores. Export the post URLs and author profiles. Reach out with context — you know exactly what they're looking for.
Running it well
Search symptoms, never brands. This is what separates the workflow from competitor intelligence. Query the pain — 'spending hours on', 'is there a better way to', 'still doing this manually' — rather than any product name. Brand queries find people already evaluating; symptom queries find people before anyone else has reached them.
Layer in the phrase patterns that mark an active search. 'Looking for a tool that', 'what does your team use for', and 'anyone tried' consistently outperform topic keywords, because they capture the moment a person moves from complaining to shopping.
Score for intent rather than reading everything. AI Intent Detection at two credits per post ranks each one, and the practical threshold is seven and above for immediate outreach, four to six for a watchlist. That turns several hundred collected posts into a short actionable list.
Close the loop with identity. Reddit is pseudonymous, so the highest-intent posters are often found on LinkedIn instead, then enriched with LinkedIn Profiles. The first message references the specific problem they described, which is why this converts far better than an unprompted cold email.
How to tell it's working
Sourced opportunities is the metric — pipeline that exists because of a conversation you found, and would not exist otherwise. Mention volume is not a goal here.
Reply rate is the honest quality check. Outreach referencing a real post someone wrote should substantially outperform your baseline cold sequence. If it does not, the intent threshold is too low or the outreach is not actually referencing the post.
Speed matters more than in any other workflow. The window between someone describing a problem publicly and choosing a solution is often days, so time-to-first-touch is worth tracking directly.
Is this the right workflow?
This workflow finds people by what they are saying. If you need to reach a defined set of accounts or roles regardless of whether they have posted about anything, Lead Generation builds that list from an ICP description instead. To track named competitors rather than unbranded problems, use Competitor Intelligence.
Estimated credits
~60 credits for 200 posts with intent scoring
Credits only charged for results returned.
Sources used
Frequently asked questions
- How is this different from competitor intelligence?
- Opposite direction. That workflow searches named brands to inform positioning; this searches unbranded problem language to find prospects nobody is talking to yet. Different queries, different metrics, different readers — that one produces battlecards, this one produces pipeline.
- Which platform works best?
- Reddit for depth, because people explain their problem at length and unprompted. Twitter for immediacy, when reaction speed matters more than context. Running both is common and they surface genuinely different people.
- Should I reply directly on the platform?
- Rarely on Reddit — community norms punish anything that reads as selling, and a badly judged reply costs more than the lead is worth. Most teams find the person elsewhere and reach out with context instead.
- How do I keep the noise down?
- Filter to posts with engagement, and use intent scoring rather than reading the raw collection. Reviewing a ranked shortlist is what makes this sustainable as a weekly habit rather than a one-off experiment.