📋 Available on: Startup, Growth, and Enterprise plans. Deep Research is available to all customers on 2026 Growth and Enterprise plans.
Deep Research is a self-serve feature inside Archive that generates AI-powered research reports from your social listening data.
Instead of manually analyzing thousands of posts, you select a content view, pick a report type, and get a polished, structured report in minutes.
How it works
Navigate to the Deep Research tab in Archive
Click "New Research" and select a Social Listening content view (your data source)
Choose one of 5 report types
Optionally customize the analysis focus (e.g., "Focus on Nike vs. Adidas")
Your report generates in minutes, analyzing up to 10,000 posts
Sharing reports
Every report gets its own shareable link. Anyone with the link can view it without needing an Archive account.
Reports can be:
Downloaded as PDF
Copied as Markdown
Shared via an external link that anyone can access without logging in
Who can use Deep Research?
Deep Research is available on Growth and Enterprise plans.
Growth Plan — 70,000 credits/month included
Enterprise Plan — Higher credit limits and volume pricing, ideal for multi-brand, multi-market teams
Not available on Startup plans.
Pricing
1 credit per post analyzed. No base fee, no minimums.
A report on 1,000 posts = 1,000 credits
A report on 10,000 posts = 10,000 credits
You can also run a Test mode (100 posts) to preview a report before committing to a full run.
You can also check your current credit usage under Settings.
What problems does Deep Research solve?
Challenge | How Deep Research helps |
"I have all this data but no time to analyze it." Teams capture thousands of posts but lack the bandwidth to synthesize insights. | Point any view at the report engine and get a polished analysis of up to 10,000 posts in minutes. |
"Leadership wants a briefing and I'm building slides manually." Weekly reporting means hours pulling screenshots and writing summaries. | Generate a shareable report URL that leadership or clients can consume directly. No slides, no formatting. |
"We're reactive to brand crises instead of ahead of them." Negative sentiment builds before anyone notices. | Sentiment Analysis reports surface negative content patterns and emotional drivers so your team can respond before it escalates. |
Report types
5 report types are available:
Report Type | What it does | Best for |
Weekly Trends | Trending topics, content formats, engagement patterns | Weekly team briefings, content strategy |
Sentiment Analysis | Audience sentiment breakdown, complaint themes, emotional drivers | Brand health monitoring, crisis detection |
Share of Voice | Brand presence, sentiment, competitive positioning | Market positioning, executive reporting |
Creator Partnership Analysis | Deep analysis of paid creator content performance | Creator program optimization |
Brand Diligence | Investment-grade brand intelligence | Due diligence, brand audits |
Key benefits
Instant strategic intelligence — Go from raw social data to structured insights in minutes, not days
Shareable with anyone — Report URLs work for leadership, clients, or partners who don't have Archive access
Built on data you already have — No new setup or integrations. Reports pull from existing Social Listening views
Multiple lenses, one workflow — Trends, sentiment, competitive landscape, creator partnerships, share of voice — choose the analysis that fits the question
Frequently asked questions
How do I create a report?
Navigate to the Deep Research tab in Archive, click "New Research," select a Social Listening view, pick a report type, and hit create. Reports generate in a few minutes.
Can reports be shared with people who don't use Archive?
Yes. Every report has a shareable URL. Anyone with the link can view it — no login required.
Is there a limit on how many posts a report can analyze?
Yes, the current cap is 10,000 posts per report. You can also run a Test mode (100 posts) to preview before committing to a full run.
How does Deep Research select which posts appear in the report?
Every post in your view is analyzed. The posts that appear as cited examples in the final report are selected based on engagement (high-performing posts are prioritized), diversity (a mix of formats and content types), and relevance (5-10 representative posts per theme or trend). For Sentiment Analysis, negative sentiment posts get priority.
Can I create my own report templates?
Not yet. The 5 templates are pre-built. However, you can customize the analysis focus using a free-text prompt field when creating a report.
How is Deep Research different from AI Insider?
AI Insider is a broad, automated weekly digest of all activity across your workspace. Deep Research is targeted and on-demand — it runs against a specific saved content view you choose and generates a structured report for a specific use case. Think of AI Insider as "here's everything that happened" and Deep Research as "here's a deep analysis of this specific slice of data."
Do Share of Voice or Brand Diligence reports analyze competitor data?
All report types analyze the content in the view you select — your own data. Share of Voice will compare brands if competitor content is mentioned or captured within your view, but it does not pull external data.
Why does Deep Research pull from views instead of collections?
Views contain live data — they reflect the most current content based on your filters. Collections are static snapshots. Using views ensures reports are generated from the freshest, most relevant data.
How can I build the best view for Deep Research?
To maximize coverage, use Magic Fields, sources, and other filters to construct your content views rather than relying solely on SuperSearch keyword queries.
Are there any known limitations?
In rare cases, reports may include dead links to social media posts or show content that doesn't perfectly match a given section. If this happens, try regenerating the report. The 10,000-post cap per report cannot be increased at this time.
If you have questions about the technical capabilities of this feature, don't hesitate to contact our support team via Intercom chat or at [email protected].