See the pattern.Keep the evidence.

Shopcue starts from one public Shopify product page, studies public TikTok, Instagram Reels, and YouTube Shorts evidence around it, and keeps Content, Creators, Competitors, Recreate, and Memory inside the same research Agent.

ILLUSTRATIVE PREVIEW · NOT LIVE PROVIDER DATA
Barrier Serum Agent4 patterns
Three illustrative skincare short-form video frames showing a serum routineIllustrative source clip frames
Top patternRoutine interruption
86product fit

One Agent.Six focused views.One sharedevidence graph.

The global product stays a quiet Dashboard with New Agent, recent Agents, View all, Settings, search, and provider status. Research modes appear only after opening a product.

01Dashboard

Continue the product that changed—not a generic trend feed.

See recent Agents, latest successful run, next run, provider health, new patterns, creator changes, competitor moves, and research that needs review.

  • Recent Agents
  • Research Delta
  • Provider status
  • Usage
02Content

Turn individual videos into evidence-backed patterns.

Normalize and cluster public videos by Hook, angle, mechanism, proof, format, CTA, sound, creator adoption, and lifecycle—always through the current product lens.

  • Rising and evergreen
  • Hook anatomy
  • Source videos
  • Counter-evidence
03Creators

Rank fit and repeatability, not follower count.

Compare public performance against each creator's own baseline, detect repeat outliers, show hit concentration, and bind every candidate to relevant product patterns.

  • Repeat outliers
  • Pattern fit
  • Own baseline
  • Competitor overlap
04Competitors

Research the product that competes for the sale.

Seed explicit public competitor product pages, then trace content coverage and change across products, videos, creators, hooks, proof, offers, platforms, and time.

  • product watchlist
  • Coverage preview
  • Pattern overlap
  • Whitespace
05Recreate

Adapt the mechanism without copying the asset.

Freeze a source Pattern and current Product Fingerprint, generate three original directions, run product-truth and rights checks, and optionally request a 9:16 draft.

  • Three directions
  • Script and storyboard
  • Creator brief
  • Veo draft
06Memory

Make every week smarter than the last.

Keep runs, evidence versions, saves, rejections, creator tracking, competitor seeds, generation lineage, failures, and lifecycle changes in one append-only research history.

  • Run history
  • User decisions
  • Version lineage
  • CSV and JSON export

No hidden magic.Every stage hasinputs and failurestates.

Runs are queued, leased, retried, and preserved. Each platform can fail independently. A failed refresh never deletes previous evidence or pretends that a stale result is new.

  1. 01
    Build lens

    Extract public product facts and wait for the user to confirm claims, use cases, audience, and search terms.

  2. 02
    Collect

    Query configured public-content providers independently by platform and record exact coverage.

  3. 03
    Normalize

    Reject unsupported fields, deduplicate canonical content, and preserve observation time.

  4. 04
    Cluster

    Group related hooks and mechanisms into stable Pattern versions across creators and platforms.

  5. 05
    Score

    Calculate lifecycle, outlier, quality, product fit, saturation, and confidence from inspectable inputs.

  6. 06
    Prepare

    Surface opportunities and Delta without erasing older evidence when the latest run fails.

Ranked by
visible inputs.

Shopcue does not output “viral probability.” It calculates evidence grades from versioned public observations and explains why one opportunity ranks above another.

01

Breakout

Reach relative to creator baseline, velocity, freshness, and independent adoption.

02

Creator opportunity

Repeat outlier rate, engagement quality, hit concentration, and pattern consistency.

03

Hook state

New, rising, evergreen, fading, or unproven from time-series evidence.

04

product fit

Lexical and structural overlap with verified product problems, outcomes, use cases, and audience.

05

Confidence

Evidence count, creator diversity, platform diversity, freshness, and missing fields.

Enough to audit.
No raw provider dump.

Shopcue stores normalized public evidence, product truth, user decisions, and lineage. Provider secrets and raw payloads stay out of browser records and Memory exports.

  • Public source URL, platform ID, creator handle, publication and observation time
  • Views, likes, comments, shares, follower baseline, and the fields that were unavailable
  • Pattern membership, Hook family, mechanism, lifecycle, product-fit terms, and exclusion reasons
  • Product Fingerprint version, approved facts, forbidden claims, market, language, and Research Lens
  • User saves, rejections, tracking decisions, competitor seeds, Recreate versions, and provider failures
research Agent
A persistent research agent bound to a single product, keeping Content, Creators, Competitors, Recreate, and Memory in one shared evidence graph.
Product Fingerprint
The versioned set of approved product facts, use cases, audience, claims, and forbidden claims that every research decision is measured against.
Research Lens
The product-specific research frame — category, problems, competitors, adjacent products, and search terms — used to interpret public evidence.
Outlier
A video that significantly outperforms its creator's own baseline — not just raw views. A creator-relative outlier is a signal the market already validated before you spend a dollar testing it.
Pattern
A stable group of repeated hooks and mechanisms across creators and platforms, with evidence, lifecycle, and momentum behind it.
Creative Cue
When multiple outliers converge on the same hook, format or angle, Shopcue turns them into a Cue — a direction worth testing, backed by the source videos that prove it. Not a vague AI score.
Memory
Append-only research history: runs, evidence versions, saves, rejections, tracking decisions, competitor seeds, generation lineage, and failures.

Public content in.
Product-specific learning out.

No TikTok Shop orders, attribution, ad automation, creator outreach, publishing, or guaranteed performance. Every unsupported provider and paid capability remains labeled.

What Shopcue isWhat Shopcue is not
Shopcue starts from one public Shopify product page, studies public TikTok, Instagram Reels, and YouTube Shorts evidence around it, and keeps Content, Creators, Competitors, Recreate, and Memory inside the same research Agent.No TikTok Shop orders, attribution, ad automation, creator outreach, publishing, or guaranteed performance.
Every cue traces back to real outlier videos — the hook, the format, the creator baseline and the source URL. Not an invented AI number.Shopcue tracks only product-matched public evidence. It does not provide private contact data, automated outreach, pricing, or Creator CRM.
A public Shopify product page defines the product, and a server-side public-content provider supplies research evidence.Shopcue does not require Shopify Admin, TikTok Partner approval, ad-account access, customer PII, or manual CSV imports for the Shopcue MVP.
Unconfigured providers, failed runs and unavailable private metrics stay visible. The interface never claims success it didn't earn.Shopcue reports source-level health and partial coverage instead of substituting demo records when a provider is unavailable.

Sources

Research evidence is collected from public TikTok, Instagram Reels, and YouTube Shorts pages and profiles only. Private metrics such as spend, orders, ROAS, and audience demographics are never collected.

Official platform documentation

Review the data boundary