Build-A-Bear (BBW) — Social Demand Tracker

A small, GitHub-Pages dashboard that measures whether the adult Build-A-Bear fandom is real, growing, and durable — using several independent social signals.

Thesis. Build-A-Bear is not a dying mall retailer for 4-year-olds. It is a high-recognition brand (90%+ U.S. aided awareness, ~20M loyalty members) whose customer base is already ~40% teens & adults ("kidults"), comparable to Warhammer collectors. If adult demand is accelerating, search, community, culture and the resale market should all point the same way at the same time. This tracker tests that.

Live dashboard: enable GitHub Pages (Settings → Pages → Branch: main, folder /docs), then the URL appears here.

Modeled on the Warhammer demand dashboard.

---

How it's built

docs/                 # the static site (GitHub Pages serves this folder)
  index.html
  app.js              # renders KPIs + Chart.js charts from data/*.json
  styles.css
  data/               # the data the page reads (versioned in git)
    manifest.json     # thesis, status, last-updated
    kpis.json         # headline figures (real, cited)
    google_trends.json reddit_subscribers.json reddit_topics.json
    social.json roblox.json resale.json retail.json
scripts/              # data collectors — you run these to refresh data/
  fetch_roblox.py            # live Roblox "Build-A-Bear Tycoon" visits
  fetch_reddit.py            # current subreddit size (Reddit API)
  backfill_reddit_wayback.py # subscriber history (Wayback Machine)
  ingest_google_trends.py    # parse Google Trends CSV exports
  snapshot.py                # dated data save points
data_raw/             # (git-ignored) drop Google Trends CSVs here

Data lives as JSON inside the repo, so every refresh is a git commit — the demand history is visible in the git log, and the dashboard is re-runnable and diffable over time.

Sections

Search · Community (Reddit) · Social · Digital (Roblox) · Resale · Retail & commercial · Signal convergence · Methodology. Each analytical section ends with a "What this shows & why it matters" institutional read-through.

• Real / live data: KPI strip, retail & commercial segments and locations (SEC filings), Roblox visits (live API), resale prices (eBay sold).
• Illustrative (flagged on the page): the Google Trends and Reddit time-series and the Reddit conversation taxonomy — seeded for design and overwritten when the collectors run. The Reddit Dec-2023 point (27,351) and the latest Roblox point are real anchors.

Refresh the data

pip install -r requirements.txt

# 1) Roblox — live "Build-A-Bear Tycoon" visits (no keys needed)
python scripts/fetch_roblox.py

# 2) Reddit — current community size  (needs a free Reddit "script" app)
export REDDIT_CLIENT_ID=xxxx REDDIT_CLIENT_SECRET=xxxx
python scripts/fetch_reddit.py

# 3) Reddit — historical growth from Wayback Machine
python scripts/backfill_reddit_wayback.py

# 4) Google Trends — export CSVs to data_raw/, then:
python scripts/ingest_google_trends.py

git add docs/data && git commit -m "data: refresh signals"
git push

Each script header explains its 30-second setup.

Saving & rollback

Everything is versioned, at two levels:

1. Named project save points (git tags). Each milestone is tagged (see CHANGELOG.md). To look at or roll the whole repo back to one:

git tag                       # list save points (v0.1.0, v0.2.0, …)
git checkout v0.1.0           # inspect that exact version
git checkout main             # back to latest
# to truly revert main to an old version:
git revert --no-commit v0.2.0..HEAD && git commit -m "roll back to v0.1.0"

2. Data snapshots (browsable files). Before/after any data change, run:

python scripts/snapshot.py "what changed"
git add snapshots && git commit -m "snapshot: ..."

This drops a dated copy of every docs/data/*.json into snapshots/<stamp>/ (visible as plain files on GitHub) and logs it in snapshots/README.md. Restore one file with:

cp snapshots/<stamp>/reddit_subscribers.json docs/data/

Because data also lives in git, the full history is in git log too — snapshots just make specific results easy to find and compare.

Roadmap

1. ✅ MVP — search + Reddit + culture, on GitHub Pages.
2. TikTok / Instagram hashtag view counts as a tracked series.
3. Resale market (eBay/Mercari sold listings for retired & limited bears) — the hardest loyalty proof: adults paying premiums on the secondary market.
4. Retail / international franchising + job-posting expansion pipeline.
5. Optional automation: a GitHub Actions cron that runs the collectors and commits.

Not investment advice.