Skip to the content.

EARLY ACCESS

Know what your competitors are promoting. Today, not next week.

py-leaflets turns the leaflets your category prints — every flyer, every catalog, every weekly circular — into a structured promotional database. Refreshed daily. Queryable in seconds. Built for distribution companies, brand managers, and retail teams who can't afford to learn about a competitor's price drop on Monday's sales report.

11–27%
of CPG revenue is spent on trade promotions — often the second-largest line on the P&L after COGS.
Promotion Optimization Institute
~70%
of trade promotion investments don't return their cost. Brands are flying blind on which ones are working.
Promotion Optimization Institute
1–2%
of revenue can drop straight to the bottom line just by reallocating spend away from underperforming promotions.
Industry benchmarks (POI)
$8B
spent annually on grocery feature ads in the US alone — almost equal to retailer net profit margin.
Stanford GSB

What py-leaflets delivers

01 — VISIBILITY

Every promotion, every retailer, in one place

Stop pulling PDFs out of inboxes and screenshotting rivals' websites. Every leaflet from every retailer in your category, captured automatically and stored in one queryable archive — with full historical depth.

02 — STRUCTURE

Promotions as data, not pictures

Vision-AI extraction turns each printed page into structured rows: product, brand, price, was-price, discount, validity dates, category. Plug straight into Power BI, Tableau, your data warehouse, or your forecasting models.

03 — SPEED

Decisions in hours, not weekly reports

New leaflets land within hours of publication. Set price-alert thresholds, monitor competitor cadence by category, and respond before your next sales-out report tells you you're losing share.


How the pipeline works

1. CAPTURE Aggregator sites, retailer microsites 2. ARCHIVE SQLite + image store, versioned 3. EXTRACT Claude vision AI, structured rows 4. NORMALIZE Brand & category tagging, dedupe 5. DELIVER DB / API / dashboards RAW STRUCTURED ACTIONABLE Idempotent. Resumable. Every stage backed by a SQL view that answers "what work is left?".

A 5-stage pipeline: from messy PDFs and image-heavy circulars to a queryable promotions table.


Built for the people who feel the pain

Distribution companies

See which retailer customers — and their competitors — are running which promotions, time replenishment, and negotiate trade terms with hard data instead of anecdotes.

CPG brand managers

Audit retailer compliance with agreed promotions. Benchmark promotional pressure by SKU, region, and season. Catch off-strategy pricing in days, not quarters.

Pricing & revenue teams

Feed competitor promotional data into pricing models, demand forecasts, and elasticity studies. Stop building models on stale weekly POS extracts.

Retail category managers

Watch how rival chains are merchandising your category. Spot category-wide pricing moves before they hit your shelves.


Why now

Where the trade-promotion dollar goes today ~70% never returns its cost ~30% drives results 0% 50% 100% Reallocating just the worst-performing slice can lift bottom-line revenue by 1–2%. Source: Promotion Optimization Institute, industry benchmarks The competitors who know your prices first are the ones rebalancing fastest.

The fastest way to stop wasting promotion spend is to know what’s working and what isn’t — yours and theirs. Leaflet data is the cheapest, freshest signal of competitor intent on the market. py-leaflets turns it into a dataset you can act on.

Read the full business case →


What’s available today

Self-hostable pipeline

Available Python package + CLI. Five-minute install, runs end-to-end against the first source. Your data, your infrastructure, your control.

One source live

Available wowdeals_ae — UAE retail aggregator covering major hypermarkets and pharmacies. Source-pluggable for new sites.

AI extraction

Available Claude Opus 4.7 vision with structured outputs and prompt caching. Configurable model and effort to control cost.

Hosted service

Pilot Q2 Managed ingestion + extraction with REST and webhook delivery. Bring your category, we run the pipeline.

Multi-region sources

On request MENA, Europe, APAC retail aggregators added by category demand. Tell us which sites matter to you.

Dashboards & alerts

Roadmap Built-in price-change alerts, promotional pressure dashboards, and category benchmarks.

Ready to see your category's promotion data?

Pilot a source, integrate with your stack, or just take a look at the architecture.

Explore on GitHub