Competitive intelligence built in Brazil,
for Brazil.
We are Batedor — the competitor-monitoring platform born to solve the real problem of the Brazilian retailer: finding out too late.
The rider who scouts ahead.
In medieval warfare, the batedor (scout) was the rider who galloped ahead of the army. His mission: to look first. To report what the enemy was doing, which route they took, which trap they were setting.
With intelligence, the commander decided. Without it, he marched blind — and lost the battle.
In e-commerce, your competitors are that front line that shifts every day. New coupon, aggressive campaign, launch, price change. Whoever finds out late reacts in panic. Whoever knows first commands.
Batedor is your digital scout. It rides ahead, observes, reports.
You command.
Chapter I · Problem
What we solve.
The traditional process of monitoring competitors is structural — and structurally broken.
How it used to be (and still is in 90% of stores)
- 📋 Spreadsheet updated weekly with screenshots pasted by hand
- 📱 24h stories slip away because no one checks every day
- 💬 A customer tells you on WhatsApp that a competitor has a better coupon
- ⏱️ An analyst spends 4-8h a week collecting, capturing only ~35% of the signal
- 💸 Hidden cost: R$ 800-1,600/month in analyst hours
- 🔻 Margin burns from late reaction (typically 1.5-3 p.p. on the contested SKU)
- 📊 No structured history to compare BF 2024 vs 2025
How it becomes with Batedor
- 🤖 Automatic 24/7 monitoring with no dedicated analyst
- ⚡ Story captured within 15 min of publishing, kept for 90 days
- 🔔 Real-time alert via email / Slack / Teams
- 📈 16 commercial types classified by AI with 0-100% confidence
- 💰 Starter plan R$ 199/month (vs R$ 800+ in analyst hours)
- 📑 PDF reports ready for the sales meeting in under 2 min
- 🗄️ Structured history to compare periods and cycles
Chapter II · History
How we got here.
A timeline of Batedor's 18 months.
2024 · Q3
The pain that sparked the product
The founding team ran a DTC brand and kept learning about a competitor's promo from customers — always 24-48h too late. Spreadsheet + WhatsApp screenshots ate up 4-8h a week per analyst and captured only ~35% of the real signal.
2024 · Q4
Validation with 12 e-commerces
A Figma mockup + interviews with pricing managers. 11 of the 12 confirmed the pain is real and structural — it won't go away with yet another Looker report.
2025 · Q1
MVP in production
First 5 beta tenants: campaign detection across Instagram + Facebook + website, with local AI on open models. The very first week it captured a flash story that would have escaped manual monitoring.
2025 · Q2-Q3
Scale and calibration
Added classification across 16 commercial types, banner OCR, PDF reports, multi-tenant with RLS isolation. The base grew to 80+ stores in pilot tests.
2025 · Q4
A Black Friday on fire
First BF at scale: 3,200+ campaigns detected over the 5 peak days. A stackable-coupon pattern identified in 67% of stores (vs 38% in 2024). The learnings became a public playbook on the blog.
2026 · Q1
Public release
Platform opened with 3 plans (Starter / Professional / Enterprise) + a 14-day no-card trial. Google login, PagBank integration, e-invoicing, LGPD with an active DPO.
2026 · Today
Where we are
480+ Brazilian stores monitored, 28,400 campaigns classified. Near-term roadmap: TikTok (Q3), LinkedIn (Q4), REST API for Enterprise.
Chapter III · Mandate
Mission & Vision.
Accessible, continuous competitive intelligence
For any Brazilian e-commerce operation, from the small retailer to the manager of an established brand.
- Close the information gap between SMBs and large retailers
- Turn repetitive collection into an automated cycle
- Give analysts back hours for real cognitive work
To be the standard commercial radar of Brazilian e-commerce
Where every competitor decision begins, on any cycle (daily, weekly, seasonal).
- Coverage of 100% of national commercial dates (BF, Mother's, Father's, Children's, etc.)
- Support for 100% of relevant channels (IG, FB, YT, TikTok, LinkedIn, website)
- Adoption in 5,000 Brazilian stores by 2028
What drives us.
Six principles stamped on the newsroom wall — each with what it means and how we apply it day to day.
Informed decisions
Guesswork doesn't set price, coupon or campaign.
What it means
We believe every relevant commercial decision must be anchored in an observable signal — not in intuition, assumption or copying what the competitor does.
How we apply it
AI confidence on every detection (0-100%). Threshold ≥85% to trigger an alert; 60-85% requires manual review; <60% is a weak signal. Always traceable back to the primary evidence.
Speed matters
Whoever knows first, reacts first.
What it means
In competitive e-commerce, every hour of lag between a competitor publishing an offer and your reaction costs margin (typically 1.5-3 p.p. on the contested SKU, Bain 2023) or a lost sale.
How we apply it
Stories captured within 15 min of publishing. Manual re-analysis in real time. Alerts to email/Slack/Teams the minute we detect them. SLA 99.5% on Starter; 99.9% on Enterprise.
Data sovereignty
Your data is yours. Nothing goes to a third-party cloud.
What it means
The competitive intelligence you build up in Batedor is your own strategic asset. We don't believe sending it to OpenAI/Anthropic/Google in exchange for convenience is acceptable.
How we apply it
AI runs on our own infrastructure with open models (Qwen 2.5, vision models). Datacenter in Brazil. AES-256-GCM on sensitive secrets. Multi-tenant with row-level scoping in Prisma.
Public content only
Only what anyone sees without logging in. No gray zone.
What it means
Ethical competitive intelligence uses public sources only. We don't ask for a competitor's credentials, we don't bypass paywalls, we don't use unauthorized APIs to reach private content.
How we apply it
Crawlers identify themselves as the Batedor bot (transparent User-Agent). We respect robots.txt. No capture of closed profiles, private groups or content behind a login.
Responsible AI
Models run in our own environment, on Brazilian soil.
What it means
Every automatic classification carries a cognitive cost (a wrong decision becomes a wrong action). We don't treat AI as a black box — quantified confidence, traceable prompt, human review where it matters.
How we apply it
A confidence score on every detection. The prompt used kept per piece of evidence. Internal quality telemetry to detect drift. Quarterly audit of the thresholds.
Customer as partner
Feedback shapes the roadmap. A customer who suggested it became a feature.
What it means
We build the product with customers, not for them. We don't believe in the “hire us and see what we ship” model. The roadmap is an open discussion with the active base.
How we apply it
A direct channel to the product team at /ajuda (tickets). Roadmap visible to customers. Beta program for new features (TikTok, REST API). A reply within 4h during business hours.
V-07 · Flag
Brazil first.
Calibrated for our market, our currency, our dates. Coupon, BOGO, extended installments — the Brazilian commercial vocabulary is what we understand best.
Made in
🇧🇷 São Paulo
Chapter V · Differentiation
Why not use a global alternative?
SimilarWeb, Crayon, Brand24 are great — they just weren't built for Brazil.
| Dimension | BR lineBatedor | GlobalSimilarWeb / Crayon | ManualSpreadsheet + analyst |
|---|---|---|---|
| Focus | Brazilian e-commerce | Generic or US enterprise | Variable |
| Language and vocabulary | Native PT-BR (coupon, BOGO, free shipping, 12x installments) | Native EN, translated PT | Depends on the analyst |
| 24h story capture | Within 15 min, kept for 90 days | Limited or nonexistent | Manual = misses ~65% |
| AI processing | Local in Brazil (our own models) | OpenAI/Google/Anthropic (US) | — |
| Price (SMB) | R$ 199-499/month | US$ 500-5,000/month | R$ 800-1,600/month (analyst hours) |
| Support | Portuguese, 4h SLA | English, 24h+ | In-house |
| Native LGPD | Active DPO, documented legal basis | GDPR-only or adapted | Operational risk |
Comparisons based on public price and feature listings from the vendors cited (consulted in May/2026).
Chapter VI · Operation
How we operate.
Record 01
Focus on Brazilian e-commerce
Black Friday, Mother's Day, Father's Day, Children's Day, Back to School, Valentine's, Christmas — Brazilian dates, Brazilian coupons, Brazilian offer formats. The Batedor understands “12x interest-free” and “free shipping over R$ 199”.
Our own calendar, our own vocabulary, classification calibrated on 28.4k Brazilian campaigns. Global tools treat the BR market as a special case; for us it's the central case.
Record 02
Local AI, in Brazil
Models running on our own infrastructure, in a Brazilian datacenter. Nothing leaves our environment. If you manage an e-commerce, this matters: the commercial strategy that Batedor delivers to you doesn't become someone else's model training.
Open models (Qwen 2.5 14B for classification, open vision models for OCR) on servers we operate. It costs more than using an API — but it's what makes sense for a product that monitors sensitive competitive data.
Record 03
Open by default in communication
Public roadmap. Visible status page. A bug tracker with items flagged as known issue. When something breaks, we say it broke.
Direct communication with the customer, no “Customer Success Manager” filtering what you say. Every engineer answers tickets when the subject calls for it.
Record 04
Real support, in Portuguese
A 4h SLA during business hours for the Professional plan, 4h round-the-clock (8h on weekends) for Enterprise. No offshore queue, no first-tier chatbot, no “open a ticket on the portal”.
Every ticket opened at /ajuda sends an email to all super admins. A reply typically in hours, not days.
Chapter VII · Team
Who's behind it.
A lean, Brazilian, remote team, with a background in the problem the product solves.
Engineering
Platform, infra, AI
Veterans of Brazilian fintech and SaaS startups. They built dynamic-pricing systems, scrapers at scale and ML pipelines.
Product
Direction, UX, roadmap
They came from fashion e-commerce and marketplaces. They know what hurts because they've felt the pain.
Content & Marketing
Blog, tutorials, positioning
Economic journalists and market analysts. Every blog article comes from primary research + our own data.
Support & Success
Onboarding, tickets, retention
Experienced in Brazilian B2B SaaS. A reply in hours, not days. They speak the language of retail.
A deliberately lean team. We believe quality B2B SaaS comes from small teams in direct communication with the customer — not from layered departments. Open positions occasionally at /contato.
What we promise publicly.
Every commitment comes with the technical evidence that backs the promise.
Native LGPD
Promise
Active DPO, documented legal basis, deletion within 24h, predictable retention.
Evidence
See /privacidade for details. The DPO replies at dpo@batedor.com.br with a 5-business-day SLA.
Layered security
Promise
TLS 1.2/1.3 + HSTS preload, Argon2id on passwords, AES-256-GCM on secrets, multi-tenant with RLS isolation.
Evidence
Quarterly external pentest. Full AuditLog. Daily backup with RPO 24h / RTO 4h.
Infra on Brazilian soil
Promise
Datacenter in Brazil. AI models run locally. Data does not leave the country without explicit consent.
Evidence
PostgreSQL on an isolated docker network, with no host port. Cloudflare proxy with BR geolocation.
Transparent roadmap
Promise
What's coming in the next quarters is public. Customers see what's being built, in testing, in production.
Evidence
Q3/2026: TikTok + REST API. Q4/2026: LinkedIn + Microsoft SSO. Updated with every release.
Assertive communication
Promise
No dark patterns. One-click cancellation. No forced retention. Automatic e-invoicing.
Evidence
Cancellation tested: under 30 seconds from the dashboard. No call, no form. Full refund within 7 days.
Predictable uptime
Promise
Starter/Professional: target 99.5%. Enterprise: contractual SLA 99.9% with credits for breach.
Evidence
Public status page (coming soon). Notification of planned and unplanned downtime. Monthly recovery tests.
