There is a silent gap in most e-commerce operations: the team collects data, organizes it in a spreadsheet, saves it to Drive — and when it is time to decide, goes right back to gut feeling. Competitive intelligence only creates value when it completes the path signal → pattern → decision.
3.4×
is the margin gain on contested SKUs at companies that decide with structured data vs deciding on intuition.
McKinsey Global Institute, Analytics in Retail, 2023
You cannot manage what you do not measure. But even more dangerous is to measure and not decide.
The gap between collection and action
Three typical symptoms of an operation stuck at the collection stage:
- Competitor spreadsheet up to date, pricing decision improvised.
- Weekly meetings report “what happened”, they rarely decide what will happen.
- Data is presented by volume (crowded slides) but without an actionable recommendation.
The competitive decision pyramid
Signals (what shows up today)
A competitor launched a 30% coupon. An isolated signal, little useful information. Deciding on that signal alone is the equivalent of gambling.
Patterns (what repeats)
A competitor drops a public coupon every Tuesday at 9am for 4 weeks in a row. A clear weekly acquisition pattern. Here there is actionable information.
Decisions (what you do)
Two choices, both defensible:
- Pull forward: drop the offer on Monday, capturing demand before the competitor’s Tuesday.
- Counter-attack: a coupon on Tuesday itself, targeted at repeat customers (it preserves the competitor’s acquisition, shields your base).
The choice depends on context (margin, capacity, positioning), not on the isolated signal.
3 quick frameworks to use tomorrow
1. Depth vs frequency (the campaign’s mode)
Deep + rare
acquisition (calibrate a long-term response)
Deep + frequent
desperation or clearance (don’t copy blindly)
Shallow + frequent
retention (shield your repeat base)
Shallow + rare
maintenance (probably ignore)
2. Audience match (relevance of the threat)
Before reacting, ask: is the audience of the competitor’s campaign mine? Criteria:
- Income bracket compatible with your average ticket?
- Geography overlaps with your fast-delivery area?
- Demographic profile (age, gender) matches?
- Discovery channel is the same (organic Instagram vs Google Shopping)?
If fewer than 2 criteria match, ignore it. The competitor’s campaign does not threaten your revenue.
3. Reaction window (timing is everything)
< 6h
proactive reaction — captures ~80% of demand at risk
6-24h
tactical reaction — captures ~50% of demand at risk
24-48h
late reaction — only worth it if the SKU is central
> 48h
too late already, turn it into planning for the next wave
Commercial meeting: what to show and what to hide
Show in the meeting
• Signals with confidence ≥ 0.85
• Patterns mapped over ≥ 3 occurrences
• Recommendations with BLUF
• Year-over-year comparison (BF 2024 vs 2023)
• Estimated cost of inaction
Hide in the meeting
• Isolated signals with no pattern
• Subjectivity (“it felt aggressive”)
• Technical collection detail
• Data with no traceable source
• Crowded slides with no recommendation
Referências e leitura complementar
- McKinsey Global Institute (2023). Analytics in Retail — The Margin Premium. McKinsey & Company link .
- Deming, W. E. (1986). Out of the Crisis. MIT Press.
- Davenport, T. H. (2018). The AI Advantage: How to Put the Artificial Intelligence Revolution to Work. MIT Press.
- Conversion (2024). Brazilian E-commerce Yearbook. Conversion / B-Capital.
See your first competitor in minutes
14-day free trial, no card. Within minutes, the first detection shows up on your dashboard.
Create free account
