ai-ecommerce
0/15 complete

Module 4: Pricing Intelligence · 25 min

Building a Competitor Price Tracker With Sheets and AI

// sabak

Turn this lesson into one checked practice output

By the end, you should be able to explain the core idea behind “Building a Competitor Price Tracker With Sheets and AI” in your own words, apply it to one small real or sample task, and identify what still needs human review.

  1. 1

    Learn

    Read the 25-minute lesson without copying an output blindly.

  2. 2

    Try

    Use a small, non-sensitive example that you can inspect line by line.

  3. 3

    Review

    Check facts, fit, and risk; save one improvement note for next time.

A price tracker should explain comparable offers over time, not encourage automatic undercutting. Prices differ because bundles, variants, delivery, warranty, seller reputation and campaign terms differ. Normalize those facts before asking AI for a pattern.

Capture Comparable Observations

Use one row per dated observation:

capture_date | listing_id | url | seller | sku_equivalent | quantity
variant | displayed_price | shipping | voucher_condition | normalized_unit_price
placement | stock_note | warranty | notes | checked_by

Keep formulas transparent. normalized_unit_price = (displayed price + unavoidable shipping - universally applicable discount) / quantity. Conditional vouchers stay in their own field; do not subtract a discount most buyers cannot use.

Google Sheets can flag changes and create pivot views. AI can summarize supplied rows, but require listing IDs and dates for every observation. Never ask it to scrape accounts or bypass platform controls.

Worked Example

Three sellers appear to price storage jars at PKR 1,399, 1,599 and 1,850. Normalization shows the first is a four-pack, the second a six-pack with shipping, and the third a six-pack with thicker material and warranty language. The cheapest displayed price is not the cheapest equivalent unit.

After four weekly captures, the tracker shows campaign discounts that reverse afterward. The business decides not to chase each change. It sets a sustainable floor, monitors position, and tests whether clearer quantity and material information supports its price.

Turn the Tracker Into a Decision Log

Add decision, reason, approved_by, effective_date and review_date. Most observations should produce “no change.” If you do change price, write the hypothesis first—for example, clearer bundle value at the current price versus a lower price—and keep a baseline. The tracker then explains why a decision happened instead of becoming a colorful archive.

Failure Cases to Diagnose

  • Tracking screenshots without dates or URLs.
  • Comparing different quantities or specifications.
  • Treating conditional vouchers as universal prices.
  • Auto-lowering price whenever a competitor changes.
  • Storing competitor or platform credentials in a script or sheet.

🇵🇰 Pakistan Angle

Separate imported-input price changes, currency effects, campaign subsidies and city-specific delivery. A sustainable Pakistani seller may need a higher price than an informal competitor who has omitted packaging, tax, returns or warranty costs.

Hands-On Exercise

  1. Select five genuinely comparable listings.
  2. Capture them on three dates.
  3. Normalize bundle and unavoidable shipping differences.
  4. Ask AI for row-cited patterns and verify each.
  5. Set a pricing review rule that does not auto-undercut.

Completion Rubric

  • Every price has URL, date and offer conditions.
  • Quantities and specifications are normalized.
  • Conditional discounts remain visible.
  • AI output cites supplied rows.
  • Price decisions respect a documented floor.

Sources

Key takeaway: Track comparable value over time; do not automate a race to the lowest displayed number.

Self-check

Before you mark Lesson 4.1 complete

  • Can I explain “Building a Competitor Price Tracker With Sheets and AI” without reading the lesson back word for word?
  • Did I complete the lesson’s practice step on a real or clearly labelled sample task?
  • Did I check the result for invented facts, private data, unsafe actions, and mismatch with the brief?