How a Pricing Team Recovered 8 Margin Points on 340 SKUs — Without a Single Manual Reprice

A mid-sized ecommerce business was repricing manually. Every week, a pricing analyst pulled competitor data by hand — visiting sites, copying prices into a spreadsheet, making judgment calls on which SKUs needed adjusting. It worked, in the sense that prices got updated. But it was slow, incomplete, and by the time the data reached the pricing team it was already between 5 and 9 days old.

Within 60 days of switching to an automated pipeline, they identified 340 SKUs where they were underpriced by more than 15% versus the market — and raised margins on that cohort by an average of 8 points without losing volume.

This is how that happened.

The Problem with Weekly Manual Repricing

The analyst wasn't doing anything wrong. The process was just structurally broken for the speed at which pricing moves in competitive ecommerce.

In fast-moving categories, competitor prices can shift multiple times per day. A manual process that produces a weekly snapshot is not measuring the market — it is measuring where the market was a week ago. By the time the pricing team acts on that data, competitors have already moved again.

The second problem was coverage. Manual repricing at scale is impossible. The team was working across a catalogue of over 3,000 SKUs. In practice, the analyst was covering maybe 200–300 high-priority lines each week and sampling the rest. The majority of the catalogue was effectively unmanaged from a competitive pricing standpoint.

The third problem was invisible: the team didn't know what they didn't know. They had no systematic way to identify which SKUs were underpriced. They could flag items that looked obviously wrong, but a 15% gap is easy to miss when you're manually comparing prices one row at a time.

What the Automated Pipeline Looked Like

We built a competitor pricing pipeline that scraped pricing data across all 3,000 SKUs daily from relevant competitor sites and marketplace listings. The pipeline ran overnight so the pricing team arrived each morning to a dashboard showing:

• Current price vs. competitor median and lowest price for every SKU
• SKUs where their price was more than 10% above or below the market
• 7-day price trend per SKU and per category
• New SKUs added by competitors that week (new competitive threats)

The dashboard was built into their existing BI tool — no new software to learn, no separate login. The data landed in the same environment their pricing team already used every day.

The technical stack: Python with Playwright for JavaScript-rendered competitor pages, rotating residential proxies for anti-bot mitigation, a nightly orchestration job with retry logic and completeness checks, and a PostgreSQL table that fed directly into their BI connector. Field coverage ran at 98.4% across the full 3,000 SKU catalogue.

What the 60-Day Analysis Revealed

The first month was diagnostic. Before changing any prices, the team used the new daily data to audit their existing pricing across the full catalogue. That audit produced the finding that changed everything: 340 SKUs where they were underpriced by more than 15% relative to the current market median.

These were not obscure items. Many were mid-tier volume lines — products moving regularly, with healthy demand — where the team had simply not had the data resolution to see they were leaving margin on the table. At 7 days of data latency, price gaps of that size are invisible. At 24-hour resolution, they are obvious.

The pricing team raised prices on that 340-SKU cohort gradually over four weeks, testing in tranches to confirm volume held. It did. The average margin improvement on that cohort was 8 points. No marketing changes, no promotions, no product changes — just correcting prices to where the market already was.

Why This Pattern Keeps Repeating

This is not an unusual outcome. The margin recovery pattern shows up in almost every engagement where a team moves from manual or infrequent pricing to daily automated competitor monitoring. The reason is simple: the longer your data latency, the more your prices drift from the market without anyone noticing.

Ecommerce pricing is not a set-and-forget problem. Competitor pricing strategies change constantly — promotional cycles, stock availability, new entrants, algorithm-driven repricing on marketplace platforms. A team that checks weekly is flying on instruments that update once a week in a market that updates every hour.

The margin recovery is just what becomes visible when you finally turn on the lights.

The Build vs. Buy Question

This team initially explored building the pipeline in-house. The estimate from their engineering team was 8–10 weeks to build, test, and deploy a scraper covering their full SKU catalogue, plus ongoing maintenance as competitor sites updated their layouts.

The managed pipeline from JustMetrically went live in 6 business days. Maintenance — parser updates when sites change, proxy pool management, monitoring and alerting — is handled as part of the service. The engineering team's time stayed on their roadmap rather than on scraper maintenance.

The ROI on the data cost was recovered in the first month from the margin improvement on a single SKU cohort. The pipeline has been running daily since.

What to Look for in Your Own Catalogue

If your team is repricing manually or on a weekly cycle, the diagnostic questions are:

• What is your current data latency — how old is the competitor pricing data your team acts on?
• What percentage of your SKU catalogue is actively monitored versus sampled or ignored?
• Do you have a systematic way to identify underpriced SKUs, or are you catching them ad hoc?
• How much analyst time per week goes into pricing data collection versus pricing decisions?

If the answer to any of these is "we don't know" or "more than a few days", you likely have margin gaps in your catalogue that daily data would make visible within 30–60 days.

Getting Started

A pricing intelligence pipeline scoped to your catalogue typically goes live within one to two weeks. We handle target site selection, parser build, proxy infrastructure, quality monitoring, and delivery format — structured data into whatever surface your pricing team already uses.

If you want to run the same diagnostic audit this team ran — identify underpriced SKUs before changing anything — that is where every engagement starts. The data tells the story. You just need the resolution to read it.

Reach us at info@justmetrically.com or explore our ecommerce data scraping service and managed pipelines. We scope new projects within one business day.

#EcommerceData #PricingIntelligence #CompetitorPricing #WebScraping #DataPipelines #EcommercePricing #PriceOptimization #ManagedData #RetailTech #EcommerceAnalytics