Every experienced retailer knows that intuition gets you started, but data keeps you profitable. For Elena Torres, owner of Roots Market — a three-location specialty grocery in the Pacific Northwest — the turning point came when she looked at her year-end numbers and realized something troubling: her revenue had grown 14% over three years, but her gross margin had barely budged. She was working harder, selling more, and somehow ending up with less to show for it.

This is the story of how her team used sales analytics to find the margin they were leaving on the table — and what happened when they started acting on the data.

The Business Behind the Numbers

Roots Market carries approximately 2,400 SKUs across three locations, ranging from local produce and specialty pantry items to natural supplements and craft beverages. As with most specialty grocers, the product mix is as emotionally curated as it is strategically planned. Elena had built deep relationships with local vendors and prided herself on carrying unique items customers could not find elsewhere.

That commitment to curation, while admirable, had a cost that was not appearing anywhere on her radar.

When her accountant flagged that gross margin had slipped from 36% to 32% over three years, Elena initially attributed it to commodity price inflation. But a closer look revealed something more systemic: the product mix had ballooned, vendor terms had grown inconsistent, and reorder decisions were still being made the same way they had always been — by gut feel and walking the floor.

The average specialty retailer operates on gross margins between 30% and 42%, so Roots Market was not in crisis. But the trend line was moving in the wrong direction, and Elena knew it. "I had a sense we were carrying too much," she said. "I just did not know how much, or where."

Step One: Understanding What Was Actually Selling

The first step was not complex. With an analytics platform connected to her POS system, Elena's team ran a straightforward product performance report: revenue, units sold, gross margin contribution, and days of inventory on hand — organized by SKU and category.

What they found was striking.

The top 400 SKUs — about 17% of the product mix — accounted for 71% of total gross margin dollars. Meanwhile, roughly 680 SKUs had generated fewer than 10 transactions per location in the past 90 days. These slow movers were tying up capital, consuming shelf space, and being reordered on autopilot because "we have always carried it."

The analytics team segmented the full catalog into three tiers:

  • Core performers (A-tier): High velocity, strong margin contribution, reliable demand patterns
  • Seasonal and specialty (B-tier): Lower velocity but acceptable margins; meaningful for customer experience and differentiation
  • Dead weight (C-tier): Low velocity, thin or negative contribution after accounting for shrink and carrying costs

This tiering framework took less than a week to build and immediately gave Elena's team a shared language for product decisions that had previously been entirely subjective. For the first time, conversations about what to carry were about performance, not opinion.

Step Two: Fixing the Reorder Problem

The second discovery was about how purchase decisions were being triggered. Staff had been using a combination of visual shelf checks and gut feel to generate orders. For top-selling items, this occasionally led to stockouts on peak shopping days — Saturdays and the days before major holidays were particularly bad. For slow movers, it meant the opposite: product arriving to already-full shelves, inflating the weeks-on-hand metric and burying working capital in products that moved slowly.

By analyzing 18 months of historical sales data, the team identified consistent velocity patterns for each SKU — not just average weekly sales, but day-of-week and seasonal coefficients. A 12-pack of oat milk might move 6 units per week on average, but that average masked a real pattern: 9 units Thursday through Saturday, 1 to 2 units Sunday through Tuesday.

The team used this data to set dynamic reorder points — minimum inventory levels calibrated to actual velocity and supplier lead times, not rough estimates. The result was meaningful:

  • Stockouts on core A-tier items dropped by 38% in the first quarter after implementation
  • Excess inventory on slow-moving SKUs declined by 22% as the team stopped reflexively reordering C-tier products
  • Average days of inventory on hand fell from 31 days to 22 days, freeing up significant working capital without increasing stockout risk

The insight here was not sophisticated — it was simply using historical data to set better triggers than "the shelf looks low."

Step Three: Vendor Negotiations Backed by Data

The third piece — and arguably the highest-impact change — was using analytics to restructure vendor relationships.

Before this process, Roots Market's purchase orders were largely reactive. A vendor rep would visit, present new items, and leave with an order that looked a lot like the previous one, perhaps with a few additional SKUs added. Terms had been negotiated years ago and rarely revisited.

Analytics changed that dynamic entirely. For each vendor, Elena's team produced a vendor scorecard showing:

  • SKU sell-through rates versus the vendor's promised velocity projections
  • Return on shelf space, measured in margin dollars per linear foot
  • Shrink contribution relative to category average
  • Lead time reliability across the previous 12 months

Armed with this data, Elena went into vendor conversations with specifics rather than impressions. With her top three produce suppliers, she negotiated net-30 payment terms based on demonstrated sell-through reliability — the data proved her business was a dependable partner. With two specialty beverage vendors whose products had consistently underperformed, she renegotiated minimum order quantities downward and secured promotional cost support in exchange for continued shelf placement.

Across the vendor portfolio, the renegotiated terms and improved buying decisions added an estimated 2.5 margin points on their own — before any changes to the customer-facing product mix.

Step Four: SKU Rationalization — The Hard Decisions

The emotional piece of this transformation was SKU rationalization. Removing products from the shelf — especially items from local vendors or long-standing relationships — felt uncomfortable. Elena had carried some of these products for years. They were part of Roots Market's identity.

Having the data made the conversation more manageable, even when the decision was difficult.

The team developed a simple decision rule: any SKU with fewer than 6 transactions per month per location AND a margin contribution under $1.50 per unit would be placed on a 90-day performance plan — given time to prove itself through a promotional push or coordinated placement improvement. If it still did not meet the threshold after 90 days, it was discontinued.

Over two review cycles spanning approximately six months, Roots Market removed 312 SKUs from active inventory. Some were consolidated to single-location exclusives, which preserved vendor relationships while reducing carrying costs across the system. Others were discontinued entirely, with conversations handled respectfully and transparently.

The shelf space freed by C-tier removal was reallocated to expanding depth on A-tier performers — allowing the team to carry larger buffer stock on their most reliable sellers without increasing total inventory value. The payoff was double: fewer stockouts on high-demand items and a cleaner buying experience for customers.

The Results: 12 Months Later

The transformation did not happen overnight, but the compounding effect of multiple data-driven changes added up faster than Elena expected.

After 12 months of structured analytics work:

  • Gross margin improved from 32% to 40% — an 8-point gain that translated to approximately $180,000 in additional annual profit on essentially flat revenue
  • Inventory turns increased from 9.2x to 13.1x annually, freeing up roughly $65,000 in working capital that had previously been trapped in slow-moving stock
  • Stockout rate on core SKUs dropped 38%, improving customer satisfaction and reducing the invisible cost of lost sales
  • 312 underperforming SKUs discontinued, simplifying operations and reducing the cognitive load on purchasing staff
  • Vendor terms improved with 6 of 14 key suppliers, adding an estimated $42,000 in annualized margin benefit

"I always knew we needed to be more data-driven," Elena told us. "But I assumed that meant hiring someone expensive and waiting a year to see results. What surprised me was how quickly the data revealed things we should have known — and how simple some of the fixes turned out to be."

What Made This Work

Not every analytics implementation produces results this quickly. Several factors made Roots Market's transformation particularly effective.

The data was complete and clean. Elena's POS system had reliable transaction-level data going back 18 months. With clean data as the foundation, analysis was fast and trustworthy. Retailers with inconsistent data entry or fragmented POS systems face a harder road to these kinds of results.

The team was involved. Rather than handing down directives, Elena brought her purchasing staff into the process. They saw the SKU performance data, helped categorize the product mix, and owned the execution. Buy-in at the operations level was essential — data insights that staff do not trust will not change behavior.

Decisions were made incrementally. Rather than attempting a full SKU rationalization in one pass, the team worked category by category over 90-day cycles. This reduced risk, allowed for learning along the way, and gave vendor relationships time to adjust gracefully.

Analytics were connected to action. The real value was not the reports themselves — it was tying those insights directly to purchasing decisions, vendor conversations, and inventory targets. Data that lives in a spreadsheet and never drives a decision does not move margins.

The Bottom Line

Roots Market's story is representative of what is possible for small and mid-sized retailers who move from intuition-based decisions to data-driven ones. An 8-point margin improvement does not require a price increase or a new concept. It often requires simply looking clearly at what you already have — and making better decisions about what to keep, what to cut, and where to double down.

Key takeaways:

  • Product mix bloat is a silent margin killer — most specialty retailers carry 2 to 3 times more SKUs than their top-line revenue actually requires
  • Reorder logic based on historical velocity consistently outperforms gut feel, especially for preventing stockouts on fast-moving products
  • Vendor negotiations improve dramatically when you come with data — scorecards turn subjective conversations into specific, productive ones
  • SKU rationalization is emotionally difficult but operationally necessary — a 90-day probation model reduces conflict while maintaining performance standards

At Chapters Data, we help small and mid-sized retailers build the analytics foundation that makes decisions like these possible. Whether you are running a single location or managing multiple sites, clean data and structured analysis are the difference between working harder and working smarter — and between margins that drift and margins that grow.