Boulangerie Meva, Antananarivo: -30% Stock Loss in 3 Months

Meva Bakery in Antananarivo — Fournil case study

Context

Boulangerie Meva is an artisan bakery located in the Analakely neighbourhood, in the heart of Antananarivo, Madagascar's capital and most commercially active city. Founded in 2019 by Nirina Rakoto, the bakery employs 6 people full-time and produces a daily range of breads, mofo gasy (traditional Malagasy fried rice cakes), viennoiseries, pastries and celebration cakes. The Analakely market district sits at the commercial centre of the city, providing consistent foot traffic from market visitors, office workers and local families. In 4 years of operation, Meva had built a strong local reputation and a loyal customer base that extended beyond the immediate neighbourhood to include office workers and families from several surrounding districts, attracted by the quality of both the traditional products and Nirina's own specialities.

By late 2023, the bakery was facing a problem of success: growing demand — including an increasing volume of catering orders for corporate events, weddings and family celebrations — was pushing the limits of what the existing management approach could handle. The 6-person team was managing stock, recipes, supplier relationships and customer orders through a combination of a handwritten notebook, a calculator and the owner's memory. This system had worked when the business was smaller; at its current scale, it was generating daily stress, unmeasured losses and an administrative burden that fell almost entirely on Nirina. In early 2024, Nirina was evaluating the possibility of opening a second location in the Ivandry district, which offers strong commercial foot traffic and a growing professional clientele willing to pay for quality artisan products. But the prospect of replicating the current operational approach across two sites — with all its inefficiencies and blind spots — was a serious concern. The bakery sector in Madagascar benefits from a positive economic outlook supported by <a href="https://www.worldbank.org/en/country/madagascar/overview" target="_blank" rel="noopener noreferrer">national development initiatives</a>, making it a good time to scale — but only for businesses with solid operational foundations.

Problem

Before adopting Fournil, the management of Boulangerie Meva relied entirely on manual processes with no data layer — a common situation for artisan bakeries that have grown organically without ever stopping to formalise their operational systems. Supplier orders were estimated from memory and rough observation rather than from any reliable record of what had been used and sold. The result was a stock loss rate of 25% — far above the 5 to 15% sector average identified by ADEME — driven primarily by two structural causes: overproduction of baked goods relative to actual daily demand, and the spoilage of perishable ingredients such as butter, eggs, fresh cream and seasonal fruit fillings that were ordered in quantities beyond what recipes and sales actually required.

Every Sunday, Nirina spent 6 to 8 hours on manual accounting: reconciling the week's till receipts by hand, updating stock figures by comparing physical counts against rough expected consumption, and attempting to calculate which products had been profitable. She had a general sense of the bakery's overall revenue but no reliable visibility into margins at the product level. She suspected that several items in the range were being sold below their true cost — a guess based on observation of ingredient prices and labour effort rather than precise calculation. Without a reliable cost price for each recipe, she had no way to identify which products to focus on, which to reprice, and which to remove. The idea of opening a second location with these unknowns unresolved felt like a significant financial risk. <a href="/blog/reduce-stock-loss">Stock loss at 25%</a> alone was enough to make the expansion economics very challenging.

Solution

In March 2024, Nirina adopted Fournil for point-of-sale management, stock tracking and business activity reporting. The implementation was structured over 5 days: 45 recipes imported with their full ingredient lists and current raw material costs, initial stock levels configured for all active ingredients, and the full 6-person team trained on their respective modules. The POS module was operational from day one, immediately beginning to capture accurate sales data for each product sold. The production manager could log output for each item after each batch, and the system began building the historical sales data that would later support data-driven production planning informed by actual daily demand rather than memory and observation.

Automatic stock decrement from each sale eliminated the need for daily manual stock counts on most ingredients: the system tracked consumption in real time based on sales and recipe formulations, applying the precise raw material cost of each ingredient against each unit sold. FIFO rotation protocols were implemented simultaneously across all cold and dry storage areas, with each delivery labelled with its receipt date. Spoilage alerts were configured to notify Nirina when any perishable ingredient — butter, fresh cream, eggs, seasonal fruit preparations — was approaching its use-by date with more than a threshold quantity still in stock, flagging overstock situations before they became waste rather than after. Within the first 2 weeks, the system had already surfaced several production patterns that had been invisible on paper: specific products overproduced on specific days of the week, and ingredient orders that were systematically 15 to 20% above what actual consumption required — a direct driver of the chronic spoilage losses Nirina had been experiencing.

Results

Within 3 months of full adoption, Boulangerie Meva's operational metrics had improved substantially across every dimension that had previously been a source of concern and loss. The team of 6 had moved from a management system built on memory and handwritten notes to a fully digital operation with real-time visibility across every product, ingredient and revenue figure. The stock loss rate dropped from 25% to 10% — a 15-percentage-point reduction that far exceeds the sector average improvement reported by ADEME for bakeries that adopt structured waste tracking. This was achieved through the combination of real-time spoilage alerts, FIFO rotation enforcement across all perishable storage, and production quantities progressively recalibrated against 6 weeks of accurate historical sales data. On typical artisan bakery revenues, a 15-point reduction in waste rate translates directly to improved net margin in a sector where margins average only 3 to 7%.

<a href="/blog/calculate-real-margins">Net margins</a> improved by 18% overall as Nirina used Fournil's recipe cost analysis to identify 8 products that had been sold at a loss — their selling prices had been set by market comparison without accounting for their actual raw material cost, labour and overhead allocation. Those products were repriced to reflect their true cost prices, with minimal customer resistance: the quality of Meva's products had always supported higher prices; the barrier had been Nirina's uncertainty about how to justify them. Several additional products were adjusted upward after markup coefficient analysis revealed they had been systematically underpriced relative to comparable offerings in the neighbourhood.

The administrative overhead that had consumed Nirina's Sundays dropped from 6 to 8 hours per week to under 1 hour, with automated reports handling the reconciliation, stock summary and margin analysis that previously required manual compilation. The team saves approximately 15 hours per week collectively on administrative tasks — stock counting, bookkeeping, order reconciliation — that are now handled automatically by the system. Building directly on these operational and financial results, Boulangerie Meva opened its second location in the Ivandry district in September 2024, managed centrally through Fournil with consolidated reporting across both sites from a single dashboard. The second location was profitable from its second month of operation. For a deeper understanding of the stock loss reduction methods that drove these results, read our guide on <a href="/blog/reduce-stock-loss">how to reduce stock loss by 15%</a>.

Before Fournil, I didn't even know which products were actually making me money. Now I manage two shops from my phone and I've got my Sundays back.

Nirina Rakoto, Founder, Boulangerie Meva