Retail product data: why your 4 systems don't talk to each other (and what it really costs)
Alexandre Point
April 30, 2026
Your ERP knows the prices. Your PIM knows the attributes. Your WMS knows the stock levels. And your crawling tool knows the competition. But none of these systems truly communicate with each other – and your entire organisation pays the bill every day, in lost hours and decisions made on incomplete foundations.
Table of contents
The myth of centralised data
Most retail organisations have, on paper, a clear picture of their product data architecture. There is a system for each dimension: purchase and sale prices in the ERP, product attributes in the PIM (which often also hosts images, or references them from the CMS), sales volumes in the data warehouse, competitive data in the monitoring tool. At the foundation of all this: the GTIN code, the global product identification standard that should, in theory, serve as the common key across all these systems.
What is missing is the binding layer. The layer that assembles these fragments into a single, coherent, up-to-date product record – by SKU, by market, by channel.
In the absence of this layer, the assembly happens elsewhere. In shared Excel files. In Python scripts maintained by a single person. In manual exports scheduled every Monday morning. It is fragile, slow, and does not scale. As we show in our analysis of pricing in food retail, organisations that have not solved this foundational problem find themselves unable to deploy their pricing strategies effectively.
The integration layer your systems are missing
Mercio is not one more system. It is the missing integration layer between your existing tools – one that ingests each source in its native format, applies your business rules, and produces a unified product record, synchronised continuously.
What concretely happens when data doesn't converge
Data silos are not an abstract problem. They have precise, daily, measurable operational consequences.
A multi-country product launch becomes an obstacle course
Imagine a new product to be listed simultaneously in the UK, Germany and Spain. The product attributes arrive from the supplier in one format. Country-specific purchase prices sit in the ERP, with different VAT rules. Images are in the PIM or need to be retrieved from the website and CMS, but some do not match the visuals validated for each market. And the regulatory requirements – legal notices, units of measurement – vary from country to country.
Without a unified product data pipeline, each local team reconstructs its own version of the truth. Consistency errors are only discovered once the product is live – sometimes after customers have already placed orders based on incorrect information.
The urgent price update arrives too late
A competitor drops prices in a sensitive category. Your pricing team wants to react within two hours. But to calculate the right response, they need to cross-reference the current selling price, the purchase cost, the minimum authorised margin, recent sales volume, and the validated competitor price – five dimensions from four different systems.
If this assembly is not automatic, it takes time. And in a market where prices move several times a day, that time has a direct cost: you react when the window of opportunity has already closed. As we analyse in our article on intelligent orchestration of dynamic pricing strategies, reactivity is only possible when the data is already ready at the moment of decision.
Performance analysis never draws on the right figures
When sales, margins and product attributes live in separate silos, any cross-cutting analysis requires a preparation phase. Before being able to answer "what are our 50 most profitable products in the beverages category in Germany this quarter?", someone has to extract, join, clean and reconcile data from multiple sources.
This work takes time – often several hours for a question that should be answered in seconds. And it is repeated with every new question, because the data has changed in the meantime.
Data already assembled when a decision needs to be made
Mercio continuously synchronises all dimensions of a product. When an analysis need or a competitive reaction emerges, the data is already assembled, validated and available – not in the process of being prepared.
The real cost of data silos: three angles for measurement
The cost in analyst time
In the retail organisations we work with, data teams spend between 30% and 50% of their time on assembly and reconciliation tasks – before any analysis even begins.
The cost in missed opportunities
Every pricing or merchandising decision made on an incomplete data basis is a suboptimal decision. Across a catalogue of 50,000 references, the effects accumulate silently.
The cost in human dependency
When the assembly relies on undocumented scripts, the organisation becomes dependent on a handful of individuals. One departure, one absence, and the data flow stops – or produces errors that no one detects.
This finding is corroborated by market trends: according to the Mordor Intelligence report on the PIM market, resolving product data silos is today the primary driver of investment in retail integration solutions, with a market growing at more than 18% per year.
What "well integrated" means in practice
A well-integrated product catalogue is not another data warehouse. It is an operational guarantee: for every product, in every market, all relevant dimensions – price, cost, margin, attributes, images, sales volume – are available in a single, up-to-date, reliable record.
This means that business rules – VAT by country, price floors and ceilings, local regulatory requirements – are applied automatically, not manually at every export. It means that when an attribute changes in the PIM, the change propagates without human intervention. It means that your pricing, merchandising and reporting tools all consume the same single source of truth.
| Dimension | Without unified integration | With the Mercio pipeline |
|---|---|---|
| Multi-country launch | Each local team reconstructs its own version of the data | A single product record per market, propagated automatically |
| Pricing reactivity | 2 to 4 hours of manual assembly before a decision is possible | Data already assembled – decision possible within minutes |
| Performance analysis | 1 hour to several days of preparation per cross-cutting question | Answer in seconds on up-to-date data |
| Pipeline maintenance | Fragile internal scripts, dependency on a handful of individuals | Infrastructure maintained by Mercio, scalable without intervention |
| Multi-market business rules | Applied manually at each export, a source of errors | Encoded once, applied automatically at every synchronisation |
A single source of truth for all your systems
Mercio produces a unified product record that all your tools – pricing, merchandising, reporting, e-commerce – consume directly. No more discrepancies between what the pricing team sees and what the website displays.
Why this is not an ordinary internal project
Many retail data teams have attempted to build this integration layer in-house. Some succeed – at the cost of a considerable engineering investment, documentation that must be continuously maintained as sources evolve, and ongoing vigilance to keep the data pipelines running when a supplier changes their export format.
The problem is not the competence of the teams. It is that this work is never finished. Data sources change format. Markets are added. Business rules evolve. Building and maintaining this retail data infrastructure in-house is a permanent commitment that diverts data teams away from their higher-value analytical work.
Technical debt accumulates. Scripts become unreadable. And the day the only engineer who understands the data pipeline moves on, the organisation discovers how much its decision-making capacity rested on a hidden fragility.
A maintained infrastructure, not a permanent burden
At Mercio, maintaining the data pipeline is our responsibility, not yours. When a source changes format, when a supplier updates their API, when a new market is added – our team absorbs the complexity. Your data teams remain focused on analysis and decision-making.
What Mercio builds concretely for you
Retail product data consolidation is at the heart of the Mercio infrastructure. For each client, we build a bespoke data pipeline that ingests your sources – whatever their formats, update frequencies, and market-specific characteristics – and transforms them into unified, enriched product records, synchronised continuously.
A single product record, always up to date
Mercio aggregates in real time the data from your ERP, PIM, WMS, CMS and competitor crawling tools. For every SKU, a complete profile – price, cost, margin, attributes, images, sales volume – in a single interface.
Business rules applied automatically
VAT by country, price floors and ceilings, local regulatory requirements, image source prioritisation – encoded once in Mercio, applied at every synchronisation. Zero manual exports.
Competitive reactivity that is finally operational
When a competitor moves their prices, all the dimensions needed to calculate the response are already assembled and validated. Your teams no longer wait for data – they decide.
An infrastructure that scales without effort
New market, new data source, 20,000 additional references – the Mercio pipeline absorbs the complexity without mobilising your data teams. Designed to grow with you.
Analyses that reflect reality, not an outdated copy
Because the Mercio pipeline is synchronised continuously, your catalogue performance analyses, price elasticity studies and category profitability assessments draw on data that reflects the actual state of your catalogue at the moment you consult it.
- The question "what are our 50 most profitable products in Germany this quarter?" is answered in seconds – not after two hours of preparation.
- Data anomalies are detected and flagged automatically, before they impact your decisions.
- Every team – pricing, merchandising, data – consumes the same source of truth, at the same moment.
Mercio is built for you if...
- Your data teams spend more than 30% of their time reconciling exports before they can analyse
- Your pricing must account for different price, VAT or regulatory rules across markets
- Your data pipeline relies on scripts maintained by one or two people
- Your pricing teams cannot react to competition in under two hours
- A multi-country product launch mobilises several teams for several days
- Your profitability analyses are already out of date by the time they are finished
- You have tried to build this integration layer in-house and have not been able to sustain it
Your product data deserves better than manual exports
Let's talk about what Mercio can put in place for your organisation – whatever the number of markets, source systems or references in your catalogue.
Get in touch with MercioAlexandre Point
Alexandre leads our development teams to make Mercio's pricing application the best and most innovative on the retail market. He graduated from Ecole Polytechnique and Stanford. As a second time entrepreneur, his curiosity and pursuit of excellence drive relentless innovation for our retail clients.
Everything you need to know about product data consolidation and retail data pipelines.
What is the difference between a PIM and a unified product data pipeline?
A PIM (Product Information Management) system manages product attributes: descriptions, images, technical specifications. A unified product data pipeline goes further: it aggregates all dimensions of a product from multiple source systems (ERP, WMS, PIM, crawling tool) into a single record. The PIM is one source among many; the pipeline is the integration layer that assembles these sources and applies business rules automatically.
How much time do retail data teams lose because of data silos?
In the retail organisations we work with, data teams spend between 30% and 50% of their time on data assembly and reconciliation tasks — before even starting the analysis. These are experienced, costly profiles applied to tasks that could be entirely automated with a well-designed pipeline.
Why is building this data pipeline internally risky?
The problem is not the competence of internal data teams: it is that this work is never finished. Data sources change format, markets are added, business rules evolve. Building and maintaining this infrastructure creates permanent technical debt and dependency on a few individuals. One departure or absence is enough to block the data flow or generate undetected errors.
How does Mercio handle multi-country specifics (VAT, local regulations, images)?
Business rules specific to each market - VAT by country, floor and ceiling prices, local regulatory constraints, image source priority - are encoded once in Mercio and applied automatically at each data synchronisation. When an attribute changes in the PIM, the update propagates without human intervention to all production systems, in all local versions.
Can Mercio absorb new markets or data sources without mobilising data teams?
Yes. The Mercio architecture is designed to scale without additional effort from your teams. Whether you add a market, a new data source or 20,000 references to your catalogue, the retail data pipeline absorbs the complexity automatically. Data teams are freed for higher-value analytical work - not mobilised for every infrastructure change.
