RETAIL & FAST-MOVING CONSUMER GOODS
AI and data at the heart of retailers’ challenges
Our expertise
Reinventing business models in the face of a persistent crisis and changing consumption patterns
Retail in the broad sense is heavily influenced by technology and data, from the creation of the offering to its distribution, including procurement, performance management and CSR impact.
Distributors operate in an environment where AI is profoundly transforming their businesses. Optimising product ranges, ensuring price consistency, personalising the e-commerce client experience and managing stock now require robust approaches that can reconcile operational performance with client satisfaction. Beyond the relationship with the consumer, the challenges affect the entire chain: automation of low value-added tasks, sales forecasting for ultra-fresh products, warehouse maintenance, traceability via blockchain, and the development of retail media. In a rapidly changing world fraught with uncertainty, data and AI give you the agility you need to build robust business models.
OUR OBJECTIVES
Build your offering around your clients
Boost the performance of your product range by identifying growth drivers, maximising margins and building in-depth knowledge of your ultra-segmented consumers.
AI Pricing and Revenue Growth Management
Optimise and manage your prices based on demand sensitivity and your strategic objectives. Integrate AI and predictive modelling into pricing and commercial policies for increased revenue and margins.
Invest resources more effectively and commit to CSR
Support stakeholders in directing their resources towards what matters most: streamlining the supply chain, strengthening positive impact and tangibly measuring their commitment to responsibility. Make sustainability a factor in competitiveness.
Build a new asset from your data
You have a wealth of data on your clients and their purchasing behaviour. Leveraging and utilising this data must be part of a strategic roadmap to extract maximum value and create a new asset.
Sales forecasting
Effectively anticipate demand using advanced predictive models to confidently adjust your inventory, pricing and marketing campaigns.
Creating value by combining your data, our modelling and AI
From mass market distribution to specialised B2B distribution, Veltys has been supporting major retailers in a sector facing price pressure and changing consumer habits for over ten years.
Using data intelligence, we identify performance levers, improve commercial management, detect additional business opportunities and build client segmentation. Our interventions are results-oriented and accompanied by relevant success indicators.
OUR BUSINESS CASES
Optimising catalogue targeting to save 2–5% in costs
Context
For a mail order company, visibility and client activation are essential. Historically, the company relied heavily on sending paper catalogues, a medium particularly appreciated by an older clientele. But with the rise of digital channels, it became strategic to evaluate the relevance of mass catalogue use.
The difficulty: the multiplicity of activation channels makes it complex to identify the real impact of each lever. In this context, any decision had to remain profitable, with a tolerated margin of error of less than 3%.
Method
The approach combined a detailed analysis of client profiles (purchasing habits, frequency, amounts, products, preferred channels, age, geography, length of relationship) with a study of promotional campaigns (catalogues and emails) over different periods.
Veltys modelled the precise impact of each catalogue on orders and distinguished between the most responsive and least responsive clients. Activation thresholds were optimised to maximise margins, and windows that were not conducive to sending catalogues were identified.
Finally, an experiment was set up to validate the recommendations before large-scale deployment.
Results
The experiment confirmed the robustness of the recommendations: removing catalogues resulted in a decrease of less than 2 points in orders per segment. The company was able to identify 2 to 5% of mailings to be removed, depending on the strategy adopted, and significantly improve profitability while maintaining commercial efficiency.
This optimisation illustrates how rigorous analysis and an econometric approach can align cost reduction with maintaining client performance.
Segmenting the clients of a B2B distributor to tailor pricing to individual profiles
Context
To boost margin growth, the management of a major B2B distributor wanted to regain control of its pricing practices, which had previously been largely left to sales representatives in a highly decentralised organisation.
The challenge was significant: a catalogue of more than 1.5 million items and a client base of more than 300,000 with extremely diverse profiles, ranging from residential tradespeople to industrial manufacturers. Not all clients buy the same products, nor do they buy with the same regularity or price sensitivity. Even among clients who appear to be comparable, purchasing behaviour can vary greatly. Segmentation therefore became essential in order to offer accurate pricing tailored to each client.
Method
The approach consisted of segmenting the client base according to economic criteria (annual expenditure, regularity, growth trend, distribution of purchases by product category) and behavioural criteria.
An agnostic AI analysis revealed purchasing patterns, which were then translated into simple indicators of expenditure concentration. These results were challenged by business experts to ensure their relevance and consistency with the reality of client activities.
Finally, a cross-referencing of statistical analyses and business rules made it possible to refine the segmentation and identify the most representative consumption profiles.
Results
The analysis led to the creation of around twenty distinct client segments based on their purchasing balance and dynamics. Algorithms made it possible to distinguish between ‘specialist’ clients and ‘cherry-pickers’, who partially source their supplies from competitors.
Thanks to this segmentation, the company now has a solid basis for applying differentiated and relevant pricing, aligned with actual client behaviour. This approach paves the way for better margin control while strengthening commercial competitiveness.
Structuring data governance in a decentralised retail organisation
Veltys supported the Chief Digital Officer in implementing governance tailored to her regionalised model in order to unlock the potential of data and AI initiatives across the business lines.
Context
The group operates under a highly decentralised model, with autonomy being part of its DNA, both at the point of sale and in regional entities.
Any data governance approach must take this complex organisation into account to ensure the alignment and transformation of data and AI uses. The challenge is to create a structured framework without stifling local initiative.
Method
We have structured governance based on accountability, with defined roles and community leadership.
Operational processes have been designed around data quality: existing documentation, interviews, IT data feedback, multi-stakeholder redesign workshops.
Actions have been focused on a concrete objective: data quality and its real impact on business lines. Structuring tools such as a project prioritisation matrix and quality monitoring dashboards were custom-designed.
Results
The quality dynamic has been launched, piloted and connected to business challenges.
Roles have been defined and integrated into a structured data programme that gives meaning and creates buy-in. Business issues have been addressed, with immediate fixes and sustainable solutions as part of the redesign and migration of data tools. The programme is grounded in reality, with concrete benefits for team performance.
Structure purchasing and store layouts based on actual demand to boost margins and sales outside of sales periods.
A retailer wants to streamline the business model for its own brands by purchasing and stocking each item according to its sales potential. Veltys conducts a detailed analysis of demand to predict potential and de-average purchase volumes.
Context
The retailer noticed a decline in sales and margins, offset by an increased reliance on sales. Purchase volumes were too uniform: the best products were missing from the shelves, while the least successful ones were piling up in stock.
The situation was complicated by a reduction in purchasing credits, more difficult sourcing, MoQ constraints and the lack of reliable management tools. Manual processes, which were prone to errors, did not allow the wealth of available data to be exploited. The actual heterogeneity of sales per item was not reflected in purchases.
Method
We analysed the performance and contribution of each product characteristic (colour, type, material, location) using an econometric model.
This model, fed by several seasons of sales data, enabled us to predict the expected sales volumes for each item, both during and outside of sales periods. The teams enriched the approach with their fashion and aesthetic expertise, which was integrated directly into the tool.
On this basis, we formulated precise recommendations for purchase volumes, calendar placement and store placement per item.
A customised collaborative tool was designed, linked to other tracking files (collection plan, etc.), enabling buyers to make confident decisions and reduce data entry errors.
Results
For six seasons, the brand has relied on this system for its purchasing. The results are tangible:
- +10 points in sales rate,
- +5% in turnover,
- +10 points in margin generated outside of sales.
The collaborative tool is used by all purchasing teams, facilitating work with product teams and offering significant time savings. The predictions are robust and integrated into the reality of the business: they combine the power of models with field expertise, enabling each item to be placed as close as possible to demand, at the right time and in the right place.
Discover our other business case studies
THEY PLACE THEIR TRUST IN US
OUR OFFERS
Discover all our offers
Our sectors
Discover our other sectors
Our insights
Analysis, insights, and studies to help you understand the major changes underway and provide you with Veltys’ perspective.
Our publications are designed to help you take action: in-depth analysis of key issues, real-world case studies, and proven methods. What they all have in common is content that helps you make informed decisions and drive lasting change.
PhD careers beyond academia
En savoir plus
Leave us a message
Have a question or comment?
Feel free to contact us any way you like.
The Offices
32 rue de Ponthieu
75008 Paris
Métro :
Franklin Roosvelt
Saint Philippe du Roule
Station Vélib’ :
8050 Boétie – Ponthieu
8039 Colisée – Champs-Elysées