Retail / distribution · footwear & fashion
A large customer base, part of which had quietly gone dormant. We wake the right customers at the right moment, lift the average basket and cut acquisition cost, against aggressive competition.
Visit siteThey move forward with BouclR.
Retailers and brands running the loop on their own priorities. Our pilot clients.
Chocolate · boutique network
Retention and average basket sustained through peak seasons, B2B prospecting via the dedicated agent, and lighter logistics so teams stay close to customers.
Visit site
Hospitality & dining
A digital and online-sales overhaul, and winning business clients (seminars, groups) through the prospecting agent. KPI steering lets them adjust the strategy earlier, on business data rather than gut feeling.
Visit site
Electric mobility · e-commerce
A complete, multilingual ecosystem: B2C and B2B acquisition, retention, and end-to-end reading of the customer journey. Every decision rests on reality, not guesswork.
Visit siteOther situations you will recognise.
Scenarios built from our field experience, to show the method and orders of magnitude. They do not reproduce any client’s results, but they draw on them.
B2B distribution · technical supplies and equipment
A multi-branch technical distributor
€22m in revenue, 6 branches, 28 sales people between the counter and the field, a B2B e-commerce just starting. A legacy business ERP, separate accounting, a CRM installed but barely fed, consolidated margin on a spreadsheet at month end. The company is growing, its systems do not talk to each other, and decisions are made faster than the figures come back.
The diagnosis
Three gaps stand out that the monthly reporting did not show. Margin dispersion: 19% achieved by one salesperson against a 26% average, the same product sold at 4 different prices, and 28 customers on a volume discount they no longer reach. DSO at 67 days, or €1.5m of working capital tied up. And 4 quotes out of 10 with no follow-up and no answer.
The structuring
A pricing grid by product family, with defensible margin floors. The arguments that let the team hold them in meetings. The ideal customer profile, to steer prospecting toward high-potential accounts. And the mapping of the data to connect.
The setup
Six agents switched on in the order in which they complete each other: KPI steering first, then margin defence, collections, quote follow-up, prospecting and customer retention. We connect the ERP, accounting, the CRM and the mailbox. Start in assisted mode: the agent proposes, the salesperson approves.
The run
Every week the indicators are consolidated and the gaps raised. Before each meeting, the salesperson has the real margin on that customer and the price to defend. Pending quotes are followed up with their history, receivables are worked account by account. Above all, the system learns: the price elasticity of each customer, the reminder that triggers payment, the quotes that convert. DSO falls, margin recovers, and each quarter the proposals gain in precision.
Industry · manufacturing with direct installation
A manufacturer that installs
€18m in revenue, a target of €23m, installation crews in the field, continuous quoting and a portfolio of maintenance contracts. A business ERP, separate accounting, monthly reporting. Management wants to catch margin drift before it settles in.
The diagnosis
The overall reporting was hiding two gaps. Jobs involving masonry rework take 1.5 times the quoted time, and they are the ones absorbing the margin. 50% of maintenance contracts are not renewed, for lack of being tracked at their renewal date. Site tracking happens by phone, with no usable data coming back.
The structuring
A quote built on value rather than cost, with margin floors and the arguments to hold them. Then the structuring of the site data, so that it finally comes back: without it, no correction is possible.
The setup
Five agents switched on: KPI steering, quoting, site tracking, margin defence, after-sales and contracts. We connect the business ERP, with a connector dedicated to quoting. The installers get a mobile app. Quoting starts in assisted mode.
The run
The installer closes the job from the app. The system compares actual duration against quoted duration, and the next quote carries the observed gap. Maintenance contracts are followed up 90 days before their renewal date. Before each meeting, margin is argued product by product. The system learns real durations by type of work and by crew: margin stops drifting, the renewal rate climbs back, and quoting corrects itself job after job.
Software publishing · established B2B subscription
A business software publisher
€8m in revenue, 180 client accounts, a solid development team but no commercial engine. Prospecting is still done by hand, and 14% churn is discovered at renewal, too late to act. Usage data is rich and unused.
The diagnosis
Churn is endured for lack of being anticipated. Some accounts have outgrown their scope and are ripe for an upsell, without anyone offering them anything. Prospecting is neither targeted nor sequenced. The diagnosis puts a figure on avoidable churn and on the upsell potential in the installed base.
The structuring
Here the contribution is not technical, it is commercial. Ideal customer profile, prospecting sequences, framing of the offer, and structuring of the usage data so that it serves. A go-to-market method, brought to a team that already knows how to build its product.
The setup
Six agents: usage tracking and renewal first, on the installed base, then intent signals, qualification, prospecting and demand generation. We connect product data, subscription billing and the CRM. Prospecting and follow-up start in assisted mode.
The run
Signals say when to approach a prospect, rather than at random. Usage tracking spots the accounts slipping away 4 months before renewal, while there is still time to act. Accounts ripe for an upsell surface on their own. The system learns the departure signals specific to this publisher: churn recedes, NRR climbs back, and the pipeline fills with qualified business.
Retail · network of specialist stores and online sales
A multi-site retail brand
8 stores and an e-commerce, €12m in revenue, a loyal customer base part of which no longer comes back, aggressive competition. Point of sale, e-commerce and marketing each run separately, and consolidation happens on a spreadsheet.
The diagnosis
The diagnosis measures the share of inactive customers and their recoverable value, the acquisition cost climbing for lack of reactivating the existing base, and the average basket within reach. It puts a figure on the revenue lying latent in the installed base.
The structuring
Segmentation of the base, a calendar of key moments by store and by channel, and the building of the reactivation narrative. Without it, campaigns go out on guesswork.
The setup
Four agents: KPI steering, customer retention, active loyalty, production of the visuals. We connect the point of sale, the e-commerce and the marketing tool. Every campaign is approved before it goes out.
The run
Customers to reactivate surface one by one, at the moment they still count. The key moments run to the calendar, the visuals are produced, the results measured. The system learns the drop-off signals specific to the brand and refines its win-backs season after season: the average basket rises, acquisition cost falls, and the installed base becomes active again.
Frequently asked questions
What we get asked about proof.
We publish no typical timeframe, because it depends on the state of your data and on the issue you pick. What is constant is the order: measurement is installed before action, so the effect reads on a precise point rather than on a general impression.
Retail and distribution, chocolate and a boutique network, hospitality and restaurants, mobility and e-commerce. These companies are our pilot clients; their detailed results and testimonials are coming.
The question is not size but data. If your activity already leaves usable traces · orders, customers, stock, exchanges · there is enough to run a loop. A small business with clean data often moves faster than a large group whose systems do not talk to each other.
Yes, when the client concerned agrees and their case resembles yours. Describe your context: we will tell you plainly whether we have a comparable reference, or whether we do not have one yet.
We see it, and we say so. That is exactly why measurement is installed before action: a system you cannot judge is a system you keep out of habit. An agent that produces nothing is stopped or rebuilt.
See what the loop can transform at your company.
