Heritage Coins
From customer data in two systems to one environment for dashboards and customer selections.
The question
Which customers should we approach this month, through which channel, and who should we leave alone?
Starting point
Heritage Coins sells collectable coins in the United Kingdom and runs its own contact centre that approaches customers by phone. The team wanted to select customers on buying behaviour, products and payment position, and to build a prioritised list for each campaign. Customer data sat in two systems: the Odoo ERP system and the previous system. A complete view of the customer needed both sources.
Our approach
We started with a secure portal in the team's own language, with invitation-only login, two-factor authentication and permissions per role. We then quickly analysed the data sources and built the data models, set up to fit the selections the team makes later. Next we migrated the customer and purchase data from both systems into this new system and built a connection to Odoo that delivers the latest data every night. On that data we built two dashboards: the CEO Cockpit for turnover, returns and top products, and the Listmanager, which shows per month and channel which customers are on no list yet and which selections the team could run next. The selection tool turns those insights into a list. An employee describes who should be on it in plain language or by voice, and AI converts that into the tool's filters. Customers who may not be approached through the chosen channel stay off the list automatically.
How it works
You see what happens at every step.
- 01automated
Analysing, migrating and connecting the data
We analyse the data sources and build data models that fit the selections. Customer and purchase data from both systems then moves into the new system. After that, customers, products, invoices, returns and consents come from Odoo through the connection. On the Odoo side the connection only reads.
Role of AI, automation and people: The connection runs automatically every night.
- 02automated
Grouping customers
Each customer is scored on how recently, how often and how much they bought, which places them in a segment.
Role of AI, automation and people: Calculated automatically on fixed rules after each full sync.
- 03automated
Spotting opportunities
The CEO Cockpit shows turnover and returns. The Listmanager shows by segment, value and channel who has not been selected this month, with suggestions for new selections. From the dashboard, the team turns a group of customers straight into a selection.
Role of AI, automation and people: The figures and suggestions come from the data automatically. The team picks which suggestions to pursue.
- 04AI + person
Building a selection
An employee types or speaks who should be on the list, for example by product, spend, last purchase or segment.
Role of AI, automation and people: AI converts the request into filters and can only choose existing filter values. On request, AI summarises the result.
- 05automated
Cleaning the list
Non-contactable customers, excluded customer groups and customers on other lists are removed. Changes can be traced per selection.
Role of AI, automation and people: Contactability per channel is checked automatically. The employee sets exclusions and deduplication.
- 06person
Putting the list to work
The list gets a campaign date and channel and goes to execution as an export. It holds customer numbers, not names.
Role of AI, automation and people: The export date is recorded automatically. An active or exported list closes to further additions.
Result
Customer data from two systems now sits in one secure environment, updated through the Odoo connection every night. The portal gives insight into all migrated data and turns it into action. Management sees turnover, returns and top products in one overview. The contact centre sees by segment, value and channel who has not been selected yet and builds its own selections and lists from the same screen. Anyone who may not be approached stays out of the selection automatically.
How we keep building
The portal grows with daily use. The next step is adding call outcomes once they are available, so selections, campaigns and realised sales can be followed in one line.
