Coinage
From unused customer data to a view of every campaign, journey and return.
The question
How do we turn the data we already have into insight we can use to judge our commercial performance and steer it?
Where they started
Coinage is a direct marketing organisation in collectibles, operating from Poland, Czechia and Slovakia. The company had a CRM system and other data sources, but could not put that information to use when making decisions. The performance indicators and reports needed to judge results were missing. The task was to make the available data accessible for analysis and turn it into information people can act on.
Our approach
We first mapped which data sources were available, with Azure as the central source, and worked out how they could be turned into usable data models and performance indicators. Most of the work sat in that foundation: the ETL process that pulls the data out of the source systems, processes it and makes it ready, and the data models resting on top. On top of that we built a secured online BI portal where employees analyse the data from different angles. Three weeks after the first idea the first employees signed in and saw campaign reports, sales reports and an overview of performance. From that point we kept building on what daily use called for: the full customer journey, analysis at product level, returns analysis, forecasts and prediction models, and AI help that lets employees ask questions about their own figures.
How it works
You see what happens at every step.
- 01person
Map the data sources
We take stock of which data is available, including what sits in the CRM system, and decide which sources matter for the reports.
Role of AI, automation and people: The data comes together in Azure, the central source underneath the portal.
- 02automated
Set up ETL and data models
We pull the data out of the source systems, process it and make it ready in data models built for fast analysis, with the performance indicators alongside. Most of the work sat here.
Role of AI, automation and people: The ETL process runs automatically, so every report rests on the same, current data and definitions.
- 03automated
Follow campaigns and sales
In campaign, sales and overview reports, employees judge which campaigns are performing and how sales are developing.
Role of AI, automation and people: The reports update automatically, so the team sees the current state without anyone stepping in.
- 04automated
Follow the customer journey
The portal shows how customers move through the database: when someone arrives, what their first purchase is and where they stand in the journey now. Also visible: which products customers buy and which products are bought together.
Role of AI, automation and people: The journey and the product relations are calculated from the data models and shown per customer group.
- 05AI + person
Investigate returns and outliers
Employees see which products drive returns and for what reasons, and can adjust range and communication accordingly. Questions like “why are returns high?” they ask straight in the portal.
Role of AI, automation and people: AI supports investigating and interpreting the data. The conclusion stays with the employee.
- 06AI + person
Look ahead with forecasts
Forecast reports and prediction models support judging expected developments, alongside the figures on results already booked.
Role of AI, automation and people: The prediction models work the expectation through; the team decides what to do with it.
Result
Three weeks after the first idea the first employees were working in the portal, with campaign, sales and overview reports. Coinage now sees quickly which campaigns perform and how customers develop in the database, from the first purchase to where they stand in the journey today. At product level it is visible what customers buy and which products are bought together. On returns the team knows which products come back often and why, which gives it something to work with in range and offer. Decisions that used to rest mainly on experience and instinct are now backed with figures, and follow-up questions are explored in the same place with AI help. More than twenty employees use the reporting toolkit, several times a day and across the countries Coinage operates in.
How we keep building
Coinage has one secured environment where the team analyses results already booked and judges expected developments. The portal keeps being extended: what comes up in daily use decides what we build next. Because the ETL process and the data models underneath are already there, a new question is usually a new report rather than a new project.
Read more
Want a dashboard like this for your organisation?
Our page on AI dashboards explains how we bring your data sources together, what you get and how quickly a first version is ready.
More work
Take a look at these too
- LinkedIn messageVariant B
Good morning {first name}, I noticed {company} is growing fast. May I ask you a quick question?
AI writes, per audienceActiveSidon Water Benelux
Benelux, Spain
From looking for new customers to a large European fitness chain as a client.
Sidon wanted steady contact with the people who buy its products, in the Benelux and Spain. On LinkedIn, owner Casper Pruijn’s profile grew from 200–300 to almost 8,000 connections, with lead lists from an ICP analysis and our own tool for the network building. We also built the mailing lists and wrote the campaigns per audience. One of their best customers came from the email marketing.
almost 8,000connections, up from 200–30020–25%reply rate on messages - SegmentsTestedPhysiotherapistsHotelsProperty developersRehabilitation centres
Sportuok
Baltic states
From selling through the webshop to partnership requests within two weeks.
The webshop reaches the buyer, but not the physiotherapist or the hotel that recommends the equipment. We used AI to work out who holds the most promise and approached them through LinkedIn.
2 weeksto the first requests