Skip to content
Thnkers

Data and integrations

Three systems. Four spreadsheets. Zero truth.

One data layer on top of what you already have — no rip-and-replace, no year-long ERP project.

What’s missing

The ERP says one thing. The spreadsheet says another. The CRM says a third.

Every meeting starts with ten minutes of arguing about which number is correct. By the time the right figure is agreed, the conversation that was supposed to happen hasn’t started. This is life when the data lives in three places and nobody owns the reconciliation.

The management report takes a week to assemble by hand — and it’s out of date the moment it’s sent.

Someone pulls figures from the ERP. Someone else updates the spreadsheet. A third checks the CRM. The report is formatted and sent. Two days later a decision gets made on numbers that were already stale. A week of work produced a document, not a view of reality.

Replacing the legacy system is too expensive and too risky. Living with it is also a problem.

The system works, in the sense that it hasn’t stopped. But it doesn’t connect to anything, doesn’t export cleanly, and every new initiative hits the same wall: the data is scattered and there’s no clean place to start. Too embedded to replace, too isolated to build on.

Every new initiative hits the same ceiling.

A new sales channel. A new reporting requirement. A process that needs data from two systems that don’t talk to each other. The idea is sound. The data infrastructure makes it impossible to start without a manual workaround that immediately becomes someone’s full-time job.

The promise

We don’t propose starting over. We build a single data layer on top of what already works: the systems stay, the data connects, and the director gets one screen that shows what’s actually happening. No big bang, no year-long implementation, no replacement project that costs three times the estimate.

What we do

  • Landscape audit

    One to two weeks. We map which systems exist, where the data lives, where the duplications and gaps are, and produce a data map and a prioritised integration plan. The first integration we recommend is almost always the one that produces something the director can see — usually a dashboard — because that’s what makes the rest of the programme credible.

  • Integrations

    We connect ERP, CRM, warehouse, e-commerce and spreadsheets through APIs and reliable pipelines. Where an API doesn’t exist — and with legacy systems it often doesn’t — we find another path: scheduled exports, direct database connections, structured parsing. Every integration has a defined owner, schedule and failure behaviour, so it doesn’t silently break and produce wrong numbers.

  • Unified data layer and dashboards

    Sales, production, finance — in one place, updating automatically. The dashboard the director uses every morning, not the report someone assembled last Tuesday. We build the data model before the interface, because the model determines what questions you can ask later.

  • Migrations without loss

    From bots, spreadsheets, no-code stacks and old databases into structured systems. The principle: the business rules transfer exactly as they are. The interface changes, the infrastructure changes, the rules stay the same — how the Alfa Homework migration worked, three bots into one application without changing what the logic does.

  • Platform mergers

    Two systems with overlapping data become one, with a unified user model, role-based access and single sign-on. For KI Ecosystem, two separate loyalty databases became a single platform with one login and one view of the customer journey. That unification was the technical foundation the marketplace was built on.

How we work

Audit first — one to two weeks — to understand the landscape and find the highest-value first integration. Then a data map and a plan. Then the first integration, chosen because it produces something visible and useful fast, usually a director’s dashboard. Then the rest in iterations, each building on what’s already connected.

Where we have done it

Questions

  • Will we have to replace our ERP?

    Almost never. The data layer lives on top of the existing systems. The ERP keeps doing what it does. We connect to it rather than replace it.

  • How long until we see a result?

    The first working dashboard is usually three to six weeks from the start of the integration work. That’s the milestone we aim for first, because it makes the value of the programme visible and concrete.

  • Our system is a custom-built legacy. Nobody understands it.

    Everyone says that. “No API” is not the end of the conversation — it’s the start of a different one. Scheduled exports, direct database connections, structured parsing: there’s usually a path. We find it in the audit.

  • Who maintains the integrations after they’re built?

    We offer a retainer for maintenance and development. The integrations are documented and handed over completely, so you can take them to any team. We don’t build dependencies on our continued involvement.

Order a data audit

Send us a list of your current systems. We’ll tell you what we’d connect first and what that would give you.

Order a data audit