A Map of Your Business: Understanding the Semantic Layer

Two dimensions: horizon and resolution

Once these pieces begin to come together, a practical question follows: how complete is our picture? Most companies know some parts of their operation in considerable detail and see others only indirectly. Sales may be well documented, while the work required to serve those customers is spread across emails, time sheets, and the experience of a few colleagues.

It helps to think of this picture as a map with two dimensions. The first is its horizon: how much of the business and its environment it takes into account. Adding service operations to a model that previously covered only sales expands that horizon. So does identifying a supplier portal, a contractual dependency, or the person responsible for an exception.

The second is resolution: how much we can distinguish within the area already mapped. A total service cost offers one level of detail. Knowing which customers, products, or recurring issues account for it makes different questions answerable. In the inventory example, the distinction between physical, reserved, and available stock has the same effect. Satellite imagery has the same two properties: how large an area a capture covers, and how small a detail a single pixel can show. One image may cover a region at thirty metres per pixel, another a single site at thirty centimetres.

These dimensions develop independently. A narrowly focused model can be very detailed, while a broad one may still contain areas that are poorly understood. A known but unconnected source is already within the horizon as a possible source of information; its contents are not yet available for analysis.

The aim is to develop the map where it helps the business most. If service effort is changing customer profitability, that may justify more attention than another breakdown of sales already understood. And because greater detail can expose contradictions as well as answer questions, the map must retain the source, freshness, and limitations of what it describes.

A description that develops through use

Consider what happens when the warehouse manager reviews a proposed stock transfer. The report shows enough units, but she knows that a large share is reserved for a customer. The agreement is real; the reservation simply has not found its way into the data used for the analysis. Her response reveals both a missing piece of context and a weakness in the process that records it. Remote-sensing teams call this ground truthing: checking what an image suggests against what is actually on the ground.

After the agreement is checked, the reservation can become part of the shared description, with its source, responsible person, and period of validity. The next analysis can then distinguish stock that is physically present from stock that is actually available. What was known by one person becomes knowledge others can use.

There is also an opportunity to improve the underlying work. The team can agree where future reservations should be recorded, who maintains them, and when they must be updated. Showing the responsible colleagues how this prevents conflicting delivery promises gives the additional data collection a clear purpose.

An open implementation, owned by the company

The time spent agreeing definitions, documenting exceptions, and connecting operating knowledge creates something of lasting value. For an SME, this description holds experience that would otherwise remain scattered across systems and individual people. We believe it should be treated as a company asset.

This is the thinking behind Meridian’s open implementation. Customers can inspect their semantic layer: how a figure is defined, which relationships support it, where information comes from, and which gaps remain. They can review and extend that description as their business develops.

The implementation is designed to interoperate with each platform’s definition of a semantic layer, translating existing models into Meridian and company definitions back into the structures supported by other platforms. The business knowledge remains available to the customer as their technology changes. Cartographers keep survey data independent of any one projection, so the same coastline can be drawn for a globe, a nautical chart, or a city plan. Company definitions deserve the same independence.