Cost models in a data mesh
In a centralized cost model where the cost of analytics is borne by the company and the ROI is calculated at a much higher level, cost modeling is simple. All costs discussed in the previous section are borne by a central cost center. The complexity is shifted to selecting projects to execute given the limited budget.
In a distributed cost model where the central data mesh itself is a cost center and so are all the products, the cost is borne by individual products. The individual products have to muster the budget to fund their project and provide an ROI. They could build the product in isolation or join the data mesh. The advantages of joining the data mesh are manifold, such as the following:
- Shared costs of management and monitoring
- Easy access to data from other products
- Shared governance
- Shared best practices
If they choose to join the data mesh, then they need to pay the data mesh department the charges described in the...
