Much like a business plan, a roadmap is a higher-level visualisation of your organisation. It provides a holistic view of where you intend to go as a business and how you will get there.
Now, what happens if individual teams started defining that roadmap, encapsulating their individual objectives and strategies and drawing the future of the organisation with their own pen?
This is an illustration of one typical situation where business people dream about implementing the most ambitious use cases and data projects without having the means to do so.
This is also the case of most of the case studies, discussed in previous articles, that naively design sophisticated data programs and yet have little to no technical capacity to bring those projects to life. Most of them did not have a data office. As a result, every business use case was developed by the business department and the data pipelines were set up by the BI/EDWH department. However, because of the complexity of the analytics (ML, AI), the size and the diversity of the data, the non-functional constraints (such as real time), every use case was developed ad-hoc. The estimated cost of the first two use cases was respectively 1 M €* and 1.2 M € (only including development costs, excluding production, infrastructure and licence costs). The most puzzling part is that the second use case was as costly as the first one. The reason for this lies in (1) the complexity of the underlying technologies which, in the absence of the right architecture and development strategy, are complicated to capitalise on and (2) the lack of qualified human resources to develop this kind of use cases. As a matter of fact, it is already a challenge to set up a core team to build a data foundation, but it is a far bigger challenge to staff the business lines with data experts capable of building every use case.
Results, lessons learned and what we can do to help
Although capital-intensive investments were allocated to the development of new business cases, the generated ROI did not meet expectation and hence, the use cases initiatives led by the business lines were deactivated for non-profitability reasons.
- The skills to develop these new data-driven use cases are scarce. Numbers show that the staff turnover of developers with those special skills is way higher than the regular developer’s one. However, the complexity of such data-driven use cases is directly shown in the investment they require. At Digazu, we have seen that building such data-driven use cases without benefiting from a data platform costs on average between 500 K and 1 M €. With a well-designed data-platform, this cost can be reduced by half.
- Adopting a platform strategy drastically helps control costs, reduce the projects’ time to market and decrease the TCO of each use case.
Our platform strategy, Digazu, is an integrated data engineering solution that encapsulates the complexity of high-end technologies and brings in a data-centric architecture capable of supporting the most ambitious use cases and data projects.
Digazu helped many companies optimise their data project lifecycle. As a turn-key and low-code data platform, Digazu helped with the orchestration and automation of the data project stages. From the collection, storage, transformation to the distribution (full streaming) of their data, companies were able to act faster and better. The data assets were assembled back properly and distributed to the tools (reporting, data crunching, new apps) made available to data consumers in a standardized and fully managed way.