The essence of simplification is grasping what is and what is not important. Conveying Insight is a core EA-task why simplification is a must-do activity.
Apply simplification to the ever so complex Logical System Landscape and you have The Conceptual System Landscape. The Conceptual System Landscape is a strong tool to communicate the business-IT relationship, the year-by-year transformation roadmap and more to business units and CxO-level.
One of my better EA-days was when the CEO required the CIO to provide annually updated technology transformation roadmaps presenting a 3 years plan on top of the Conceptual System Landscape! EA providing business-IT insight right to the top management.
Mapping the current technology landscape provides you with a strong business-IT insight, but in the process you are likely to realize:
- The logical system landscape is a killer artifact enabling interaction with and guidance of projects and constructive dialog with a multitude of stakeholders.
- The logical system landscape is far too complex to communicate in most situations, especially to upper management.
EA needs to bring the crucial insight we attain into a more digestible form – the Conceptual System Landscape. The purposes of the conceptual level is to:
- Insight to how business is supported by technology – where and how.
- Facilitate a dialog of benefit and deficits of the current technology support.
- Establish a platform enabling business and IT to engage in a qualified discussion regarding future changes and consequences.
- Boil it down to a communication friendly one-pager.
As the Enterprise Architect you need it, the business need it and IT need it – the only real question is how to go about the significant simplification, making sure it captures the essence, is easy to grasp and uphold a consistent layout across time.
To share some of my experiences here an example:
Amanda’s first shop did really well and she has now expanded to five stores. Every store earns good money and customers asks for more sales points, but Amanda fears, that this will fail, as she is starting to lose focus on the individual stores. To make things more complex the stores have individual IT setups. To expand the business she needs to simplify as well the store concept, business processes and the technology setup.
Amanda asked the skilled Enterprise Architect, Peter, to help her out on the technology. As the business yet is fairly simple and Peter mapped similar businesses several times before he tracks down and maps the core of the current technology landscape. 4 POS systems, 4 WFM systems, 3 mail solutions, 2 eStores, 1 intranet used by 2 of the stores, and further 15 mapped and 30 crudely mapped IT solutions. Peter can see how all solutions adds value. A very significant duplication of functionality is also obvious. The full technology coverage would be value adding to all stores, but none of the stores currently benefit from the full technology coverage. In addition, annual cost would be significantly reduced if all stores were consolidated on the same technology. Peter also notices that the processes are very different across the stores, that no common stock inventory exist and that some time-consuming tasks like price updating and performance analysis could be coordinated across the stores, which would free-up several employees. To efficiently handle economy, inventory and perhaps do cross-store-ordering a consolidated ERP should be considered.
Amanda knows how to send e-mails do a bit of number juggling in a spreadsheet and some of the store systems. She is very strong at understanding her business. However, when she saw the early draft of Peters system landscape she got completely lost. Moreover, this did not at all show the complexity added by servers, databases, interfaces, SLAs, processes, misaligned perceptions and definitions of data, need of multiple data entry or the multitude of issues Peter noticed during his analysis.
Peter chooses not to share his logical system landscape. Instead, he devises a conceptual landscape showing only one POS, but indicating that more than one POS is used. At the same time he groups a number of systems – e.g. the WMSs, salary and clock in/clock out systems in a group called ‘Workforce Management’. To simplify the overview the landscape he divide it into 3 levels – IT, Backend and Frontend and structures the systems according to Amanda’s value chain.
After a very brief introduction to the conceptual system landscape Amanda mention that she would like to reduce to only one POS as she has observed that it is problematic when employees assist in other stores. Peter informs her that having 4 different POS’s as well brings a significant extra cost, but the cost of consolidation have to be investigated. After planning an investigation, Amanda also as Peter to investigate the cost of extending the intranet to serving all stores.
Peter with satisfaction concludes that the conceptual landscape leads Amanda – on her own – to successfully appoint some of the problems. With Amanda’s dawning IT insight, he believes that they during their next meeting may look at some of his other observations and after further one or two meetings even may be able to discuss some of the most critical of the technical issues.
The conceptual landscape in this example could look like this:
Simple to overlook and enabling the business swiftly to grasp the essence of the technology landscape and how the business is supported by technology. A good platform to start questioning and discussing the pains and benefits of the current setup, and discussing where the most important changes should be initiated.
When building the conceptual landscape, you are likely to fall into the trap of decision paralysis when considering best layout and what to exclude/group. Do not aim for perfect; a later rework will not cost much and it will be based on feed-back. Make a choice, use it and then improve if necessary – Lean in ‘SOLID EA’.
Tools for the Conceptual System Landscape
As this often ends up in presentations and you will want to try out many layouts, standard desktop tools is my favorite choice (e.g. Powerpoint or Visio). This does mean a continuous dual maintenance of the conceptual and logical landscape, but the layout is so different, that you should expect this anyway. Anyway, in this case the extra maintenance is a conceivable task.
Good luck building your Conceptual System Landscape.