While CAD companies are criticized for this the most, choosing one system over another almost always gets you at least partially locked in. Moving your data from one system to another is never simple. That is one reason that I suggest that industrial companies model their data and processes outside of their chosen solution so that they understand it in detail if a moving day ever comes.
To me, this is only going to get more important with cloud-based offerings under subscription. At least with on-premise solutions that were perpetually licensed you could stop paying maintenance and continue running (at your own risk of course).
I think this question is important and can become one of the most interesting questions in the future as more manufacturing companies are discovering the value of SaaS PLM solutions and the industry is in general moving towards SaaS and cloud offerings.
Let’s bring some order and speak about different types of SaaS systems, The first group is cloud-hosted single-tenant solutions and the second is SaaS multi-tenant platforms.
Single-tenant hosted Cloud PLM Solutions
These solutions are in general no different from on-premise PLM. The only difference is that this solution is run on some cloud servers (data center or IaaS provider), but from the PLM architecture standpoint, everything stays the same. As a customer, you’re locked to the database supported by the vendor and when a “moving day” will come, you will need to figure out how to get the data and migrate to another PLM solution (or database). The PLM industry was dealing with such kinds of migrations all the time. It is an expensive and painful exercise.
Multi-tenant SaaS platforms
SaaS solutions using multi-tenant cloud and data architecture are completely different from their predecessors, Their data architecture usually scales globally, provides a logical network layer to provide access to granular users, teams, and companies while the overall data architecture can span across multiple data services and storages. At first, you may think that getting the data from such a platform is much more complex compared to old monolithic single-tenant solutions, but in fact, it can be only an illusion.
Modern Data Architecture and Data Portability
One of the biggest advantages of modern new multi-tenant solutions is their reliance on new technological stacks and includes multiple layers such as polyglot persistence, micro-services, and user experience. In addition to that, all modern PLM SaaS systems contain an advanced DevOps layer, which provides facilities for data archiving, environment migrations, and upgrades. Such services can simplify data portability and allow companies to provide granular access to the data. Besides that, because of the need to operate with large complex data structures and a global data management environment, the administration and management functions of these systems have more tools to handle data migration.
Moreover, customers of SaaS systems are from the beginning very sensitive to the question of data locking. When a manufacturing company is buying an on-premise PLM system, the question of data portability might not be so obvious. However, coming to SaaS systems, the ability to export data at the “moving day” is a natural thing and often asked before companies even consider starting using new SaaS Services. Global online platforms such as Google and some others always provide a way to get your data out in a way of a data dump of some sort of standard output – Excels, CSVs, XML, and other formats.
Data and SaaS Business
I’d like to suggest another perspective of data standards and SaaS. While old on-premise PLM were sold, implemented, and used for decades without changes, SaaS business is different. In most cases, the vendors give you a subscription and are interested in renewals. Which changes the business relationships and brings up the question of data services and interoperability. I can see a moment when SaaS PLM companies will convert into Data as Service solutions providing access to the information to other services. It will, of course, requires new business models that are different from traditional PLM data locking paradigms.
What is my conclusion?
SaaS data architecture and business bring a new perspective in data management, data standards, and data portability. When you plan to embark on your SaaS PLM journey, check your SaaS vendor carefully. You need to ask how data can be extracted, how open is the system from the integration standpoint, and what are terms of getting your data out of the system are. The most interesting trend is turning SaaS PLM systems into data services and providing the data in an open and transparent fashion to all other SaaS services to consume the data. Which will bring us one step closer to the true digital transformation in the PLM industry. Just my thoughts…
Best, Oleg
