Unless you’ve been under the rock for the last decade, you most probably heard about SaaS (Software as a Service). The term was around for a quick long time, but recently it became especially hot. It can be used as a fundamental difference in variety in cloud PLM technologies available in the market.
But let me step back. SaaS (software as a service) is a technology that extends ASP (application service provider). However, for the last decade, SaaS is used for vendors and applications that develop and use their own software compared to vendors hosting somebody else software. I realized the market confusion around SaaS, cloud, and hosting is significant. So, I published my earlier article – Multi-tenant SaaS Common Misconceptions.
The software architecture of most ASP solutions and recently hosted solutions mandated maintaining a separate instance of the application for each business, for the last decade, we’ve seen that normally SaaS solutions normally utilize a multitenant architecture, in which the application serves multiple businesses and users, and partitions its data accordingly.
The reality of the PLM market is that the top 3 biggest vendors are holding a majority of the PLM market share and remaining unchanged for the last two decades. However, their market is concentrated in the business of large global OEMs and enterprise manufacturing companies. The majority of revenue in PLM in 2020 is still earned from on-premise PLM deployment.
The presence of PLM solutions in mid-size manufacturing companies, contractors, and supply chains is very low and this is one of the biggest opportunities behind digital transformation in manufacturing and market growth in the PLM domain specifically and even beyond a traditional scope of PLM. All vendors are realizing that the future is belonging to SaaS and each of these vendors makes steps towards not losing future SaaS business. This is probably the topic for a separate blog, and today I want to focus on multi-tenancy and PLM SaaS architecture. Read some of my earlier articles here. Here is probably the one I highly recommend if you’re interested in a multi-tenant architecture.
In my blog today, I decided to focus on the top 3 things that might have a chance to be avoided by vendors.
1- Utilization, cost, data management, and business models
The truth is that multi-tenant solutions are extremely powerful when it comes to resources, consumption, and ability to optimize the software stack. It comes from the side of computing and storage utilization and can be a huge factor. Think about under-utilized servers that can be split into independent services, run and scale independently and as a result save tons of resources. Administration and data management can be significantly simplified when using multi-tenant data platforms (no need for a separate database/data servers). The important thing to understand when you speak about multi-tenant SaaS PLM Is that the cost of a single user is zero. The system will increase the cost as it will be growing, but such flexibility provides a big opportunity to lower the cost. But, I think, most vendors will never talk to you about it.
2- Tenant data organization and and data sharing
Data sharing is one of the key-value propositions behind cloud software and SaaS. But if you go beyond a simple marketing message, data organization is fundamental to understand to check if you can actually share the data. Take an old established PLM software and host it on AWS. You will get access to the data, but if you need somebody else (another organization, supplier, or contractor) to get access, all these organizations will have to use the same PLM instance. In SaaS terminology, it is called Tenant. Like apartments in the house can be isolated and you won’t get access to your neighbors, multiple instances of PLM systems are isolated as well. And unless, another company will use the same PLM system instance and database, to share data will be impossible even if you run both PLM instances on AWS. So, the value of “cloud sharing” will be diminished to zero. So, while vendors say cloud is an enabler to share data, the real value can be only achieved with the proper data architecture and features.
3- Administration and maintenance
Let’s say the question of data sharing is not relevant or you convince all contractors and suppliers to use the same PLM instance. Sounds like the problem is solved? Not really. In the world “before cloud”, each company has IT managing PLM servers. It was painful, but companies did it. Now, imagine a PLM vendor sold the system to a large number of customers. If all these systems need to be administered, maintained, upgraded, and run properly to be available to you. The cost of such operation will be quickly skyrocketing and goes beyond all possible level PLM vendors will support unless even midsize companies will continue to pay a small fortune for PLM subscriptions. So, the best-kept secret is that if PLM vendors are hosting PLM system instances for each customer, they will avoid selling them below a specific threshold, otherwise the entire operation will become deeply troubled by cost structure and unit economics.
What is my conclusion?
Cloud gives us amazing technologies and a lot of opportunities. However, to get the value, the architecture of the systems should be different from the architecture of mature PLM technologies we have in the market. PLM stacks developed 15-25 years ago are SQL database based client-server or web products. But moving from old PLM SQL database architecture to SaaS requires major software retooling and paradigm change. Properly architectured multi-tenant SaaS PLM systems have the ability to outperform traditional PLM systems hosted in single-tenant servers provided by IaaS vendors and private hosting companies from many aspects – features, administration, and, ultimately, unit economics. The last is an enabler to cut costs and develop an alternative business model. It is an opportunity that can be leveraged by newcomers and establish vendors in the PLM business. Just my thoughts…
Best, Oleg
Disclaimer: I’m co-founder and CEO of OpenBOM developing cloud based bill of materials and inventory management tool for manufacturing companies, hardware startups, and supply chain. My opinion can be unintentionally biased.
