This post is based on a mix of interactions I had the last two weeks in my network, mainly on LinkedIn. First, I enjoyed the discussion that started around Yoann Maingon post: Thoughts about PLM Business models. Yoann is quite seasoned in PLM, as you can see from his LinkedIn profile, and we have had interesting discussions in the past, and recently about a new PLM-system, he is developing Ganister PLM, based on a flexible Graph database.
Perhaps in that context, Yoann was exploring the various business models. Do you pay for the software (and maintenance), do you pay through subscription, what about a modular approach or a full license for all the functionality? All these questions made me think about the various business models that I encountered and how hard it is for a customer to choose the optimal solution. And is the space for a new type of PLM? Is there space for free PLM? Some of my thoughts here:
PLM vendors need to be profitable
One of the most essential points to consider is that whatever PLM solution you are aiming to buy, make sure that your PLM vendor has a profitable business model. As once you started with a PLM solution, it is your company’s IP that will be stored in this environment, and you do not want to change every few years your PLM system. Switching PLM systems would be affordable if the PLM system would store their data in a standard format – I will share a more in-depth link under PLM and standards.
For the moment, you cannot state PLM vendors endorse standards. None of the real PLM vendors have a standardized data model, perhaps closest to standards are Eurostep, who have based that ShareAspace solution on top of the PLCS (ISO 10303) standard. However, ShareAspace is more positioned as a type of middleware, connecting between OEMs/Owner/Operators and their suppliers to benefit for standardized connectivity.
Coming back to the statement, PLM Vendors need to be profitable to provide a guarantee for the future of your company’s data is the first step. The major PLM Vendors are now profitable as during a consolidation phase starting 15 years ago, a lot of non-profitable PLM Vendors disappeared. Matrix One, Agile, Eigner & Partner PLM are the best-known companies that were bought for either their technology or market share. In that context, you might also look at OnShape.
Would they be profitable as a separate company, or would investors give up? To survive, you need to be profitable, so giving software away for free is not a good sign (see the software for free paragraph) as a company needs continuity.
PLM startups
In the past 10 years, I have seen and evaluated several new PLM companies. All of them did not really change the PLM paradigm, most of them were still focusing on being an engineering collaboration tools. Several of these companies have in their visionary statement that they are going to be the “Excel killer.” We all know Excel has the best user interface and capabilities to manipulate a collection of metadata.
Very popular is the BOM in Excel, extracted from the CAD-system (no need for an “expensive” PDM or PLM) or BOM used to share with suppliers and stakeholders (ERP is too rigid, purchasing does not work with PDM).
The challenge I see here is that these startups do not bring real new value. The cost of manipulating Excels is a hidden cost, and companies relying on Excel communication are the type of companies that do not have a strategic point of view. This is typical for Small and Medium businesses where execution (“let’s do it”) gets all the attention.
PLM startups often collect investor’s money because they promise to kill Excel, but is Excel the real problem? Modern PLM is about data sharing, which is an attitude change, not necessarily a technology change from Excel tables to (cloud) shared tables. However, will one of these “new Excel killers” PLMs be disruptive? I don’t think so.
PLM disruption?
A week ago, I read an interview with Clayton Christensen (thanks Hakan Karden), which I shared on LinkedIn a week ago. Clayton Christensen is the father of the Disruptive Innovation theory, and I have cited him several times in my blogs. His theory is, in my opinion, fundamental to understand how traditional businesses can be disrupted. The interview took place shortly before he died at the age of 67. He died due to complications caused by leukemia.
A favorite part of this interview is, where he restates what is really Disruptive Innovation as we often talk about disruption without understanding the context, just echoing other people:
Christensen: Disruptive innovation describes a process by which a product or service powered by a technology enabler initially takes root in simple applications at the low end of a market — typically by being less expensive and more accessible — and then relentlessly moves upmarket, eventually displacing established competitors. Disruptive innovations are not breakthrough innovations or “ambitious upstarts” that dramatically alter how business is done but, rather, consist of products and services that are simple, accessible, and affordable. These products and services often appear modest at their outset but over time have the potential to transform an industry.
Many of the PLM startups dream and position themselves as the new disruptor. Will they succeed? I do not believe so if they only focus on replacing Excel, there is a different paradigm needed. Voice control and analysis perhaps (“Hey PLM if I change Part XYZ what will be affected”)?
This would be disruptive and open new options. I think PLM startups should focus here if they want my investment money.
PLM for free?
There are some voices that PLM should be free in an analogy to software management and collaboration tools. There are so many open-source software management tools, why not using them for PLM? I think there are two issues here:
- PLM data is not like software data. A lot of PLM data is based on design models (3D CAD / Simulation), which is different from software. Designs are often not that modular as software for various reasons. Companies want to be modular in their products, but do they have the time and resources to reinvent their existing product. For software, these costs are so much lower as it is only a brain exercise. For hardware, the impact is significant. Bringing me to the second point.
- The cost of change for hardware is entirely different compared to software. Changing software does not have an impact on existing stock or suppliers and, therefore, can be implemented once tested for its purpose. A hardware change impacts the existing production process. First, use the old parts before introducing the change, or do we accept the (costs) of scrap. Is our supply chain, or are our production tools ready to deliver continuity for the new version? Hardware changes are costly, and you want to avoid them. Software changes are cheap, therefore design your products to be configurable based on software (For example Tesla’s software controlling the features to be allowed)
Now imagine, with enough funding, you could provide a PLM for free. Because of ease of deployment, this would be very likely a cloud offering, easy and scalable. However, all your IP is in that cloud too, and let’s imagine that the cloud is safer than on-premise, so it does not matter in which country your data is hosted (does it ?).
Next, the “free” PLM provider starts asking a small service fee after five years, as the promised ROI on the model hasn’t delivered enough value for the shareholders, they become anxious. Of course, you do not like to pay the fee. However, where is your data, and what happens when you do not pay?
If the PLM provider switches you off, you are without your IP. If you ask the PLM provider to provide your data, what will you get? A blob of XML-files, anything you can use?
In general, this is a challenge for all cloud solutions.
- What if you want to stop your subscription?
- What is the allowed Exit-strategy?
Here I believe customers should ask for clarity, and perhaps these questions will lead to a renewed understanding that we need standards.
PLM and standards
We had a vivid discussion in the blogging community in September last year. You can read more related to this topic in my post: PLM and the need for standards which describes the aspects of lock-in and needs for openness.
Finally, a remark related to the PLM-acronym. Another interesting discussion started around Joe Barkai’s post: Why I do not do PLM . Read the comments and the various viewpoint on PLM here. It is clear that the word PLM unites us all; however, the interpretation is different.
If someone in the street asks me what is your profession, I never mention I do PLM. I say: “I assist mainly manufacturing companies in redesigning their business processes using best practices and modern digital technologies”. The focus is on the business value, not on the ultimate definition of PLM
Conclusion
There are many business aspects related to PLM to consider. Yoann Maingon’s post started the thinking process, and we ended up with the PLM-definition. It all illustrates that being involved in PLM is never a boring journey. I am curious to learn about your journey and where we meet.
4 comments
Comments feed for this article
February 24, 2020 at 11:05 am
Håkan Kårdén
Great post Jos. A lot of experience written down in one post. I am planning on writing a piece about the increased importance of standards and the threat with PLM Cloud if you have not planned properly. Let’s see when I get the time to do it.
On a side note – ShareAspace is using a Graph Database – so nothing with standards is blocking that.
Thanks for your feedback Håkan – looking forward to your contribution. And I agree a graph database has nothing to do with blocking standards, maybe even providing more felexibilty to adhere to multiple standards (or am I dreaming too much ? ) .
LikeLike
February 24, 2020 at 11:24 am
Håkan Kårdén
Not sure about the flexibility (simply I just don’t know) but selected because of performance.
LikeLike
February 25, 2020 at 6:41 am
Yoann Maingon
Hi Jos, Great post again. Thank you for the mention. There is not much to add, you cover it all.
Having a profitable PLM editor is key, and I’m not sure we really know when they are. It is not because they make money that they are profitable. Also when they are in larger organizations, like the big three. PLM is generating lots of revenues for service integrators, not so much for editors compare to CAD. Therefor their investment might not be great according to their overall size.
About the standard, if we could have at least a lighter set of business standards a bit more like UBL (Universal Business Language) that would help. I think it’s the only way a cloud solution with a single data model could emerge.
I know I have a great journey ahead of me. Business model is very tricky 🙂
Hi Yoann, thanks for your feedback and indeed the business model for a PLM editor has the challenge that service integrators can deliver and drive for a lot of customizations / configurations. Related to a “lighter” standard I also fully agree that with a UBL (common sense) you can reach already a lot. Thats why I wrote in the past about the importance of a PLM data model. Success with your journey. Best regards, J0s
LikeLike
February 25, 2020 at 6:43 am
Yoann Maingon
Also I totally agree with Håkan, best value of a graph database is the performance. flexibility is great for us, but the visible value for larger companies will definitely be performances at scale.
LikeLike