What does a VC look for when they evaluate a product’s viability? Is there a difference between a hardware and software startup? How do you build the right team to develop a coherent IoT solution that differentiates and delivers real outcomes?

Understand from Chirayu Wadka of SeedPlus on what makes a startup scale, and why Ofo and Mobike succeeded despite the failure of other bike sharing companies. Plus, we hear his thoughts on starting a VC fund, and the landscape of tech VC in South East Asia. 

Erik: Hello, I'm Erik Walenza, founder of IoT ONE, and this is the industrial IoT spotlight. Every episode, I interviewed one expert about a project that is impacting businesses today. Visit us at Iotone.com to learn more, or email me to start a discussion.

Peter, thank you for joining us today. Peter is joining us from iCubed4, a company based in Shanghai, China, that does intelligent manufacturing systems for Chinese and foreign companies. In China Peter is the general manager, and we're very happy to have you with us today.

Peter: Thanks, Erik. I'm in China since 10 years, and I've been working in multinational companies and then in the consulting business over the past years, focusing more on the topic of Industry 4.0 and smart manufacturing, was a panel speaker at a Global Automotive Forum and European chamber, British chamber here. And then last year, we have founded a company iCubed4 in order to focus on the demands here in China for digital manufacturing.

Erik: And one of the first technologies that you're developing with iCubed4 is that MES SaaS solution. So why is this your first focus? Maybe explain for the listeners a little bit about the China market and why this is necessary right now?

Peter: I believe, for the China market, why is this the right thing to go for? It's because of the easiness to implement a SaaS solution compared to a complex MES solution. So the situation here in China is that companies usually do not have very clear processes when it comes to producing something, and then going in with a big sculpting and a large project doesn't usually really bring the benefits they're expecting, but will end up with a lot of changes that are more or less nice to haves, instead of really focusing on improving the business processes in the company.

Then we have developed this SaaS solution in order to enable the entrance to digital manufacturing with a simple MES solution. Also, for smaller companies, those companies without processes in place, the approach is different. They start using the system as it is, they get basically a very straightforward process. And they change it along the way when they really use it, when they really realize what they need in order to improve.

Erik: Yeah, and I think this is a great topic for China, because I talked to a lot of IoT companies that are looking at entry in China, and working with manufacturers here. And I think one thing that a lot of them don't realize is that their solutions are contingent upon the manufacturer being able to provide data and manage data. But a lot of the companies here that would be potential customers simply don't have the capability. First, they need an MES. Describe for us who is a typical customer here? So if we're looking in the manufacturing space, I imagine you're not serving the bosses of the world in China, who would you be working with?

Peter: Yeah, typical SMEs in China, mainly automotive and manufacturing clients, not so much in the process industry. Those are other bigger companies. But the typical company, which really has a need for technology and MES are companies in the automotive industry, yes, especially because their customers are usually bigger. When there are tier two tier three suppliers, their customer could be companies like Bosch or Schaeffler, and those companies demand for a certain level of quality; they are looking more and more into electronic data interchange. And if you want to become a supplier to the big ones, you got to fulfill these requirements, else it will be difficult to win the deal for them.

Erik: I was talking with Mr. Lee, the chairman of [inaudible 04:34] earlier this week, and he made this point that a lot of the implementation among SMEs in China has not been driven so much by their recognition of a particular need, it's being driven by their customers making a requirement. And in order to meet that requirement, they have to invest in technology. Is that the same situation that you're finding that it's usually an external requirement and that's forcing investment?

Peter: Yeah, absolutely. In automotive, this is definitely like that, and actually, not only for the Chinese companies, also for German or other Western companies who have to fulfill certain requirements for their customers.

Erik: I know this differs case by case. But how do you connect them with a Chinese factory? Is it generally you making building a relationship with a factory GM, identifying that they have this problem? Or is it you working with, for example, General Motors and saying General Motors, you're going to identify 10 facilities suppliers that you'd like to upgrade, and introduce me, and then I'm going to work and help those guys get upgraded?

Peter: Yeah, both ways. So of course, we go out to sell the traditional way. But we are also working on a cooperation with one of the big companies to exactly do what you just said. And we especially hope to do business with the big companies because it will be easier to get into the actual user.

Erik: And then before we move on to the technology, one last question about the customer here. Who do you tend to work with an SME in terms of seniority level? Where are the decisions being made regarding the purchase decision as a whole and then when you get down to the more micro implementation and scoping decisions, are those all done at the top or is there a delegation?

Peter: That's also looks a bit different to dealing with Western companies. And Chinese companies, the best case is, of course, meeting the decision maker, the CEO or the general manager right away. And those are usually the ones who will make the final decision. And the missing part here is the link to the people further down in the organization, that the owner or the manager of the company makes a decision to implement the system doesn't talk to the actual end users before making that decision. And then of course, it can become a very challenging implementation project now.

So what we advise in the very beginning is already to establish here's a concept so that we can immediately from the beginning on when we start talking about the project normally talk to the high level people, but also to those who will use it at the end to make sure that we mitigate the risks.

Erik: Just from a different angle, my wife is working in M&A, and also has this experience often that the Chairman, the President is making the decisions but not communicating down low and you often have to coach or prod to an extent to make sure that communication does happen. Otherwise, you're going to run into some bottlenecks once you get into implementation. How does a SaaS MES differ from a traditional?

Peter: Yeah, I believe, looking at SaaS versus traditional, there are a few very big differences. First of all, in a SaaS environment, you have multiple customers using one application How we realize that we share the applications but we separate the data. So let's say we have 50 customers, 50 customers use the same application and that brings down the cost a lot and finally, also the price eventually for a customer.

The database, the separate data hosting for the different customers takes away the fear of the customers that their data is being mixed with the competition. For example, that's a big topic in China, even though there is actually no worry that the data is going to be mixed. But it's something that helps at least here. Then you, of course, in a SaaS environment, you pay a monthly or yearly charge, you don't have large investments for data centers and stuff hiring. You have very low or even no capital expenditure while you have very large capital expenditure when you go into a traditional on-premise hosting system. And you have a very clear OpEx, of course. You know what you pay per user per month.

While in a traditional environment you never know what kind of surprise you will have when you look at your operational expenditure. On the long run, staff costs increase, while it costs decreases. However, over the longer run, you will have an increase in operational expenditure when you run your own data center while we the more customers we have using on our system running on our system, we are actually reducing our costs and reducing the prices for end users.

And another factor is security. Most of the companies believe Cloud or private Cloud, and this is a private cloud solution is not secure. But the point is when you are an automotive company, your core businesses doing automotive, building parts, producing parts, it's not doing IT. Our core business is doing IT. So of course, we are in a much better position to provide redundant, reliable and secure systems than an IT department that’s two, three people. So it's just a matter of trust. The fact is security and rather reliability is actually higher in the Cloud. It is just a matter of trust.

Erik: So traditional system might be updated once a year or a couple times a year, what does it look like from a SaaS situation so both in terms of change requests, and also in terms of over the air updates?

Peter: Now, it's also a completely different approach. So as it says in a traditional environment, you get updates 1, 2, 3 times a year from Microsoft, for example, while when you run on a SaaS environment, the SaaS provider captures usage data and performance data from the application automatically. So you're able to determine which parts of each of the application is used more than others and you're able to find out how much the load is on certain parts of duplication.

So you can respond accordingly and extended and updated accordingly. You can run surveys with your customers, and then they will receive their change requests at a certain point in time, obviously, not guaranteed with the next release. But since we will do three to six releases a year, there is a high chance that customers also get their change requests, which they actually would pay for; on a long run, we'll get them free.

Erik: So obviously, one of the challenges in China is traditional manufacturers often don't have the IT capability to manage a complex solution, so there's some value in simplifying. Often, they also just don't have the need for a full service MES. What does it look like in terms of capabilities and in terms of interface for your SaaS solution?

Peter: Of course, it's web base, so it's got to work on phones. That’s very important. So certain functions, like the warehousing function, you can operate from your iPhone, if you want. There is an app for it. Each of the functions has its own user interface. Now, unfortunately, we don't have a video podcast so I could show it to you. But it's a simple, easy-to-use interface, which everybody can use. You only see what you're supposed to see. So if you're a worker on a machine or assembly station, you have only two three fields to fill in. And then it's important to use scanners to eliminate paper in production, or you capture data from machines in order to enter data into the system.

So we reduce the complexity of the user interface to minimum or we even try to take them away. That's especially for automotive important. The more steps you put into the production, the less productivity you will have. And when you have very high tech times, you don't want to have a user entering data manually; so less user interfaces.

Erik: Now, I imagine this is not as customized as a traditional, although. Obviously, you can customize it to an extent per user. If we're looking at functions or capabilities, what is the tradeoff or how does this differ from you could say our full service traditional MES system?

Peter: Yeah, the default solution doesn't include capturing machine data for example. So that we have an add-on which allows us to capture the machine data, store it locally on the customer side and then upload it periodically to the cloud system. Another one is the integration with different existing systems. We can integrate with ERP and warehouse management solutions integration with product lifecycle management for example is not foreseen. Then the master data, obviously, use bonds and routings, which we bring together and we call it a technology, that is less complex as you would have it, for example, in an SAP system, which you can run globally with global product, global materials. Those kind of characteristics we don't offer.

Erik: What about from the implementation side? So obviously, for a traditional MES, this is a big part of the cost and also the timeline for getting a platform up and running. What does it looked like, maybe if you can just walk us through the steps from scoping through implementation, and then I imagine you're going to potentially do a simple implementation and then have some iterative improvements.

Peter: Yeah, the approach is different, as we already said, in the very beginning. So there is not much blueprint, actually, as we would see it in the traditional projects. The average duration is two to three months for an implementation. And we only have three phases. We start with a preparation phase, where we meet the team have a kickoff, we go into a planning session with all users involved. We do requirements collection while we provide the training.

So we do a training together with everybody involved in the project and at the same time, collect the requirements. Users can say, okay, this doesn't really work that way for us. And after assessing these requirements to make the decision, whether they need changes to be implemented already during the project or not depends on the importance of the change. And then we get to the customer started with data collection and cleansing, provided they’re his data.

Then in the second phase, we go right into realization. So we set up the system, we implement those changes, which are relevant for the implementation, which cannot be postponed. Again, we do training, and we support the key users to train the end users. We continue with data collection and cleansing. And then finally, we come to the go-live day. And after go-live, we will support for a certain period of time, depending on the maturity of the customer. And we go through the lessons learned, which is an important thing for us in order to improve our project method on the long run and become more efficient.

Erik: And within that process, where do you see the challenges? Where do you face bottlenecks, cost overruns?

Peter: Actually, as long as we stick to our implementation model, it's fine. The discussions come when change requests are on the table, obviously. So that's more of a commercial thing. And then it depends how much commitment we get from the customer for how long he signs a contract with us, then, obviously, we can turn some screws and bend the system to a certain degree, also without extra charges. So we try to balance it out with the contract.

Erik: Oh, one point that Mr. Lee from [inaudible 18:14] had was quite insightful is that within China, companies often ask, how do I become an Industry 4.0 player? And the answer is, I think you don't, that's not the objective. The objective is to become more profitable. And the question then is how do these technologies help you to become more profitable? The becoming an Industry 4.0 is not an end in itself.

The danger of this approach for trying to become Industry 4.0 and first of all, is you're probably going to scope wrong because that's just not a business objective in itself. But the second is that in an environment like China, if you have a very sophisticated system that is significantly more sophisticated than your supply chain partners, you're going to become to an extent an island, unable to take data from partners, unable to share data with partners and a lot of the value then of the systems you've invested in, it's just not going to be realized. So how do you address this in ecosystems where the value that you're trying to create requires that partners also invest to an extent in in IT systems to be able to share data, and that they then have an also an open mind and the business process is set up to share that data in an effective way?

Peter: This is what we want to achieve with our solution. We want to help companies to move up to certain standards in order to be able to have data. And then the next logical step is how can we integrate these companies? How can we integrate demand and supply here? So we have an interface which allows you to electrically interchange information with your suppliers, the supplier just needs a web browser for that.

Then, in the next step, we're looking to find a strong partner to build up an end-to-end platform, that also customers who look for suppliers are able to electronically send demands to suppliers, get traceability and quality and production information in real time from their suppliers so that they know when they can expect the deliveries. And this kind of integration is what we are looking at to build up the midterm.

So but overall, of course, we are convinced of the business idea of like for is come up with a simple solution to allow specially the Chinese SMEs to take first step into technology in production, and with that step, being able to serve on a certain level to a certain customer base.

Erik: I think one of the core value propositions of any SaaS is typically the price point. Let's compare the technology and also the implementation cost may be a separate bucket. So what does this look like MES SaaS versus traditional implementation?

Peter: Traditional implementation, if we would run a sculpt project, also, in the past, it's round about six months standard and then it would be 1-2 million depending on the size of the project, really, depending on the requirements, so many consultants are on there plus license fees, obviously. And if you look at, for example, SAP, then you spending another few millions on the license plus 22% maintenance fee every year.

For us, we’re able to offer an implementation for 200,000-300,000 renminbi, and then we look at monthly prices of 5,000-15,000 RMB, depending on a number of users, depending on the number of functions the customer is using. And another item, of course, when you go for a traditional way is you have to host the system by yourself or you need to manage the system by yourself. No matter where you hosted in the Cloud or a new data center, you still have to manage it. While here, you only need an internet connection and a web browser.

Erik: So that's really a 5-10x difference in pricing. So before we end, I want to talk a bit about the fast follower or the copycat or you could just say the competitive environment in China. So China is fiercely competitive. If you bring a good solution to the market in China, there are other companies that are going to be watching you and they're going to be following up with similar solutions soon. With this type of solution, do you see the competitive environment as being a threat? Or do you think it's a good thing if MES gains traction, even if it means that there's more competitors in the market, how do you view the competitive environment?

Peter: I think there is no real competition yet in the SaaS MES market. I haven't really found a real competitor here in China local. There's one but it's looked rather fakie. They for sure don't provide proper SaaS service. This is just MES hosted in the Cloud and that's a big difference to offering a SaaS service. Offering systems just hosted in the Cloud doesn't mean you're actually offering SaaS service and it will not reduce your cost as a SaaS provider when you have separate installations for each customer. And then you're not able to offer these prices.

We are German-invested company, but we are offering it for Chinese prices. So right now I don't see competition. But of course, the big ones they are going in the same direction, Oracle SAP, of course, as well, but they're playing in a different league. So belief till now we are serving a segment where were the big ones don't look at. It's really the small and medium sized companies who would never be able to afford big brands like SAP or Oracle or GE or Siemens. I guess there will be fast followers coming up.

Erik: [inaudible 24:59] put out a good report on the Made in China 2025 policy recently. And one of the points that they made was that enterprise or industrial software was actually the category that's the most difficult for Chinese companies to enter into. Their forecasts was that it's going to be another decade or so before we see a really strong competitive environment in China, among local suppliers; whereas if we're looking more at the control level sensors in a lot of categories, the Chinese companies are already there, or they'll be there soon. So we'll see how this actually plays out. But maybe you will be one of the last categories to actually face some stiff local competition.

Peter: This is indeed that the biggest difference here is that usually when you come from Western company, you have a technical background, but you understand the business as well. So you're able to provide technology consulting services, instead of pure technology.

Erik: What is one interesting technology on the horizon that you think people need to be paying attention to in China because it's going to be shifting how people operate in the next five years?

Peter: MES.

Erik: Tell me how can people learn more about you? How can they contact you if they'd like to reach out?

Peter: Drop me a mail to Peter@iCubed4.com. I should probably spell this out, Peter@iCubed4.com.

Erik: Peter, thank you so much for sharing, I think a very interesting discussion into MES SaaS. I wish you all the success in the world. And I look forward to talking more in the coming years on this topic.

Peter: Thanks, Erik. And thanks for the opportunity to taking the podcast today.

Erik: Thanks for tuning in to another edition of the industrial IoT spotlight. Don't forget to follow us on Twitter at IoT one HQ, and to check out our database of case studies and IoT one.com.

Erik: Thanks for tuning in to another edition of the industrial IoT spotlight. Don't forget to follow us on Twitter at IotoneHQ, and to check out our database of case studies on IoTONE.com. We help to accelerate digital transformation by advising business leaders on how to integrate IoT technologies into their operations and products. We appreciate your thoughts, suggestions, and of course, your reviews. And if you have an interesting project, we'd love to feature you on a future edition. You can write me at erik.walenza@IoTone.com.