Mike Olson, the Chief Strategy Officer at Cloudera, eloquently articulated the philosophy behind their business model in this LinkedIn article titled The Cloudera Model. I highly recommend anyone working on open source platforms and building value on top to stop coding, and read this article. Like right now!
I would like to highlight one quote from the article:
You can no longer win with a closed-source platform, and you can’t build a successful stand-alone company purely on open source
Finally somebody said it!
For years, I have been convinced of this model. I have often wondered how pure open source companies actually make money and justify such large valuations without offering any kind of monetization other than offering training, support and development services on top of the open source projects. Some do it via dual-licensing model, as Mike points out in his example of how SleepyCat worked it before getting acquired by Oracle.
Now, I don’t know if this is the best model that Mike has articulated, but it is close to Winston Churchill’s idea of democracy where he said something along the lines of:
…it has been said that democracy is the worst form of government except all those other forms that have been tried from time to time.
I feel the similarly about the Mike’s call, and I believe this is the best model of all the models out there currently to think of how you want to build a company and balance between open source and closed source to create unique value on top of open source. For most part, it seems that people agree with Mike’s article (based on LinkedIn comments and tweets). Those that don’t agree are in denial.
Use, contribute back, innovate (both open and closed) and create value – that sounds like a good viable business model.
Is this going to be the big guys’ game or can innovative companies break through?
I am not at the conference, but this tweet got my attention. The big guys I am guessing is a reference to companies like IBM, Cisco, GE, Orange, Vodafone, Verizon, AT&T, and others in similar size and stature. However, expecting the big guys to innovate in something that is so rapidly changing in such a short time is unrealistic. This is where new, smaller, and smart companies can innovate faster and create solutions, products and technologies. Over the next 5 years, I see this as a space that will be more dominated by smaller guys with innovative approaches. Big guys sure will be involved, watching, and sometimes collaborating with the smaller guys. But this is a huge opportunity for everyone, small or big to participate. After all, just yesterday, Brian Profitt reported on Read Write Web that this industry could be as large as 14 Trillion dollars, if not bigger. This space is big enough for the behemoths and the miniscule.
Machine to Machine (M2M), Industrial Internet, Internet of Things (IoT)… different names, but they all in a way converge on the same thing, looked through different lenses. This is one of the biggest emerging trends in network computing. Mega-trends like these happen once in a decade or longer. At Sun Microsystems where I spent 10 years, we used to have a tag line from start till the end which said: The Network is the Computer. What a vision that was, and it outlasted the company! Scott McNealy and others at Sun frequently would talk about the connected refrigerator. Today, it is a reality. Smart machines are here to stay and not only stay, but to grow and thrive. A company that sells any kind of hardware (washing machines to MRI machines, cars to planes, computers to mobile devices, the list goes on to virtually every physical object sold by someone), cannot afford to sit by on the sidelines and watch this mega-trend demolish them. These companies have to react and respond, and do it now.
GE, Cisco, and many other large corporations have long realized the importance of IoT. One of my favorite ads on TV (I don’t see it on anymore) comes from GE where all kinds of machines are coming home (I found this, but those videos are no longer active on youtube, however I did find this one that still works, sort of). Also check out what Dave Evans, Chief Futurist at Cisco says about this topic.
The point is, we have not only been talking about IoT for a long time, in fact, it has silently become a reality. Those that are still asleep at the wheel of their organizations can now ill afford to not notice and take some action.
I was talking to my friends Alok Batra and Jane Ren, who just launched their new company MQIdentity. Alok and Jane are well-known thought leaders and change leaders behind the Industrial Internet and M2M related transformations at Cisco and GE. Their knowledge of this space is quite vast and impressive. So I asked them, what is MQ? They explained it stands for Machine Quotient. They went on to describe two concepts: Machine Quotient (MQ), the technical measure of efficient machines. You want MQ to be really high. And Service Coefficient (SC), the business measure of service competency. For your organization, you need to find the right combination of MQ and SC to bring the maximum effective business value out. All this sound too complex? This is where I think Alok and Jane with their MQIdentity methodology can help.
For me the most fascinating aspect dawned on me when I read their white paper titled Technology-as-a-Service (TaaS), which they just released on their website. It is a well-written, thought provoking paper that lays down a solid foundation to think about how to transform your business into the new IoT economy. They argue that TaaS will be the path of technology transformations for tech industries into service-centric economy. The paper is quite detailed and there is a lot to think about. But if I were to distill it to a few key take aways, here they are 1:
TaaS gives you a new way to expand your proprietary IP and technology, by unlocking its access to new customer base leveraging other proven services model (IaaS, SaaS, etc.).
Your business is heavily disrupted by technology trends and you cannot stand by any longer and watch your customer base erode to newer solutions while you still have huge value locked inside your proprietary technology/products that can generate new revenues and expand in ways that you can’t otherwise think of expanding.
Your business is also disrupted by new startups who are going to offer the state-of-the-art solutions at a much lesser price, even though their solution may not be as excellent as yours, you will see customers erode to the “good-enough” solutions.
In short, this is a space (M2M/IoT) that I am totally fascinated about, and this space is going through tremendous innovation. As a result, in the coming years, there will be a new emergence of creative solutions, products, business models. In my opinion, Alok and Jane are right in the thick of it. So watch this space, it will be exciting!
1 – There are many other implications, consequences and micro-disruptions at work in this area, read their white paper or their executive summary for more details. Don’t just take my word for it.
LinkedIn advises you to only accept connections from people you know. That’s not how it happens in the real world, at least not for me. You accept connections from people who you think you know or a friend knows, or who you just met at a party and thought you should connect for future interactions. And over time, as you accumulate hundreds of connections, it gets difficult to tell who is who in your own list of connections.
For instance, suppose a friend of a friend requests to connect with me, I usually accept that connection based on the transitive trust factor. And later, when I revisit my contacts after a while, I forget who this connection is, there is no way to tell me the context around which the person became a connection. Of course I can look at the shared connections, but that is not any more insightful other than telling you that your friend is a mutual connection with this contact.
What I think is that LinkedIn is missing an opportunity to make their social network more smarter and richer. All they have to do is to provide the ability to (optionally) annotate our contacts with additional tidbits of information. This can really enrich the network intelligence and also provide more insights about my own contacts. Think of all the interesting graph searches you can do if you had such richer data…
So to start with, I just think two small pieces of information I can add to a contact at the time of accepting a new connection will be great help, like:
Why did I connect with this person? (e.g. “Wants to work with me” – someone was looking for a position and contacted me, I didn’t have an opening, but seemed like an interesting person to keep in mind for later).
Where did I meet this person? (e.g. “Met at Disrupt” – met at a conference and exchange business cards, and later connect on LinkedIn, and after a while, forget about it).
While at it, let me also tag the contact at the time of accepting the connection. I never ever have time to go back and tag them later. Even if I had, I wouldn’t know what to tag them with. Also tagging does not address 1 & 2 as it does not carry the same level of semantics.
If you are going to give an award, put together a group of real people, make up a criteria you can openly share with everyone, select the winners and announce the awards. Don’t just put up a silly survey with hundreds of possible categories and self-nominated awardee choices. Most people who go there will not know why and what they are voting for, with the exception of, my friend or boss or colleague asked me to do this. This is flawed.
I saw this tweet from Pivotal asking us to vote for them for the “Technology Innovator of the Year”.
First of all, I don’t know who is running this awards survey. It just said “V3 Awards Survey”. A bit of googling and I find this.
V3 launched its fourth annual Technology Awards on Thursday evening with an event in Mayfair for winners from previous years, and 2013 contenders. The night saw the unveiling of this year’s categories, and attendees were able to nominate themselves or others during the evening. But don’t worry if you weren’t able to make it along, as there is still plenty of time to get your firm’s entries in.
So anyone can nominate themselves in any category they want? This explains why the entries made no sense to me.
Now voting for Pivotal, I don’t even know what they innovated frankly. I am not saying that they haven’t, I just don’t know how to vote for them versus the other shortlisted options in that category like Nokia for Lumia 1020, Samsung for Galaxy Gear Smartwatch, etc. This list just doesn’t make sense.
Coming to Pivotal again, I really don’t know what they have innovated so far. Cobbling together a bunch of assets that existed previously into a new company is hardly innovation. Pretty much everything listed in the products page on Pivotal are pre-existing products under VMWare/EMC: Spring, RabbitMQ, vFabric, tc Server, Pivotal Analytics (was Cetas, acquired by VMWare), Cloud Foundry, Greenplum, GemFire, the list goes on. I do see one small piece of potential innovation that is new in this mix, HAWQ, as part of Pivotal HD, the Hadoop distribution offered by Pivotal.
Is that why I should nominate Pivotal for the technology innovator of the year and not Kaspersky Lab GrEat team for work on Flame and MiniFlame? Now, don’t get me stated on that, I have no idea what Flame and MiniFlame are, and I am not going to google them now. And no, I am not going to propagate this by publishing the survey link. There now, I got things off my chest. Feels better. I just need my 5 minutes back.
In most startups, you begin with a hypothesis of how you are going to address a need, and hope to build a product or a solution that can fulfill that need and build a business out of it. The time to market is one among the many pressures weighing down on your mind. How do you execute in this scenario?
Build: You can build everything on your own. Though you don’t know what to fully build yet, you will discover along the way.
Buy: You can find other technologies that address certain problem in a generic way and bundle (integrate) those to create your product. 1
Build and Buy: You can build what you think is the core competency that your startup is going to focus on, and buy/acquire the other peripheral components to create your full offering.
Sometimes building everything on your own is stupid, why would you want to waste your time and resources in solving problems already solved. And you can’t just buy everything and cobble together a unique product that is worth something. So you have to really come up with a Build and Buy strategy for your product to both be uniquely valuable and addresses time to market. Time kills startups.
So if you are in charge of building the product, you are bound to face all these pressures coming from your CEO, CTO, Marketing and Sales visionaries in your company. And this is where if you are not careful, you will end up building something that probably takes more time, is not elegant, and does not add any real value. And thus, if the product has no real value, how can you expect your company to establish itself as a viable business. The paranoid around you will start proposing ideas, most of them tend to be hare-brained, spur of the moment types coming from really loud voices in the organization. This is the time when you have to stand strong. After all, you are there to build a new product that offers a new and unique value proposition and change the dynamics in the market.
I have gone through this experience myself and wondered if I and our product team caved to these pressures and simply bundled what is out there already in the market, would we really be a product company or just another integrator? Where does innovation rank? Where is the IP that is going to make your company highly valuable and set us apart? As a product owner, you need to really think hard and have the courage, conviction and vision. You need to lead and convince others that what you propose to build will be more valuable than just cobbling together a piece of other technologies. At JackBe, we established very early on that innovation was going to be the key value my team was going to live (or die) by, and as such, we went against the opposing forces time and time again to innovate and really build the core technology and IP that established JackBe as the market leader in enterprise mashups and real-time intelligence. Our product gained customer admiration, won several awards and eventually, the company was acquired by Software AG because of the value of our unique product and IP.
I am not saying that your complete product has to be 100% home-grown, every line of code written by your team. There are plenty of FOSS products for you to build your unique product or platform. And there may be commercial products or components that you might want to OEM and license to build your product. There is no need to build those from scratch. That would be silly and a stupendous waste of time and resources. The key is to identify what is core to your differentiation and what is commodity, and then build your core by leveraging the commodity components out there and that’s how you balance the forces and address the time-to-market. Some common commodity components to consider (I only mention a few to just illustrate the example)2:
Server Side (Java Based): Application Server (Tomcat/Jetty/etc.), Application Frameworks (Spring, Akka, Play, etc.), Utility Libraries (Apache Commons, Apache Axis, Saxon XML Parser, etc.)
Visualization: Fusion Charts, High Charts, D3, NVD3, etc.
So my message to you is this. Don’t just bundle other COTS to create your product. It won’t help you to build a revolutionary new product. Instead make your own unique and secret sauce and mix it with what is commodity already to create a high value game changing product.
1 – Buy: For the purpose of this discussion, I am using ‘buy’ very generically. In some cases, you don’t really buy, but you can just acquire it via partnership. For open source products and components, you just use it (beware of what open source license they come with and their restrictions there upon. My favorite FOSS licenses are Apache 2, MIT, BSD. I tend to stay away from GPL (all versions) and LGPL. 2 – Free or Commercial: Note that some of the components are open source, and some are commercially available sold by other vendors in an OEM friendly license that you have to acquire from the respective vendor for a price.