• March 22, 2018
  • If you study economic cycles you can watch the evolution of a disruptive technology throughout its lifecycle from a specific product, to a competitive industry. The last phase in the evolutionary chain is often the formation of a utility. For example, over a couple of centuries we’ve seen the evolution of electricity from a curiosity, to a business, to a group of public companies. Along the way there are the inevitable mergers and acquisitions to enable a winnowing field of competitors to achieve the scale needed to compete in very large markets.

    It wasn’t just the electric industry that went through this evolution. The telephone, gas, and cable industries did in their own ways. Local or reginal utilities that provide sanitation and water still dot the landscape too. Banking is in a similar position that is manifests differently. In fact, any industry that attracts the term “too big to fail” is showing signs of utility status.

    When your business becomes so big that it affects large segments of society it can’t be allowed to fail lest it crater the economy or cause massive disruption injuring many people. At that point government has a compelling interest in preventing failure and along with that comes regulation of the riskiest corporate behaviors.

    The latest example to hit the radar might be social media which has completed many steps of the lifecycle with blistering speed in just over a decade. This speed notwithstanding, we are now at a point where what happens in social media affects all of us.

    Regulation is a thorny issue wrapped in individual freedom. But it is also a logical way out of an impasse. The recent interference in the US election, which all the American intelligence agencies have confirmed, is a proof point that social media is now a utility and needs some form of regulation.

    In a recent article in Wired, “Bad Actors are Using Social Media Exactly As Designed,” writer Joshua Geltzer makes the point that popular social sites including Facebook, Twitter, and Air BNB, all provide tools that enable users to find and segment groups that are best subjected to targeted messaging. His point is simple, the bad actors didn’t pervert social media or hack its code. They simply used the tools provided to mount a campaign to upend a US election and there’s evidence of similar activity elsewhere.

    By this measure, social media is now too big to fail; it is too essential to a large segment of society and its potential excesses must be managed so that it does not consume the society and the users who depend on it. Geltzer’s article clearly, but inadvertently, makes the case,

    When Russia manipulates elections via Facebook, or ISIS recruits followers on Twitter, or racist landlords deny rentals to blacks and then offer them to whites through Airbnb, commentators and companies describe these activities as “manipulation” or “abuse” of today’s ubiquitous websites and apps. The impulse is to portray this odious behavior as a strange, unpredictable, and peripheral contortion of the platforms.

    But it’s not. It’s simply using those platforms as designed.

    So, what would a social media utility look like and how would it be different from what we see today? First off, social networks should be regulated with as light a touch as possible. Grandmothers sharing baby pictures shouldn’t have to change their use habits, for example. Second, to achieve positive ends, regulation should be implemented at two distinct levels, the source and the periphery.

    At the source, regulation comes down to access for any person or entity with a beneficial and productive need for the utility’s services. In the electricity markets this means stable pricing for all and a commitment to serve as a common carrier. It wouldn’t be much different with social network regulation; the key is beneficial and productive use.

    Common carrier law began in railroads and shipping. In return for its use of public lands and roads, the carrier commits to serve all parties equally. In broadcast industries (radio and TV), slices of electromagnetic spectrum play the role of roads  that the broadcasters use as grants (licenses) from the people. In transport, the waterways are also owned by the people, so are the roads, and railroads have historically received government help because they provide a useful service to society. It goes on, but you can see that source regulation amounts to giving all participants a fair shot at using the public’s assets and insisting on beneficial and productive use.

    Regulation at the periphery takes on a different cast, notably in America where so much of modern utility regulation evolved. A great deal of peripheral regulation occurs through certification and licensure. People interested in careers involving one of the utilities often serve apprenticeships, learning from a master before earning journeyman’s status. They must also pass tests to prove their knowledge and skill.

    Barbers, beauticians, and other personal services professionals go to school and sit for certifying exams. Other professions are similar. Doctors, dentists, lawyers, and many others must take many years of education, pass tests, and serve different forms of internships before practicing on their own.

    The point here is that we already regulate the day to day best practices of many industries. We do it with a light touch and in the interest of the culture and the society functions quite well despite, or more likely because of, this light approach to regulation.

    My take

    Perhaps the time has come to consider lightly regulating parts of the tech industry, something we have never done. But the age of information and telecommunication, a 50-year economic cycle called a K-wave, is reaching its natural endpoint and that’s often when utility status and regulation has come to the forefront in prior cycles.

    Elevating social media use to professional status, seems a logical thing to do. Establishing a certification or licensing process plus capturing a user’s license number when accessing some of social media’s higher functions would give an uncomplicated way of keeping bad actors out of the networks or at least making them traceable. In case you are wondering this is the basic process of getting a building permit.

    This approach need not apply to lower level personal use. But trying to reach millions of people on a social network is functionally like climbing a utility pole and messing with the wires. For this one should need certification.

    Last point, part of certification in any industry is training in the ethical use of the tools and techniques of that industry. As a society we have not engaged in such a dialog for social networks yet, but one is overdue.



    Published: 2 months ago

    I’ve recently looked at the changes in markets and companies as industries grow and as a disruptive innovation commoditizes. Generally, prices come down because the innovation becomes commonplace and competitors fight for every scrap causing margin erosion. Eventually markets equilibrate and a monopoly or small oligopoly sets in. Typically, as Geoffrey Moore pointed out, mature markets have three competitors, the leader that gets most of the business, a fast-following challenger, and one or more niche players.

    We can see this playing out right now in various industries. Mature industries like databases have an oligopoly set up with Oracle and IBM leading. SAP with Sybase xxx is one of several niche players along with MySQL and SQLServer. Recently new approaches to data storage have come to market such as no SQL databases that offer a different approach. Amazon has developed an in-house product that has been a discussion topic for people thinking of dethroning Oracle but that’s not likely. More likely is that Oracle and IBM will be the database leaders until the planet no longer needs databases. It could happen.

    On the way to oligopoly we see vendors making mistakes that could cost them their leadership positions. Mature market leaders have figured out how to put customers in the forefront and their solutions look more like a broad array of services than products. That’s because they have come to understand the role of whole product and are willing to do whatever makes sense to keep customers happily engaged.

    There’s one practice, though, that flies in the face of whole product and keeping the customer satisfied. It’s the software audit. Essentially a vendor builds into its contracts the right to periodically inspect a customer’s systems to understand use patterns and to ensure that the terms of its licenses are adhered to. Typically, a contract for so many seats, or applications, or gigabytes can remain under a preset limit but should not go over. If there is an overage the customer would owe a higher license fee.

    All of this seems fair in the abstract but in practice customers resent the intrusion of an audit especially when the audit does not turn up a flagrant violation but instead identifies a possibility of one. Vendors have been known to bill on the possibility and that can cause, shall we say, disharmony in the customer base.

    In the future, the software audit will likely be seen as an artifact of the licensed software era. Subscription vendors and their customers don’t have the same issues because it’s clearer when the customer needs more seats and they make a purchase.

    But the software audit is here today and in my mind, it represents a mistaken approach to revenue generation by mature market vendors. It also opens a legitimate seam for the competition to exploit. Recalling that markets trend toward commoditization and that features and functions tend to converge on a market ideal, the cost of attrition, even in a licensed software situation, declines over time. So, the risk of customer attrition over some aspect of the whole product, i.e. policies and procedures, increases.

    You can see this scenario playing out in the document management space as Nitro, a document management company with a focus on whole product does battle with the leaders. Nitro is not a small upstart. They’re well over 10 years old and have a customer base of roughly 650,000. Nitro differentiates itself on a UI that’s very Windows-like and on price. But it doesn’t simply offer a lower cost approach to document management. Its flexible licensing policy avoids the software audit process. This seems to matter to its customers who want to deploy technology to all members of the company freely, without obsessively checking license allocation or fretting about fines each time a new hire comes on board. For them the audit can be a show stopper.

    My two bits

    Regardless of the size of the market or its installed base, a company needs to grow, preferably faster than its organic rate—when customers grow their business they buy more licenses. But in mature markets growth is a zero-sum game. For one company to grow it has to take customers away from competitors; usually, that means taking from the market leader(s) who have a majority of the customers.

    So the software audit, as useful as it might be for generating incremental revenue, has a downside. It is not the face of whole product that a company in a mature market ought to present because it can easily prompt a rush for the exits. That’s why so many vendors avoid such contention and focus on product line extension and new product development. Those activities are harder to pull off than swooping in with auditors but they are ultimately better ways to hold onto customers and grow markets.

    Finally, you might say that cloud computing with its per-seat pricing can avoid all of this but that’s not the point. A vendor that can adopt audits can also find ways to throttle use and charge dearly for increasing use. The real point is developing modern approaches to customers that generate revenue on actual products and services delivered rather than gotcha gimmicks.

    Published: 2 months ago

    In 1906 President Theodore Roosevelt signed the Pure Food and Drug Act into law. It had been a long time coming and corporate interests had done their best to derail and delay reformers who advocated the contents of the act. The meatpacking industry had been a major focus of reform as was the over the counter drug industry, each for different reasons.

    Meat was unwholesome especially in the warm weather. Cattle were driven by cowboys to railheads in Kansas City Missouri and then transported live by rail to Chicago which became a major distribution point after the American Civil War. Meat packers slaughtered cattle in Chicago and shipped to markets in the east. This prompted the need for refrigerated railroad cars or reefers. At first, reefers were cooled by ice and not necessarily very well. People in the east got sick from tainted meat and Upton Sinclair’s novel, The Jungle was a story about abuses in the industry that led to unsafe products.

    Over the counter drugs contained high quantities of alcohol and opioids and their makers over promised what they could do for patients. Newspapers, which relied on advertising revenues from the drug companies, were enjoined from expressing support for legislation regulating the drug industry on pain of losing their ad revenue.

    Muckraking which we might call investigative journalism today, came into its own and crusading journalists like Sinclair and many others published exposés of abuses that led to the Pure Food and Drug Act as well as the Federal Meat Inspection Act which Roosevelt signed into law together.

    A century later, the meat supply still has occasional problems, but no one tries to hide the truth and both producers and regulators work to correct issues when they arise. Also, producers rigorously test over the counter drugs both for safety and efficacy before going to market. Libertarians and conservatives might want to complain about the added costs involved or the rights of participants in free markets in other situations, but they are largely silent here.


    The social networking industry today is in a similar position to meat packing and pharmaceuticals more than a century ago. Our news is full of stories such as Cambridge Analytica stealing 50 million Facebook profiles while that company did little to protect its users. The US intelligence services all point to Russian active measures in interfering with the 2016 presidential elections by using the tools available with social networks as they were designed.

    All of this is awakening many people to the understanding that social networks are not free because they collect a fee in kind; namely consumer data which they use to model profiles and sell to advertisers. Some have suggested that in the face of these revelations that Facebook might be doomed or that social networking in general may be. But nothing could be further from the truth.

    Social networks are at the same place that meatpackers and drug companies were in 1906. Regulation was and is a way to enable the people to influence markets in such ways as to not interfere with any single company. Regulations apply equally to all those regulated and businesses should, after a period of adjustment continue as they were in the newly regulated markets.

    My two bits

    Regulation can take many forms. I’ve previously advocated for treating the social networks as utilities and that’s certainly one approach. Another is simply to put  forward legislation like the Pure Food and Drug and the Federal Meat Inspection Acts of 1906. The Pure Food and Drug Act ultimately spawned the Food and Drug Administration which regulates industries without treating them as utilities.

    The history of regulation in the US and much of the west has involved setting standards and fines for violation. Government agencies randomly sample many aspects under their direction and in some cases government meat inspectors work at slaughterhouses. But the regulation and enforcement is largely something that individuals take on because they are good for business.

    The point is that regulation can work well with business and it doesn’t necessarily mean the end of an industry or crippling rules. A level playing field is good for all  and often increased the size of markets.

    Social networks, and the Internet more generally, need some regulation right now for the good of the society they serve. There are approaches and models that we can adopt. The current issues we face are not unique in history and we should heed its lessons.

    Further reading

    Time to call it—the social utility has arrived. That has important consequences 

    The Social Utility


    Published: 2 months ago

    I’m going to need help with this. Mulesoft, a small tech company with less than $300 million in revenues is being purchased for $6.5 billion by Salesforce.com. Salesforce offered $44.89 per share, a 36 percent premium over the market price. Others are already saying that the transaction price represents more the value than the actual price. I don’t know, I am not that smart.

    What I do know is the Mulesoft is an integration platform that brings together data from other clouds and even legacy apps. That’s really important as companies all over the landscape race to build what I’m calling the information utility. For that to happen you can’t have data sitting outside of the utility because, presumably, it’s all important, and it is.

    In fact we don’t have a clear idea of what the information utility will look like or do yet. Yes, we can see the broad shapes and yes, enabling apps everywhere to access and share data seems like a no brainer. That’s what a utility like this should do.

    But now we have to take things a step further. In light of the blowback from the social network shenanigans associated with the 2016 election we’re beginning to see that Mark Zuckerberg and Sheryl Sandberg are doing a revival of “Home Alone”. Adults are not in the room and all heck is breaking loose in social media.

    The social networks you see today are dinosaurs, kiss them good bye. It takes more effort to pull a plumbing permit to work on your new kitchen than it takes to get permission to mess with 50 million Facebook profiles. That’s wrong and unsustainable.

    I can see a day in the near future when we’ll need Mulesoft to integrate social networks with encryption and storage systems that protect customer identity data. The systems that encrypt and secure identity aren’t around yet but might be coming to market in a year or two.

    We’ll need those systems as governments in the EU and US ponder ways to enhance GDPR and lock down personal data in an effort to prevent future cyber war.

    Here are some links to articles that flesh out my meaning:

    Cyber War

    Social Networking as a Utility

    So, grudgingly, because $6.5 is a lot of money, I say the Mulesoft acquisition makes sense in a roundabout way. It speaks to a vision of the future at Salesforce that I can only speculate about as an outside analyst covering them. That said, I’ve been right about stuff like this before.

    Published: 2 months ago

    Oracle is in a legitimate exponential growth phase and not for the first time in its long career. Like a startup it is growing much faster than the organic growth rate of the economy or its primary market because it has some new products that are highly desirable including cloud offerings and a unique fully autonomous database. It is also growing quarter by quarter and year over year—this is not a one-time thing. But unlike a typical startup that must claw and fight for every new customer, Oracle has the opportunity to sell its new products into an existing customer base that is hungry for improvements in their price-performance ratios.

    Yesterday co-CEO’s Safra Catz and Mark Hurd plus Chairman and CTO Larry Ellison broke down the numbers for the just completed quarter and they were impressive with most of the growth coming from cloud computing including infrastructure, applications and platform services.

    Total cloud and conventional software revenues were $8 billion, and Catz gave a blizzard of other positive numbers including,

    Cloud staff revenue for the quarter was $1.2 billion, up 21% on a GAAP basis from last year in constant currency with — on non GAAP basis with Fusion Cloud revenues up 52% in constant currency. Cloud PaaS and IaaS revenues for the quarter were $416 million, up 24% from last year in constant currency. Cloud PaaS and IaaS revenue, excluding legacy posting services, saw growth of 49% in constant currency and 56% in U.S dollars. As legacy hosting services become smaller part of total PaaS and IaaS, the underlying growth of PaaS and next generation IaaS will be more visible.

    You can find the whole announcement here as well as a transcript of the call here

    Here are my observations.

    First, only about 15 percent of the customer base has even begun moving to the cloud products meaning that more good news will likely be coming in future quarters and for some time. Much of the momentum comes from net new sales.

    Second, the revenue gains are likely to accelerate as the company builds out its infrastructure support. Oracle announced earlier this year a drive to deploy xxx new data centers to support its Infrastructure as a Service (IaaS) initiative. Without those data centers being deployed and ubiquitous it can’t sell much infrastructure though even there it generated $416 million last quarter gaining 24 percent year over year.

    But more is to come, and, now, infrastructure is a potential throttle. For instance, as the company will soon be able to service more customers with the same hardware or as Hurd put it,

    As we fully deploy database multi-tenancy in our staff, let’s say. We double our capacity without spending one penny on hardware. We can help twice as many customers, twice as many transactions, twice as many users without spending one dollar.

    Third, the new edition of the core database product, which is fully autonomous, is now generally available offering a level of efficiency and security unparalleled in the industry. On the call Ellison said there’s more to come.

    Over the next few months, we expect to deliver autonomous analytics, autonomous mobility, autonomous application development and autonomous integration services.

    If anything, Ellison may be underselling the benefits of the autonomous database when he talks about the labor-saving aspects as money savers as when he said,

    Our highly automated suite of autonomous PaaS services reduces cost by reducing human labor and improves reliability and security by reducing human error.

    For enterprise users, labor is cheap and although reducing human error is important the speed with which the autonomous database acts to self-correct may be the more significant benefit. Ellison revealed at OpenWorld that his customers can take upwards of 13 months to install database patches leaving their systems unnecessarily exposed for that time. The autonomous database self-patches meaning that users can install fixes much faster. Almost any cost associated with upgrading to the new database when compared to reputational hits and law suits over compromised data are insignificant in comparison.

    Fourth Oracle is buying back shares and has a war chest of $70 billion. In the last quarter Catz said the company bought back $4 billion in shares, a process that is ongoing. The implication is that revenue and earnings numbers that are measured on a per share basis will likely improve simply because there are fewer shares outstanding.

    Also, with so much cash on hand Oracle can pay for important things like further IaaS deployment and marketing to further push its products into the marketplace.

    My take

    The notable difference between Oracle and a startup is that the company can afford to be its own venture capitalist. Another difference is that in addition to being able to attract net new customers, it has a huge installed base to bring current. They’ve previously said the process could take 10 years or more so it is still early days.

    Unlike many other players in the market, Oracle has seen the need for greater security measures to combat the threats in the world today. Oracle is the logical player to deal with security since so many of the world’s applications are based on its technology infrastructure.

    It will take some time before the technology diffuses through the industry and IT becomes a more secure environment. But it won’t take ten years. Plenty of companies will need to upgrade their applications independently of what Oracle does now that new tools are available. As it makes this turn to cloud computing Oracle is laying the foundations for a global information utility that can address today’s challenges.

    But Oracle can’t do the job alone. Current news demonstrates that the Internet and social networks are no better at security than a screen door against a winter gale. The information utility will need more encryption and professionalization of its user class. In addition to having proficiency with the technology, users need to be easily identified, perhaps through license numbers, and certified in the ethical use of the technology.

    You might not realize it but a plumber has to get a permit before working on your natural gas feed. It’s not an onerous process if the individual can demonstrate (via his or her license) the basic competence to do the job. We’re getting to that point in IT right now.



    Published: 2 months ago