Robert Scoble (still) doesn't understand enterprise software

Robert Scoble (still) doesn't understand enterprise software

Summary: Yet again, we open the thorny debate of enterprise vs. consumer software.


Several years ago, famous blogger, Robert Scoble, made outlandish statements about enterprise software that started a heated discussion and culminated in author, Nick Carr, declaring the debate a "FIRESTORM!"

Also read: "Enterprise software won't get you laid" Sexy enterprise software, part two: SAP and Workday Robert Scoble doesn't understand enterprise software Five principles of sexy enterprise software

As before, Robert goes berserk on this topic, comparing a deep enterprise application (Workday) to a lightweight consumer tool (Expensify). After expressing shock and outrage that the little Expensify app better meets his expense reporting needs than industrial strength Workday, Robert exclaims:

This is how sucky enterprise software gets chosen. The people who choose it are choosing the software that makes THEIR lives easier, NOT the lives of everyone else in the enterprise.

Four years have passed since those fateful days of FIRESTORM, but apparently, little has changed. Despite the increasing convergence of IT and consumer technology, even smart folks like Robert Scoble still misunderstand enterprise software.

If you are brave enough to handle it, here's a video in which Robert conflates enterprise and consumer software:

Enterprise vendors like Workday, SAP, Oracle, and NetSuite build solutions that serve a broad range of business processes and functional areas inside large organizations. To be useful in an enterprise environment, the software must integrate deeply (and hopefully seamlessly, but that's another issue) with many existing systems. Enterprise technology must also scale, offer robust security, high reliability, and so on. In contrast, consumer tools typically perform little more than a single function based on a very small set of features.

Enterprise solutions and consumer tools are not the same. Robert's discussion of Workday and Expensify, well intentioned though it may be, compares a broad enterprise system with a small utility program. When Scoble extrapolates assumptions about enterprise software based on a tiny subset of features in Workday, he commits a logical fallacy and falls prey to an attractive, but wrong, "sin of convenience."

In fairness, however, Robert raises an excellent point. Ideally, the user experience of enterprise software should be best in class on par with consumer apps. Because enterprise technology must retain deep, backend integration along with business process richness, accomplishing this goal is hard. Few vendors are up to the challenge, which is one reason enterprise software is so often difficult to use.

Scoble's complaints are also ironic because Workday offers the best overall user experience among the major enterprise vendors; the company's iPad app is also excellent.

That said, we all know that entering expenses is a pain in the ass and little Expensify does it really well - Workday should indeed look and learn.

Advice to all enterprise software vendors: Redouble efforts to offer best-in-class user experience while retaining the enterprise substance that makes you indispensable to virtually every major organization on the planet.

Topics: Enterprise Software, Software

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • I'm sorry...

    ...but I have to disagree with EVERYTHING printed by the author. ;-)
    • RE: Robert Scoble (still) doesn't understand enterprise software

      @IT_Fella fair enough, but what is your point of view then?
  • RE: Robert Scoble (still) doesn't understand enterprise software

    From an enterprise software selection perspective, Rackspace might have known about this expense-related functional inconvenience. However, it might have decided that this inconvenience was outweighed by the other benefits. <br><br>In any enterprise software selection project, there's going to be give-and-take. The intention shouldn't be to find the perfect software. This is an impossible goal, even in the rare case that the software was designed from the ground-up for your specific business ("perfect" as a subjective concept doesn't work, since different users will have different wants and needs). Rather, the intention should be to find the software that fits best as compared to the other alternatives (accounting for value and support, among other things).<br><br>Having said all of this, the vendors should make sure that they use customer feedback when developing software enhancements.
    • RE: Robert Scoble (still) doesn't understand enterprise software

      @Jonathang@... Thanks for the reasonable and balanced perspective!
      • RE: Robert Scoble (still) doesn't understand enterprise software

        @mkrigsman@... That's a sad defense of enterprise software. It can scale, be secure and robust, and all the rest, and still manifest great UI/UX if someone wants it to. That is a front-end issue. In the enterprise, people care too much about the IT department and not enough about the customer, who is the user. This is especially true in healthIT, a field I know well. The old enterprise providers are gradually being disrupted by newer software with better UX and less training time. Consider the cost of training people to use tools like Workday as opposed to the intuitiveness of Xpensify, and you will see that UX will win every time.
      • &quot; too much about the IT department&quot;!?!?!?!

        francine@... You have got to be kidding. The IT department is usually caught between two groups of small children saying "we want it all" and "we want it we wanted it yesterday". Those two groups of small children being senior management on the one hand and end users on the other.<br><br>See my posting below. When the boss buys the magic beans, IT is left with no choice but to climb the beanstalk and kill the matter how stupid that may be.<br><br>In the end, no one is happy - least of all the IT department.
      • RE: Robert Scoble (still) doesn't understand enterprise software

        @mkrigsman@... It is an interesting topic, though I'd stipulate that Mr. Scoble is hardly alone. While he may be guilty of fallacious logic, when it comes to the GUI and functionality of the software in question, I'd submit the question: who among us truly knows? The user experience is never the same when viewed from two seats, let alone many. There is simply too much subjectivity in the user experience to allow truly objective evaluations. The use case can be built on ladders of logic and still never achieve universal acceptance. Ask me about Windows & Office 2010. On second thought...don't ask.
  • The real way Enterprise Software is purchased

    At least in several of the places I have worked, it starts out with some Vice President getting a lot of buzz words thrown at him by some slick salesman. The VP is then convinced that buying this product is the solution to all his problems. He sees a lot demos where said salesman shows him a lot of mockups where developers can "drag and drop" and create his application in minutes. So he buys what I refer to as "the magic beans". <br><br>The magic beans get turned over to the real IT people and developers who somehow burn the midnight oil and put in lots of unpaid overtime, blood, sweat and tears, and eventually force fit the magic beans into the organization. They do whatever they have to to make it work, however bad it really is.<br><br>When all is said and done, the VP points at the project (assuming it did not fail which is sometimes does) and tells his boss what a great decision they made. The boss agrees and suddenly the VP is a Senior VP with a fat pay raise and probably a bonus.<br><br>The IT staff and developers are grateful to still be employed, knowing that had the project failed - they would have taken the blame.

    PS: Epilog to the story. End users hate the system because it is a "force fit" to the organization. We had to whittle off the organization's corners to make it fit in that round hole. But do users hate the VP who bought the magic beans? No, of course not. They hate their IT department.
    • RE: Robert Scoble (still) doesn't understand enterprise software

      @cornpie well put.
    • RE: Robert Scoble (still) doesn't understand enterprise software

      @cornpie your comments are amusing, and as far my own experience suggest, quite accurate. Disclosure: I am not an I.T. Admin, and suspect that I have been the bedazzled V.P. in your fable.
  • RE: Robert Scoble (still) doesn't understand enterprise software

    francine@... I agree with you that there is no excuse for poor user experience!!! That said, there is also a larger context of enterprise software, which is where I have an issue with Robert's comments.
  • RE: Robert Scoble (still) doesn't understand enterprise software

    The problem is, both are right, depending on the point of view, and the prioritization used in the decision.

    The user experience of most "enterprise applications" is utterly abysmal. Anyone who has been on the end-user side of an Oracle, SAP, Dynamix, or other "enterprise systems" implementation basically suffers through the entire user lifecycle with bad design, compromised implementation, and tremendous user-facing complexity that should have never been shipped in the first place.

    How did all that complexity get there? IT departments were saddled with a "do everything" product that was poorly implemented and doesn't actually deliver much of the "flexible, easy-to-integrate" capabilities that sales promised, and executive suits bought.

    In the end, the buyer (exec, procurement) is duped, the user suffers every day they have to use the app, and the implementor (IT) is screwed from day one to EOL.

    Thus goes the Enterprise software lifecycle.

    Now, trying to organize a collection of consumer-built apps in to a coherent enterprise financial system capable of passing SOX compliance?

    Honestly, I'm not sure which is harder/easier, or better/worse.
  • User experience has more to do with implementation than design.

    I've worked with a specific enterprise system for a number of years and seen it used in over 100 different businesses. In some places, I've found users absolutely hated the system, where in others, they would fight to the death anyone who would even suggest moving to something else. After working for several months with both types of users, I would find that the difference always seemed to be a combination of three things:

    1) Users in the business that "loved" the software also tended to have a better overall understanding of their entire business (rather than just their own cubicle). Users that hated the software tended to only understand or even really care about their own position in the company. Given Robert's rant on his expense software task, he probably falls into the latter category of these users.

    2) The businesses where users "hated" the system tended to be those where the system was implemented incorrectly or incompletely in some way. In many cases, users would complain that the system "wouldn't do what they needed", but when looking a bit further, it actually turned out that the system COULD do what they needed, but it wasn't SET UP to do what they needed.

    3) Users that "loved" the system tended to get a lot more training on it either when it was implemented or when they were first hired than users that "hated" the system.

    While not always the case, in many and maybe even most cases, the user-friendliness of enterprise software isn't actually dependent on the design of the software itself - it's dependent on the implementation of it. Now software selection is part of "implementation", but the selection has to follow the design, not the other way around (another major flaw in some implementations).

    And yes, THIS is why IT gets blamed - and rightly so!
    • RE: Robert Scoble (still) doesn't understand enterprise software

      @daftkey Sorry, but I think you're confusing implementation with deployment. There's a big difference.

      Blaming IT, who usually performs only part of the deployment phase is a cop out. IT is rarely consulted adequately (key word here) during the product selection process.

      Usability (or "user friendliness") is a part of the Design phase, not the Implementation or Deployment phases. Most accounts of the Software Development Cycle explain this.

      End user training and support occurs during Deployment. Most successful software developers incorporate these expenses into the cost of the application. Trying to compensate for design flaws through training and support won't fly in most cases. That's putting lipstick on a pig.
      • We can agree there are many forces at play here..


        <i>Blaming IT, who usually performs only part of the deployment phase is a cop out. IT is rarely consulted adequately (key word here) during the product selection process.</i>

        I do agree putting all the blame on IT isn't exactly fair in the context that cornpie has provided. Having said that, it's just as much a cop-out to blame an IT failure and poor user experience on a slick salesman and a witless executive.

        Based on your own response, however, I still think IT should shoulder some of the blame here. If you're not being consulted adequately, then there is at least a perception problem with your IT department, and there's probably a REAL problem with your IT department. The first things you should ask yourself are:

        Why doesn't IT have enough authority to stop projects that they "know" are a poor fit to the company? Does IT really know the company enough to make that call? Does the executive team have enough faith in IT to listen to objections that are raised? To what extent *should* IT be involved in software selection?

        In a lot of cases, the real answer is either the IT department really *WAS* consulted on a project as adequately as possible, but didn't step up to the plate, ask enough questions, and deliver the required dose of reality needed to put the project on a successful path, or the IT department is one where everyone thinks of technology before business requirements, in which case they have a real problem seeing a project from an overall business perspective and aren't really able to provide a useful contribution to the process until it's time to fiddle with the technology (at which time, it's already too late to put the brakes on).

        <i>Usability (or "user friendliness") is a part of the Design phase, not the Implementation or Deployment phases. Most accounts of the Software Development Cycle explain this.</i>

        As a lot more software is being purchased "off-the-shelf" rather than being custom-designed, user experience really has to be included as part of the selection phase of the implementation.

        Now in some ways, that allows those in charge of selecting the software more visibility in the real give-and-take of the software selection process, but as others have pointed out here, sometimes an annoying UI feature will be accepted because that negative is offset by the ability to, say, record costs against projects in the company's ERP, update a company dashboard immediately, and allow executives to identify business exceptions and make important decisions sooner.

        Usually, "bad" UI is really caused by a conflicting set of business requirements, each trying to be satisfied by a part of the system where, if one were turfed, the UI would be much simpler.
    • Never been the IT guy that's getting blamed, eh?

      @daftkey If you had you would know that the system is what it is because the combination of slick salesman from the vendor trying to make money and company executives out to make a name for themselves are in collusion to have the company buy the product in the first place. They then turn the mess over to IT with unrealistic expectations about what can be accomplished with the time and other resources available.

      IT is just making the best out of the steaming pile of crap its been handed and trying its best to make things work in spite of all the bad decisions that took place before they were even involved. And when it comes right down to it, its the executives that hire and fire and pay our salaries and so if we have to make a choice between them and end users, guess what happens. We do like to eat you know. End users can hate us and yell at us and treat us like crap - but they usually can't fire us.

      So its right to blame IT? Well I can't say never. But it's not the norm.
      • Nope - I'm probably the guy you hate the most...

        @cornpie ..

        In your words, I'm the guy serving up the steaming pile of crap, along with the overall goals of the implementation and the expectations of the organization overall.

        Really, I'm the guy that helps select the software with the overall ideas of what the software is intended to do. I'm also the guy making the decision of what is more important than what in the overall scheme of things.

        Unfortunately for IT, I'm also the guy that's going to demand that any concerns that are brought up with the implementation of the software be communicated in the context of the overall scope and goals of the project, not just some complaint by a few vocal users who are used to a different UI, a different process, or "The Old Way" of doing things. If IT cannot draw a distinct connection between a perceived flaw in the software and a specified goal in the project plan, your objection is going to be thrown on the "nice-to-have" pile.

        And you know how often the "need-to-have" pile becomes small enough that the "nice-to-have" pile even gets looked at.
  • RE: Robert Scoble (still) doesn't understand enterprise software

    As an enterprise software guy, when I read this article I naturally agreed with Michael. But then I watched Scoble's video and changed my mind. The trouble with Michael's defense of enterprise software here is that it's generic. What makes him so sure that Expensify isn't robust, scalable, secure, etc.? What makes him so sure that the poor performance and UI of Workday called out by Scoble are caused by the presence of those features? Sometimes bad software is just bad software. Scoble is right. Workday should be doing it better. Aneel Bhusri and David Duffield need to watch Scoble's video carefully and do some real soul searching. Not just "how did we let this happen?, but "what are we going to do to fix it?".
  • RE: Robert Scoble (still) doesn't understand enterprise software

    Your claim that enterprise software "scales" and is "secure", especially when compared to a "consumer" grade too, is more than a little laughable to anyone who is familiar with the security audits and the devops processes of "enterprise" vs modern "consumer" tools.

    The popular piece of "enterprise" software will usually have at most a few ten thousand users, for a VERY large org, and more likely just a few hundred, at any given installation. A popular "consumer grade" tool is still a mere toy at few hundred users, and can often go to hundreds of thousands to millions. Bluntly, enterprise software scales poorly, and in ways that make a LOT of money for Oracle, Microsoft, and Dell. And their vaunted "security" is poor at best. They are riddled with exposed databases, Bobby Tables weaknesses, scripting attacks, and require obsolete server and client versions.

    By the metrics you say that enterprise software is better, it is usually actually far far worse.
  • The Middle Ground

    Good post, Michael. I especially appreciate the fairness and civility with which you make your argument. Many writers would have trashed the person with whom they disagreed; instead, you have treated Robert Scoble with respect and taken the time to understand his point of view.<br><br>I think there is a middle ground between your and Robert's views that is about to manifest itself in enterprises. In fact, you signaled it with these words:<br><br>"Enterprise technology must also scale, offer robust security, high reliability, and so on. In contrast, consumer tools typically perform little more than a single function based on a very small set of features."<br><br>I believe we will soon see the emergence of single-purpose, easy-to-use applications within enterprises. Imagine exposing only the expense reporting functionality in Workday's suite as a stand-alone application that is designed as elegantly and simply as Expensify's. Most employees wouldn't have to navigate through Workday's standard interface, and all of its functionality, just to report expenses. Instead, they could work with an intuitive interface that captures their expenses without exposing them to the complexity of Workday that is a necessary result of supporting multiple business processes and integrating with other enterprise systems.<br><br>This approach will be a win for everyone. Employees will enjoy the experience of using a simple, well-designed application. IT staff will be able to deploy the enterprise-grade systems that they require AND see them actually used by the organization's employees. IT will be able to "control" the simple application interfaces (in both original and upgraded versions) by provisioning them through an enterprise application store.<br><br>In short, the consumerization of IT can be the solution to this debate. Use consumer software interface design as the model for enterprise application front-ends and make the necessary back end system functionality and complexity invisible to the users of those applications.