In Search of Enterprise DNA for Social Software
Posted by Bob Warfield on May 13, 2009
Some Enterprise Irregular bloggers recently raised some interesting questions about CubeTree, a new Social Software startup for the Enterprise. In general, they were reacting to a video of SAP talking about the product that they regarded as a bit too enthusiastic (some called it creepy in that cultish sort of way). They went on to argue that the video didn’t answer a lot of important questions one would have about any startup, but that are particularly important for Enterprise startups. Rather than pick exclusively on CubeTree, there was concern about whether any Social Enterprise startup could make it in the wake of these challenges:
– How are they differentiated? CubeTree is fast following something, but what? Is it enough just to bring together more Social features that others have already implemented in various point solutions? Only if we’re early in the hype cycle.
– How do they get enough credibility so big companies will buy them? CubeTree have started well by getting adoption at a Lighthouse account like SAP. But will the video play well? Will they get other Lighthouses? Is it a one off based on some personal relationship?
– How do they protect themselves against competition? The IP question. Note that there are two strong protections available to startups: IP is the obvious one, but the right partnership strategy can also be very effective. Hence my company’s partnerships with Salesforce and Oracle.
– Is their timing right? Too soon and the big guys have to do it before the startup has any critical mass. Too early and nobody has a clue what they’re on about.
Startups all grapple with these to a greater or lesser degree. I’m on my fifth startup with Helpstream, and I can tell you they are absolutely critical questions to answer. But there are other questions specific to the Social Enterprise space that I think are more interesting. What I see as a bigger problem when selling to the Enterprise among these Social startups is a lack of Enterprise DNA, and equally, a lack of savvy about platforms versus applications.
To an extent the Enterprise DNA problem for Web 2.0 companies may be a result of youth. How much can you know about Enterprise software if you’ve never built any and never worked for an Enterprise? In the case of CubeTree, they actually do have some good Enterprise experience on board, which is rare for this sort of company.
If you look at Enterprise software, there is a certain critical mass of features needed to get through the IT gatekeepers (if you view it cynically, or simply features needed to be effective in the Enterprise). They’re obvious things that people working in the Enterprise world know about almost instinctively. Things like how the security and permissions features have to work, for example. A lot of Social software has extremely vestigal permissioning that most Enterprises will find unacceptible. How many, as they review a promising new piece of Social software spend time to really understand permissions as opposed to the streaming or microblogging features? Another big issue discussed among the Irregulars was data import and export. The export, in particular, is a matter of particular concern regarding SaaS and Cloud applications.
There is a lot else and most of these folks have never had an IT department conduct due diligence on their products. They’re suprised to learn that consumer-grade LAMP stack software doesn’t impress IT.
The application vs platform issue is another one. For a long time I’ve made it a practice to always build a platform and not just an application. Do this not because you want to go sell the platform, the market will decide whether you ever get the opportunity. Rather, do it because it makes evolving the product dramatically faster and easier. Plus, it is more likely to lead to real IP and the kinds of capabilities IT and the big accounts will insist on.
It is very hard to build platforms using LAMP as well. LAMP is your platform. The good news is your productivity is boosted through the simplicity of LAMP and the ready availability of lots of off-the-shelf components you can plug in. But the bad news is you’re essentially trying to build something industrial grade with scripting software. So we have PHP and PERL based Social software. Java would have been preferred, with Ruby on Rails as a nice alternate.
This DNA goes well beyond just how the application is built, and into how it is sold and what it does. The CubeTree video was way over the top and gushingly frothy. But that’s been the rule for Enterprise 2.0. The trouble is, except for those individuals in large corporations interested in self-promotion (and there are more than a few of those around), Enterprise is not gushingly frothy. It speaks in sober terms of Business Process and Return on Investment. It asks, in short, for you to demonstrate real value and not just enthusiasm.
The ability to deliver that real value in the cold light that separates the hype from the reality is what the Enterprise craves for full adoption. Anything else is just good fun and fashion. The blogosphere is beginning to conclude it’s about time to separate ourselves from the seamier hype-laden side of the business. I completely agree with that sentiment. Various vendors are dealing with it in ways that range from getting less and less ink, to being in outright denial and attacking the Enterprise and established institutions like Business Process.
At Helpstream, we’ve addressed both the issues of permissions and import/export with Enterprise-grade solutions (we have a lot of Enterprise DNA!) as well as a whole lot more. We have a platform approach to architecture that has greatly assisted our ability to rapidly evolve the product over time. We’re seeing firsthand the benefits of Enterprise thinking for the Cloud, SaaS, and Social Software. But most importantly, we have a singular focus on creating a powerful alloy of tried and true Business Process with the new technology of Social Software to produce a combination that is more powerful than the sum of its parts. We measure that power not with enthusiasm, but with ROI. In fact our customers demand it. Helpstream sells to Customer Service organizations. If you’ve ever dealt with one you’ll know that Service professionals of all kinds (including Professional Services) are some of the most hardened skeptics out there. They have heard it all. They are extremely process and metrics focused, because that’s how you run a Call Center. They’re unwilling to trust their business to enthusiasm for the most part. Their thinking is along the lines of Robert Heinlein’s quote, “If it can’t be expressed in figures, it is not science; it is opinion.”
Towards that end, analytics is the constant companion to Business Process, and another one of those things Enterprise DNA will insist you must have out of the box. To satisfy our customer’s needs for metrics, we’ve built comprehensive analytics into the product, and we’ve thought hard about how to set up the analytics so any customer can see a real time view of their ROI. This is more than just a sales gimmick. Customers and Helpstream use these reports to tune up the performance of their Customer Service communities. In fact we’ve been able to develop several innovations and measure their real impact since introducing the reports, and there are more on roadmap to come.
Depending on where the market really is on Geoffrey Moore’s Chasm-crossing Rubicon, CubeTree and it’s video may just be the Swan Song for how this software has been sold to date. They’re totally horizontal, not focused on any particular business problem or business process, and they have plenty of gushing hype. Nice product, but their tactics seem timed poorly versus where this world is relative to the Chasm.
More Enterprise DNA, less hype needed.
dennisstevenson said
Bob,
Thanks. This made me think. I’ve been in a large Enterprise and a small private ISV, and I agree. Our current efforts are to “enterprisify” our application. And that’s a lot of work!
A question though. How do you define Platform vs. Application? From a practical standpoint, I’m wondering if what we’re doing in my world here would qualify as a platform or (just) an application. I hear what you said about Platforms being better… but I’m not sure if I’m on that track or not. How would you recommend I make that call?
D
smoothspan said
Dennis, there are a lot of responses to the question of how to tell between an application versus a platform. Let me focus on a couple that are relevant to the way I’ve used platform in this post, which is again, not necessarily focused on selling a product as a platform but rather using it as one.
The first response is a function of the talents of those involved. There is a marked difference between working on a true platform of some kind and an application. In this case, I refer to experience working on a platform that was sold. Some form of what was called “systems software” once upon a time. In particular, have any of your engineers developed a language or operating system? Or, have they largely worked on features and libraries of features? Elaborate libraries are called frameworks, but I will draw the line at including libraries.
This brings me to a second definitional thought. Why draw the line between a library versus a language? The answer is run-time malleability and an understanding of what it is that computers really do. For a deeply philosophical look at this, read the excellent Godel Escher and Bach. From a more pragmatic standpoint, is your application built of one or more domain specific languages? Do you expose those languages in some way that allows users (or at least your professional services) to change the behavior of the app? Does your UI work by altering programs in the DSL? If so, you have a strong platform focus. Your app has “meta” capabilities that are likely very powerful. You and your customers can leverage those to greatly accelerate the developmental evolution of the product. And hopefully, the thought required to have built an app in this way has resulted in a more elegant and powerful architecture. It usually does in my experience.
If you don’t have such a platform, every single features means hacking out many more lines of code. You never get features for free. Features often have tons of unforeseen, idiosyncratic, and painful interactions with each other. Your application is a monolith that will eventually be unable to grow under its own weight.
Knowtu » links for 2009-05-14 said
[…] In Search of Enterprise DNA for Social Software « SmoothSpan Blog (tags: social startups enterprise) […]