Systems management from the cloud
No one starts out designing a complicated product. But when sales and marketing become involved, enterprise software can often mutate from a simple tool into a massive, feature-packed platform. Unfortunately, over engineering and a lack of focus are synonymous with enterprise IT.
In the enterprise IT management and administration world, most commercial solutions are brought in to replace homegrown scripts and systems. It’s like buying a car to replace a handmade wagon. Sounds like a great improvement, but it comes with a downside. At purchase time, there is already a laundry list of chores any new software package will be expected to handle. Custom-made systems are built specifically for their environment and commercial software can’t anticipate every possible individual usage. You could use a new Mercedes to haul fresh-cut hay from the field, but it’s hardly the right tool for the task.
How can any static, unchanging software be expected to adapt to handle every imaginable environment? The answer so far has been massive consulting and customization projects, coupled with retraining the users to mold themselves to the new system. Think of it as going to a never-ending driver’s ed class.
Then there’s the problem of software sales. In many industries, the engineers come up with a concept, build it and then turn it over to the sales team to put it into buyers’ hands. Yet, for some reason, building software for big corporations often goes in reverse: the sales team finds a customer and then turns it over to the engineers to build.
By trying to appease every customer need, the process of creating great software is often compromised. In fact, it’s one of the biggest reasons enterprise software sucks, as this great post points out. Just imagine if we built cars by taking customer orders and trying to package them into a single vehicle: “The customer wants to go zero to 60 in three seconds, have space for his four kids, go off-roading and get 100 miles per gallon. We can build that, right?”
Or, to paraphrase one blogger: The arms race for features is over. Everybody won. Anything is possible. If you can imagine it, you can have it. Meanwhile, the user has been carpet-bombed with features.
And the bigger the sale, the more likely a new feature will be crammed into the product just in time to seal the deal. That means enterprise software packages are evolving to meet the demands of the biggest customers, and questionably useful “features” accrue quickly. Greater complexity increases the possibility of errors, because no one really understands all the interacting parts of the whole or has the ability to test them. Worse still, the end users aren’t even involved in the purchase process: their managers are trusted to pick the right tools, after they’ve been plied with conference schwag and fancy dinners.
These are big problems in the gigantic enterprise software business. Several thought leaders and companies are trying to beat these issues, fix the broken system and develop a new generation of easy-to-use programs for a range of applications.
Yet no subset of the enterprise software market is as ready for change as datacenter automation and systems management. Software suites have not addressed the need to simplify the process of working with servers. Most sysadmins still rely on homegrown automation (scripts) or just plain “hand-to-hand combat” involving manual changes.
We talk with customers all the time about the automation of their routine basic IT tasks. Increasingly, the worth of a systems administrator is measured by how much of the job he or she has automated. Automation and simplification should be the backbone of any systems management service yet that’s not what we see. This deficit is due to a few serious problems in existing enterprise IT solutions:
We want to create a systems management tool you can boot up, jack into and start using immediately. We’re focusing on the Law of the Vital Few to build it. We know that there are a handful of operations and functions that most people need and can use to drastically improve their productivity. It’s like Gartner says: simplicity focuses on the relevance of features instead of an absence of features.
Customers are beginning to demand that the software they buy be easy to install, require only basic upgrades and need only minimal user training. Our goal is to focus on the most important and essential automation tasks and make those simple, fast, and inexpensive instead of trying to solve every conceivable problem with a single product. We want to make simple tasks easy.
Today, software has to be more than just easy to use. It also has to be easy to try and to buy. Sure, everybody offers the classic 30-day trial, but setting that up typically requires preparation and a fresh environment in which to work. For example, testing out a security-auditing tool could require a would-be buyer to set up a specific Oracle database and go through all those steps just to see if what if it will work.
But you can try out cloud software in seconds and only need an Internet connection to take it for a test drive. No meetings, special sandboxes to set up, or salespeople to listen to. Boot up, jack in and start using.
To compete in today’s market, systems management and datacenter automation products must be both easier to use and easier to try. We’re going to put the sysadmin in the drivers seat. So if you want your software to take you further, start working faster and make easy jobs simple, sign up to take an early test drive of our product.
Thanks to Derrick Harris (GigaOm) for a great article about our mission, and especially like the comparison to SalesForce.com 🙂
In particular Derrick highlighted one of the most crucial parts of our story, which is the focus on the social aspects of managing systems in this new world. I’ve always been amazed that in delivering a brand new product through the internet why more companies don’t rethink the way they build their products to take advantage of the best things that their platform gives them? Hopefully more on that soon …