HP Autonomy GM: 15 Reasons HP Autonomy-IDOL OnDemand Beats IBM Watson

The Autonomy IDOL OnDemand Advantage

Robert Youngjohns, a 30 year software industry veteran with long stints at Microsoft, IBM and Sun Microsystems, is at the forefront of a Hewlett-Packard offensive to build a massive big data developer/solution provider industry around the company's Autonomy/IDOL OnDemand business. The Autonomy/IDOL offensive comes with IBM investing $1 billion to create a new division aimed at commercializing its Watson supercomputer technology. Here are 15 reasons HP's senior vice president and general manager sees HP Autonomy/IDOL OnDemand beating IBM's Watson in the big data battle.

HP Is More Focused On Creating A Big Data Web Services Industry

It is a mindset approach. I have worked at IBM, as you pointed out kindly. I also worked at Microsoft. The mindset differences between these companies and between what HP is doing [in big data] are fundamental. [In our approach,] we will succeed if we can create an industry around big data, where hundreds and thousands of developers think every day about how they can exploit the functionality we provide through IDOL OnDemand and also through Vertica and other tools that they can use to go and create applications we would never dream of creating.

HP Is Focused On Big Data Application Innovation With IDOL OnDemand Web Services

What we are engaged on now -- which is one of the most interesting things Autonomy has done -- is taking the IDOL platform and creating something we call IDOL OnDemand [at IDOLOnDemand.com] from that. IDOL OnDemand is essentially going to expose each one of these [IDOL] functions as a web service, a documented web service with a very clear interface, with very clear APIs. The idea here is to expose these APIs in a very open way so that anybody can use them to create really nice applications.

IBM Watson Is A Data Scientist-Heavy Platform; Autonomy IDOL OnDemand Is Not

Watson has a got very heavy data scientist component. In other words, when you buy Watson, whatever it is you are going to buy, you are not just going to buy some technology platform, you are going to buy PhDs and the sky is going to darken and people are going to help you define what your big data problem is.
We are building on 12 years of Autonomy history around this [IDOL] functionality. We have learned some stuff on that journey about how you do video analysis, how you do image mapping, how you do sentiment analysis, that you just can't pick up just like that. So [IBM has] got to work from scratch on that stuff. I think they are viewing it much more as a professional services, data scientist-led problem. That is okay. That may be a legitimate approach to the market. It is not the one we are going to take.

Quick App Creation And The Ultimate Partner Story

We created a Jeopardy App. Now this wasn't terribly sophisticated, because our data source was just Wikipedia. But by a combination of taking the Jeopardy Twitter feed and applying that with our tools with Wikipedia as a bigger source, we were able to get a pretty good hit rate on solving Jeopardy problems, which is one of the big things that Watson talks about that it has been able to do. Now [ours] is not very sophisticated. It is very rough around the edges. But it is meant to illustrate the power of exposing all of these APIs to developers and then letting developers create applications in spaces that we would never dream of entering or don't have the expertise to enter. That is the ultimate partner story.

The point was really not so much that we wanted to produce a Jeopardy App. It was more because we wanted to prove the concept of having these APIs and the power of these APIs.

Autonomy Is A More Multi-Cloud Developer-Friendly Platform Than Watson

[The IBM Watson approach] is not the one that is going to win the developer mindset because the developers want the platform that is clearly available, widely deployed, open in the sense that it is clearly documented, easy to use, across multiple platforms. Right now, we are doing this on the HP Cloud. But I don't see any reason why we shouldn't deploy it on any other OpenStack or even on Amazon Web Services or anybody else who wants to be in this space -- [providing it] as a set of services that can then lie alongside these other services that a particular cloud vendor may provide.

A Proven Power Platform For Analytics, Security And Search

We are working with the medical/healthcare industry on how to do better correlation between healthcare outcomes and treatment regimes.

Security is a big one in that [Autonomy] Power [business] space helping people. At one level people have got thousands of incoming video feeds from security cameras and just don't know how to process it because most of the time nothing happens. So can you use software and giga-analytics to say, 'you really need to look at this particular camera because something unusual is happening here.'

Search, which is a really complex problem when you get into very detailed and very, very large data sets, lives within that Power business. [Autonomy] Power you should think of as our information analytics business sitting within Autonomy.

HP Is Making It Easy To Build Web Services With IDOLOnDemand.Com

The philosophy is 'let's energize developers.' With a web service, adoption is everything. And you want to have as little friction to adoption as possible. You'll find out when you log onto IDOLOnDemand.com you can use your Facebook ID or your Google ID. We want to take the friction out.

The roadmap we have for IDOLOnDemand is very aggressive. I am actually stunned by what the Cambridge (U.K.) team have been able to produce in actually a relatively short space of time. And they keep on trucking. I am hoping by the March/April time frame we are going to have a lot more to say, a lot more to show. We are targeting a developer conference sometime this year. We haven't tied down the dates yet. But we want to make a big deal out of it.

A Total Focus On Mass Adoption Of IDOL As A Platform For Partners

I have a passion for this. Maybe it's because of my history. If you want to create an industry around something, you can't do it all yourself. You have got to do it by energizing and enabling the partners. Because at the end of the day we know a very small slice of the industries that we target.

The way you create the rich applications and the way you ultimately generate interest in the platform is by making the platform as widely available and getting as wide adoption as you possibly can. You have seen that model work with other software companies. It is not new. But it is one that we are determined to play here.

We will find a way [to monetize IDOL]. Right now my total focus is taking IDOL functionality, exposing it as APIs and making them widely available.

A Potential App Store For Big Data

We are even thinking about whether we should we set up a web commerce store for this, should we have a store environment for big data. Our Vertica colleagues are beginning to play with this. And it is an interesting idea. I am not sure we have the critical mass yet to make it work. But why not? Everything needs to be done to reduce the amount of friction for any given developer in getting from where they are today with a great idea to actually getting code out there with customers. That friction is not just about access to tools and APIs. It is also access to the market. And that is something that I think we can do well.

A CEO That Gets Partner Success Is Critical To HP's Success

[HP President and CEO] Meg Whitman (pictured) has made it real clear that one of the core competencies that she thinks HP hasn't put enough attention into over the last few years is its partner ecosystem. HP, as a company, one of its great successes is that it has hundreds of thousands of partners in every niche you can imagine in the marketplace, whether its printers or servers or PCs or whatever. She says every day that their success is HP's success. She is putting a lot of investment in that [partner] area with some very high caliber people.

HP Software frankly had a pretty rudimentary partner network [18 months ago] and hasn't really thought partnering very strongly. When George Kadifa came on about 18 months ago, one of the early things he did was to put a partner program in place and recruit an executive, Harry Gould, to run that partner program.

A More Aggressive Push To Embrace Partners At Autonomy

Our strategy going forward has to be extremely partner-enabled. We have done a lot of work on that and we have actually just brought a new executive in, Susan Ferguson, who was previously with Oracle and prior to that with Sun Microsystems. That's part of us putting partners front and center of everything we do.

We haven't done as much work as we should have done and will do in the big data and the analytics space which is where our focus is going forward. The numbers are good (50 percent of revenue is from channel partners). I had a partner advisory council when I was in Barcelona at Discover. Everybody is very enthusiastic about the renewed emphasis they are seeing here. Like every partner network in the world, as long as you can show them that there is a good way they can bring better solutions to their own customers, they will want to work with you.

Autonomy Is Focused On Ramping Up The Innovation Lifecycle

We have really focused on straightforward things: we focused on making customers successful. That matters to us more than anything else. We have got to make partners part of the way we operate and just not an add-on. They need to be intuitively part of the way we operate. We have got to get the innovation lifecycle in this business really running hard. I am a great believer that software businesses live or die on their ability to actually innovate. Sometimes that is a rejuvenation of an existing product. Sometimes it is creating new products. There is no room in this business for people who say, 'I have a great software portfolio, it is throwing off massive maintenance revenues, I don't need to do anything.' It is just not true.

HP Is Leveraging Partners To Build A Big Data Business

I started by thinking: there is an amazing set of software assets here. How do we really drive this from a business that has been growing partly organically, partly by acquisition and how do we really accelerate the organic growth component. The way you do that is you have got to use leverage. You can't do everything yourself. The way you do leverage is you bring partners in, you get developers in, and that is how you get the business to grow.

We have got a new adaptive backup and recovery product as part of our portfolio: great product, virtualization built-in, private hybrid cloud [functionality], all that sort of stuff.

But the truth is, if we are going to be successful with that we have got to get it sold through partners who specialize in that space. To me the [partner] leverage is a key part of making this happen. It is not about going in and finding humongous projects that require thousands of engineers and thousands of data scientists.

Autonomy Engineers Are Now Fired Up By The HP Autonomy Vision

Eighteen months ago we had 30 percent-plus attrition. We have got that down now to where most of the organization is in high single digits. In sales, it is still in low double digits. But that is way, way down from what it was and probably isn't unrealistic given the industry we are in. So I think we are stabilizing that. And we are keeping the engineers.

Engineers ultimately want to work for companies that they are excited by and where they believe they can add value. That to them is as much a motivator as all of the other things that you can do for them. Obviously you have got to pay them enough and you have got to have nice office conditions and so on. But ultimately they want to work in a business that they think has got a vision, excitement and is doing new stuff. I think we have been able to prove that to them pretty strongly over the last 15 to 18 months.

Autonomy Has Been Reborn To Drive Sales Growth

Every part of Autonomy operates in a growing business in a growing market. We have gotten to a point now where our products -- when I look at Gartner or Forrester -- are up there in the sort of top right hand corner, which is where you need to be to be one of the leaders in this marketplace. That is an incredible starting point to drive a high-growth business.

We have got a set of signs around the office [that read] "Autonomy Reborn." There is a great software company here. We can get great growth out of this business and be a huge value to HP. And that is my mission. I am not going to call success because it is far too early. But I am optimistic about what we can make happen.