An Interview with Andrew Kanter
Autonomy casts a long shadow over enterprise search, content processing, and eDiscovery. The firm is arguably the largest vendor of enterprise search systems and the dominant supplier of original equipment manufacturing search technology to other enterprise software companies.
The Autonomy organization has grown with strategic acqusiitions such as Verity and Zantaz. The acquistions have given Autonomy technology and new markets for its products.
I met with Andrew Kanter, a lawyer by training, in The Eagle, a short drive from Autonomy's modern office complex in Cambridge, England. The full text of my interview with Mr. Kantor, the chief operating officer of Autonomy's $500 million enterprise. The full text of my interview with Mr. Kanter appears below.
Thank you for agreeing to speak with me. This is a great setting.
Cambridge is a wonderful combination of the old and new. Autonomy has deep roots here. The University's mathematics and physics departments are among the best in the world. Autonomy finds the University a useful resource.
Autonomy has grown rapidly. Your firm's acquisitions in the eDiscovery and enterprise content management markets have been particularly fruitful. How does your firm manage to stay one step ahead of the other search-centric firms with which you compete?
As the volumes of unstructured information has continued to increase, Autonomy has seen much of its vision related to unstructured information come to reality. In this regards we’ve seen the enterprise search market separate into three very distinct strands: the low-end with simple keyword search products that are often free; readily available mid-market products that connect to perhaps ten repositories and do around five simple functions; and the more sophisticated end of the market. At the high-end it has become clear that technology is needed which can connect to every single one of up to 400 different repositories, at the same time respecting all security protocols, operating in every language and offering information governance features like FRCP-compliance. Autonomy operates in this high-end market.
Whilst our success is down to all of these capabilities, there’s a fundamental technology at the heart of Autonomy which is built on that basic vision discussed above. It’s the ability to work with information based on meaning that makes all of these things possible. In this high-end market being able to throw a switch and offer eDiscovery in addition to enterprise search becomes a significant advantage. CIOs no longer want to sign off on two separate purchases for enterprise search and say eDiscovery, they are looking for a single platform which can offer a complete solution. What is important to remember is that when people talk about enterprise search they’re often grouping all of these markets together. It’s like using the term “transport” which can include everything from a BMX to a jumbo jet. Only one of them will ever clear the trees at the end of the runway.
You purchased Verity in late 2006 and immediately became the largest vendor of search and the largest holder of OEM deals in the world. What's the status of Verity K2 Version 7 and IDOL? Are these now one product, or is K2 still accessed as a component of IDOL?
At the time we promised to continue to support K2, and we remain true to that promise. We also made IDOL7 pin-for-pin compatible with K2. That meant that existing Verity customers could do a like-for-like upgrade to IDOL7, delivering significant performance and scalability improvements, whilst allowing them to choose from over 500 advanced functions at the same time. We’ve seen a huge uptake of additional components from the former Verity customer base as they migrated onto IDOL7.
Google has been successful in the enterprise market. The company does not print a master list of enterprise initiatives so most people don't know about the education play, the geospatial play, or the developer play? What's your take on Google and the enterprise both as a search competitor?
We’re really operating in very different markets. Occasionally we encounter one of the mid-market players in the field but it’s usually because procurement hasn’t fully spec’ed the issue to be solved. For our customers there are major issues with some of the mid-market offerings, and what it comes down to is that enterprise requirements are very different from those of consumer search. For example, on the internet documents are already linked to one another so page-ranking algorithms are reasonably good at approximating relevance. In the enterprise none of these native links exist, so page-ranking is a non-starter. In reality, unless you have the ability to understand the meaning of information, you’re always going to have to resort to basic tricks to give the impression of being able to do what the user is asking, and those products are hardly differentiable.
Not only that, many of the repositories still in use by our customers are now obsolete but the records they contain still have to be accessible for legal reasons. Autonomy is the only provider to offer native compatibility with those repositories. There are also crucial issues around security and scalability. So the upper and the mid-markets really are very separate.
With the buzz about cloud computing increasing each day, what's Autonomy's take on cloud computing? How will it impact your enterprise business?
This is an exciting area, and our end-to-end archiving and eDiscovery solutions have been available in on-site or hosted form, with cloud computing components, for several months now. In fact we recently announced that Autonomy hosts the world’s largest hosted archive, at over 5 Petabytes and growing. There is a definite buzz in this area. Having said that, not everyone is ready to go down this route. Some CIOs are reluctant to have critical enterprise data off-site, or dependent on a server hundreds of miles away. They’d much rather keep it all under lock and key. So whilst we’re ready and available, for the moment it’s a relatively small part of our business.
I think of Autonomy as a big deal, industrial strength vendor. Do you see the financial crisis having an impact on that up market strategy?
Whilst we remain cautious, we have seen a counter-cyclical effect from recent events. The volume of litigation has exploded, with the number of lawsuits expected to rise by 300% in the U.S. alone. Companies are struggling to handle this massive increase in litigation and to find solutions to ensure that they stay compliant into the future. Autonomy is already working with 9 out of the 10 top global banks, as well as all of the top 10 law firms. Further there is a huge volume of work created by mergers and insolvencies. As a result Autonomy has closed some of the biggest deals in our history in recent months, reaching to $20 million, $70 million and more. These companies themselves often don't know what's happening within their own walls until it’s too late. In some ways it’s a sad truth but there is always going to be a demand for this type of solution, whatever happens to the macroeconomic climate.
The technology shift in search appears to be moving toward search enabled applications. Will Autonomy face push back from its OEM partners if Autonomy decides to roll out more solution centric applications, not frameworks like IDOL?
The fact that search enabled applications have become popular is evidence of how ubiquitous the problem of unstructured information is. It does not matter if you are implementing a system to administer a hospital, or a new kind of spam filter, the solution will always need to handle unstructured information. It’s rather like the database market: most applications are really an interface layer with some sort of database underneath. It’s exactly the same for unstructured information. Autonomy has over 350 OEM partners and as a result many people are already using Autonomy technology - as part of Adobe, Symantec or HP products, for example - perhaps without even realising it. Thus as unstructured information continues to take centre stage, the general software markets are a very large growth area for us and represent very low channel conflict likelihood.
There's been quite a bit of chatter about Autonomy's next big acquisition. I know you can't reveal specifics, but can you indicate whether the accusation will be primarily technical in nature or will it deliver a solution along the lines of the Zantaz acquisition?
We’re actually very averse to acquisitions. We tend to look for a very compelling rationale, such as a major strategic direction as there was with ZANTAZ, or a very specific piece of technology, as you suggest. So in general it’s never a given that there will be a “next” acquisition.
Are you seeing growth in demand for rich media solutions in addition to standard text-centric IDOL installations?
The beauty of unstructured information is that it is holistic. These days it’s very rare that we have a purely text-centric installation because we naturally communicate as human beings through rich forms of information such as voice and video. Autonomy’s IDOL is at the heart of our Autonomy Virage technology, applying the same conceptual analysis to audio and video content as to text. Further, even if demand today is only for text, most enterprises want to have the capability of including audio and video at a later date. They are bringing into their systems conference calls, webinars and other rich media content. In fact, we are seeing an increase in customers using our Autonomy Virage rich media solutions across all verticals. The strength of the platform approach is enable the architecture to evolve to accommodate the enterprise’s changing needs, without having to rip everything out.
Mobile search is touted as the next big revenue opportunity. What does Autonomy offer in this area? Is mobile search little more than location based directory look ups, not really search at all?
Autonomy launched a technology called IDOL Mobile Edition (IDOLme) in August 2005. There are two major areas in which this is used. Focussing first on telecoms providers, the technology is often used to establish new ways to monetize data, allowing them to offer a unique, personalised experience based on user interests. Whilst manual search is available to users, more important are product and information recommendations based on user interests. Customers in this area include Orange, T-Mobile and 3.
The second main market is large corporates who have knowledge workers out in the field. There are some very interesting developments we’ve recently brought to market here, such as allowing executives to approve documents via secure forms on the Blackberry or even capturing and translating foreign language notices to help users orientate themselves when abroad. In either case it’s about getting the right information to the right person. It doesn’t have to be restricted to location based look-ups.
The number of competitors in search and content processing is staggering. There's shake out taking place now. Delphes may be out of business. Mercado sold out for its projected 2008 revenues of $7.0 million and change. ZipLocal.com in Canada may be close to going out of business. How will the economic downturn and credit crunch affect Autonomy? What are you doing to make new sales such as the big pharma deal you announced and the new BBC deal which still may be secret?
With the current market turmoil we’re finding that people know Autonomy is the market leader and are less willing to take risks on unknown players. They also know that Autonomy has been a stable ship through the previous downturn, never having done any economic layoffs, and consistently profitable every quarter for nearly eight consecutive years. In terms of results, we’ve seen the size of deals accelerating to meet many of the new regulatory requirements thrown up by the current environment.
As you look toward 2009, what are the types of features your technical team will be adding to Autonomy IDOL? Will you incorporate additional speech to text components in the content acquisition component of Autonomy?
It’s interesting that you mention speech because that is set to become an even more important area for Autonomy and our customers. For example there are some changes to the UK stock market regulations which will soon govern audio information. Companies will be required to make voice recordings accessible to courts and fully auditable. In similar cases, failure to comply has lead to sanctions of millions of dollars. In actual fact because people are well aware that email archiving and policies are widely applied, these days the “smoking gun” is most likely to reside within rich media forms of information such as phone calls. So we have quite a bit of work in helping to prepare some large organisations to comply with the new requirements, a trend we expect to accelerate in 2009.
Marketing is something Autonomy is good at. Do you think search is now a marketing game, not a technology game? Will Google's growing sophistication in marketing make it more of a competitor going forward for Autonomy?
Autonomy was born and remains a technology company, led by engineers. The company was founded by world leaders in the space from the top universities, and remains at the cutting edge of research and development, with R&D remaining by far the largest part of the company. The company currently has over 130 patents.
Whilst we were, and still are, technologists first, to succeed we had to learn to market as well. In technology there is always a risk that the best mouse trap may not win. Enterprise Search is an extremely complex area and we’ve had to invest a great deal of time and effort to make sure people are really aware of the important issues, which has seen good success at the high end of the market and not much overlap with the middle portion of the market. Technology, however, will remain at the heart of what we do, and we’ll remember to keep telling the story along the way.
What are the three major changes you've seen in enterprise search and content processing the last year?
One of the biggest changes has to be with the advent of the amended Federal Rules of Civil Procedure (FRCP). The rules state that a company has to provide 100% of the information relevant to legal proceedings within a short period of time, or the CEO can go to jail and heavy fines imposed. This means that our customers have had to be proactive and put in place architecture to deal with litigation before it arises. So over the last year we’ve seen a convergence of operational information systems, with risk management systems. These new rules have converged with the overwhelming growth in unstructured information inside enterprises. Finally, the variety of unstructured information has exploded, with voice and video growth outpacing text.
What do you think about the alleged illegal behavior of Microsoft Fast in Norway? Will this issue become the "Enron of the North"?
This is a matter for the Norwegian regulators to address with FAST. It’s just unfortunate that whenever there are allegations of illegal behaviour, whatever the particulars, it’s damaging to the reputation of the European software industry as a whole.
As you look toward 2009, what are the major trends in search, content processing, and text analytics that you think will become more significant for Autonomy and the search industry?
The explosion of electronic information is leading to a convergence of operational information systems with information governance and eDiscovery, as all three disciplines attempt to control the growing volume of content in the face of increasing regulatory pressure. It’s resulted in a broadening of Enterprise Search and created the need for a platform solution that addresses all of these individual requirements, or what we call Pan-Enterprise Search. This is one area we continue to see increasing in popularity through 2009. Whether the issue is determining how long to keep a document, how to preserve, collect, review and produce that document for litigation, or simply how to make it findable amongst millions of other documents, there is an acute need for a single infrastructure solution to address all of these issues in a consistent and effective way.
In addition, we’re actually seeing a lot of demand from people who have larger deployments of Microsoft Office SharePoint Server (MOSS) and other collaborative technologies, and who are looking to extend these systems with Meaning Based capabilities. In your blog you recently pointed out that while SharePoint is an excellent department-level tool for generating and sharing content, it has some notable shortcomings as an enterprise-wide solution. IDOL is able to enhance SharePoint’s connectivity, search, performance and scalability, for example: adding comprehensive information governance capabilities based on a conceptual awareness of the documents that people are contributing and editing in real-time. Or the ability to implement policies that will operate in real-time to prevent inappropriate content being shared between employees, or even simply to add expertise recommendation based on the content people consume or create and put colleagues in different parts of the world in touch with each other. Autonomy’s technology is already embedded within some of the biggest SharePoint deployments, including the world’s largest SharePoint retrieval project, which consists of 400,000 users, 25,000 sites, and over one billion documents, and again this is an area we expect to grow considerably throughout 2009.
Autonomy has been a leader in technology, prescient acquisitions, and marketing for almost a decade. Unlike its early competitors Convera and Fast Search & Transfer, Autonomy has managed sustained growth. Going forward, Autonomy may have to compete with companies many times its size; namely, Google and Microsoft in the enterprise systems market. Autonomy has demonstrated that it can react quickly to new opportunities exemplified by its purchase of Zantaz and its entrance into the hosting sector. The company has a knack for identifying trends before some of the firm's competitors realize an opportunity exists. The firm's "portal in a box" advertisements in the late 1990s positioned Autonomy in the emerging Intranet search market months if not years before other search and content processing vendors were able to react. My view is that a competitor will have to work hard to keep Autonomy from expanding its business and finding new markets. You can learn more about Autonomy here.
November 18, 2008