April 24, 2013
I just read “Google Buys Wavii for North of $30 Million.” According to the write up:
Both Apple and Google were competing for the Seattle-based startup, and Google eventually won. Apple wanted the company, which developed its own aggregation technology and natural summarization algorithms, for its Siri division. The 25-person-strong team, including founder Adrian Aoun, will be moving down from Seattle to join Google’s Knowledge Graph division.
Interesting move. Is it for technology or people or both. My hunch is that it is people first, then technology. If the $2 million in funding is accurate, the money behind some older search and content processing outfits is going to be demanding sales sooner than later. For outfits six or more years old, generating a hefty return is likely to be a tough job.
Stephen E Arnold, April 24, 2013
Sponsored by Augmentext
April 17, 2013
Readers of this information service, which I use to keep track of information I find useful for my columns and speeches, know that I have held Autonomy’s marketing in high regard. There are some azure chip consultants and failed webmasters who pointed out that the phrase “meaning based computing” was not particularly useful. I disagreed. Autonomy—the pre acquisition version of the company—was a darned good marketing and sales organization.
What is easily forgotten in today’s “did I get more traffic on my Facebook page” world is that Autonomy excelled in three areas:
- The company was able to enter new markets such as video indexing and fraud detection when other search vendors were running around pitching, “We can index all an organization’s information in one interface.” Autonomy picked a sector and figured out how to paint a compelling story around the IDOL black box, the notion of autonomous operation to reduce some costs, and “meaning based computing.” Competitors responded with a flood of buzz words, which made sense at an off site strategy meeting, but did not translate to simple propositions like “automatic,” reduce costs, and process content in more than 400 different formats.” As a sales pitch, Autonomy did a good job and managed to stay at the top of the search vendor stack in terms of closing deals.
- The company used a combination of buying firms which would permit upsells of IDOL related products and very capable management methods to help make the deals pay off. Examples range from the Zantaz buy and the subsequent leveraging of that firm’s technology into cloud service. Autonomy bought Interwoven and pulled together its marketing services into a reasonably compelling bundle of analytics with IDOL sauce.
- Autonomy developed what I thought were clever products and services which caught the eye of certain customers and helped the firm enter new markets. Examples range from the now mostly forgotten Kenjin (a smart desktop service) to Aurasma, a virtual reality service for print advertisers.
HP’s management and advisors paid a lot of money to own Autonomy. Like most search and content processing acquisitions, the realties of running a company in this very tough sector became apparent after a few months. I am not interested in the financial and legal battles underway. What’s important is that HP purchased a company, and HP now has to make it work.
A very interesting pair of articles or semi-marketing type articles appeared in eWeek on April 16, 2013. The first is “HP’s Autonomy: 10 Ways It’s Contributing to HP’s Hardware Story.” These slideshows are ways to get page views. Please, flip through the images in the slideshow. Here’s what I noted:
First, HP seems to acknowledge that turnover and management of the HP version of Autonomy has been a problem. The slideshow calls this a “rebirth”. But the big news from a marketing historian’s point of view is that “meaning based computing” is gone and replaced by “the OS for human information.” I find this fascinating. On one hand, competitors can now carp at the scope of the IDOL technology. On the other, in this social buzzword era, “human information” is actually quite a nice turn of phrase. I won’t make a big deal of the fact that when IDOL’s fraud detection algorithms are working on content, the data does not have to be “human” at all. It can be based on a person’s credit transaction, but algorithms for fraud work on machine and human generated information. No big deal because such distinctions are not of interest in today’s here and now environment.
Second, I did not notice much emphasis on search and retrieval. For someone familiar with Autonomy IDOL, I suppose that search is self evident. Autonomy, however, is mostly an information access system. The add ons were, as I noted above, were extensions or wrappers of the IDOL core, based on Bayesian methods and enhanced in many ways since the mid 1990s. Yep, Autonomy’s technology may seem magical to HP management, but it has been around a while and does not perform some of the functions which Google backed Recorded Future performs or which a skilled SAP programmer can crank out. To me, this is a big deal because it underscores the futility of HP’s trying to make big money deals based on plain old search. Companies chasing search deals are not landing huge deals like those HP needs to make its top line grow.
Third, the “10 ways” are focused almost exclusively on Autonomy capabilities which have been available for a long time. I think that the notion of putting Autonomy functions in a printer interesting, but that idea has been floating around for years. I heard presentations from Intel and Xerox which talked about putting content processing in hardware. Interesting stuff, but the “10 ways” are useful because each makes clear to competitors where HP’s marketing and sales will be going. Examples include using Autonomy for customer support, content management.
The second write up is “HP’s Autonomy Focused on Big Data, Cloud, Mobile, Security: GM”.
This write up contains a number of quite useful insights into HP Autonomy. The “voice” of the article is Robert Youngjohns, the HP manager for the Autonomy unit. I found a number of passages which warrant quoting. I want to highlight three snippets from the three page article. You can get the complete picture in the original article which is worth reading carefully.
First, the story contains the reference to “magical”. Autonomy is math, not magic. The use of the word “magical” is fascinating. It suggests that Autonomy goes well beyond what “normal” content processing can deliver.
Second, the interview lays out the markets which Autonomy will focus upon. These are, as I understand the lingo, big data, information governance, and digital marketing. I am not sure what these phrases encompass, but it is clear that “search” is not playing a front and center role.
Third, there is acknowledgment that the content archiving market is important. The pairing of Autonomy and various HP products is significant. Autonomy will be, to some degree, baked into other HP products and services. This is, in my opinion, an extension of the formula which made Autonomy a revenue producer prior to its sale to HP.
Net net: The Autonomy for 2013 will be fascinating to monitor.
Stephen E Arnold, April 17, 2013
April 15, 2013
Color us pleasantly surprised; a traditional publisher is embracing open content. I fact, it has sunk a hefty sum into it, we learn from “Confirmed: Elsevier Has Bought Mendeley for $69M-$100M to Expand Its Open, Social Education Data Efforts” at TechCrunch. Major global science and health publisher Excelvier has long been seen as an opponent of the open license. Mendeley, on the other hand, is founded on sharing and collaboration. Does this acquisition signal a shift in the publishing industry?
Writer Ingrid Lunden delves into the details:
“Mendeley is both a technology/platform acquisition as well as an acqui-hire for Elsevier. CEO Victor Henning, one of the three PhD co-founders of Mendeley, tells us in an interview that all of Mendeley’s 50 staff are coming over to Elsevier, and Henning will become VP of strategy for the company — a sign that Elsevier may be gearing up for more activity and possibly acquisitions in this space.”
Elsevier has actually been working with Mendeley for several years, indicating it may long ago have seen the inevitability of a more open publishing model. For its part, Mendeley will continue to run its main platform and its academic edition while expanding its free services. Lunden emphasizes:
“It will also keep its API free and open to use: that API today is used by some 300 apps, up from 260 in January. Henning says Mendeley will continue to source data from different places — not just focus on what’s published or owned by Elsevier.”
That is a very important pledge. Henning also says the deal will allow Mendeley to take a longer view, as opposed to stressing over short-term revenue, as they would have done had they remained independent. That means the opportunity to proceed with projects that have been on the back burner, like the hoped-for launch of an Android app later in the year. For more information see the article and/or see the Q&A page Mendeley has created about the agreement.
Cynthia Murrell, April 15, 2013
April 8, 2013
The Swedish Axiell Group AB is developing making moves with the acquisition of the Dutch company Adlib Information Systems. Currently Axiell Group develops as well as supplies advanced IT systems and services for clients such as archives, museums, libraries and also schools. The Journalism.co.uk article “The Swedish Axiell Group in global expansion: Acquisition of Adlib Information Systems makes Axiell the largest in Europe” talks about the big merger. This new record-breaking deal will make Axiell one of the five largest players in the global market. Not only will they gain clients in 30 counties, this merger will provide the platform for them to continue to grow internationally which is one of Axiell’s ultimate goals. Joel Sommerfeldt, CEO of Axiell Group AB made the following statement.
“We are convinced that the combined expertise of the two companies will help us to boost our offer towards the museums, archives and specialist libraries sector all over the world, and we regard this as an important part of our strategy.”
Bert Degenhart Drenth, CEO of Adlib and Marijke van der Kwartel, CFO will continue their work with the company and will become a part of the Axiell management team. Bert Degenhart Drenth, CEO Adlib Information Systems had the following to say
“We at Adlib are very proud to have become a part of the Axiell group. I feel that our combined products, markets and geographic spread enables us to take the next step into the future. However, this is not only important to us, but equally important for our customers, who will benefit from a truly European and sustainable supplier for their mission critical systems. Together we can do more: offer fully integrated solutions for Libraries, Museums and Archives on a large scale.”
The Axiell Group is definitely doing big things and from here their future looks bright.
April Holmes, April 08, 2013
March 29, 2013
A recent career move from prominent algorithm designer Geoffrey Hinton has big implications, according to blogger Mohammed AlQuraishi’s post, “What Hinton’s Google Move Says About the Future of Machine Learning.” Hinton, an esteemed computer science professor at the University of Toronto, recently sold his neural-network machine-learning startup DNNresearch to Google. Though not transferring full-time to Google, the scientist reportedly plans to help the company implement his brainchild while remaining on at the university.
This development follows Google’s pursuit of several other gifted academic minds that specialize in large-scale machine learning. AlQuraishi says these moves indicate the future of machine learning and artificial intelligence. He writes:
“Machine learning in general is increasingly becoming dependent on large-scale data sets. In particular, the recent successes in deep learning have all relied on access to massive data sets and massive computing power. I believe it will become increasingly difficult to explore the algorithmic space without such access, and without the concomitant engineering expertise that is required. The behavior of machine learning algorithms, particularly neural networks, is dependent in a nonlinear fashion on the amount of data and computing power used. An algorithm that appears to be performing poorly on small data sets and short training times can begin to perform considerably better when these limitations are removed. This has in fact been in a nutshell the reason for the recent resurgence in neural networks.”
It is also the reason, the article asserts, that research in the area is destined to move from academia to the commercial sector. The piece goes on to compare what is happening now with the last century’s shift, when the development of computers moved from universities, supported by government funding, to the industrial sector. It is a thought-provoking comparison; the one-page article is worth a look.
Is AlQuraishi right, are the machine-learning breakthroughs about to shift to the corporate realm? If so, will that be good or bad for consumers? Stay tuned.
Cynthia Murrell, March 29, 2013
February 26, 2013
The IBM announcement that it will be rolling out new communication products and new upgrades for its existing social networking product, Connections hasn’t really come as a big shock to many. IBM has spent time and money acquiring new technologies and working to integrate those technologies.
CIO’s “IBM To Beef Up Content Management, Analytics In Connections Enterprise Social Product,” takes consumers through some of the basic changes they can expect to see when the products are unveiled on Monday at Connect 2013.
“At a press conference after the session, Mike Rhodin, senior vice president of IBM’s Software Solutions Group, said that the impact of enterprise social technologies in collaboration and front-office business processes like HR and marketing amounts to a “generational shift” that is transforming how companies function, and will do so for the next two decades.”
We aren’t really told which acquisitions are responsible for which upgrades and integrations but if IBM’s dreams come true, the new content management function of Connections will rival that of Microsoft’s SharePoint, a big assertion for sure.
The IBM Employee Experience Suite is one of the few newly designed products that fully explains where the new upgrades came from, in this instance, the human resource management apps are courtesy of the $1.3 billion acquisition of Kenexa.
While still a little cloudy on the content, it will be interesting to keep an eye on IBM over the next year and not just at its product reveal early next week. It’s a sink or swim time in business technology with so many up and coming developers and technologies just waiting in the wings for an opportunity. We’ll see how IBM continues to stack up.
Leslie Radcliff, February 26, 2013
February 21, 2013
We thought search was Google Plus. Now Google is morphing into a tracking outfit? Business Insider reveals that the search giant has snapped up a sales-tracking business in, “Google Spends $125 Million on Channel Intelligence to Improve Google Shopping.”
Writer Jay Yarrow reports that Channel Intelligence (CI) provides solutions that boost online traffic for retailers of products in many categories. The company’s blog speaks of working with Google’s shopping division. Yarrow believes he knows why Google made this move:
“One of the looming threats for Google is the continued strength of Amazon. When people want to buy stuff online, they will skip Google and head straight to Amazon. Inside Amazon they will search, and then buy stuff.
“Google’s business is built around people searching on Google for things to buy. That’s the most valuable search from a commercial perspective. Google is trying to improve its shopping services to combat users tendency to go straight to Amazon.
“We assume CI will be a part of improving shopping so that when people search on Google for products it will list better, more relevant results for users.”
Probably a good guess. But will it help, or is the Amazon habit already too entrenched?
Yarrow includes the press release from the tracking firm’s parent company, ICG, in his article. Channel Intelligence is headquartered in Orlando, Florida. Just a few of its many prominent clients include Target, Microsoft, and General Electric.
Cynthia Murrell, February 21, 2013
February 7, 2013
Connectors are important. If a system cannot acquire content from a system, the fanciest text processing system cannot do its work. Years ago Oracle acquired Stellent which had snapped up Outside In. Now Salesforce has followed in Oracle’s footsteps with its acquisition of Entropy Soft. (I assume the story “Salesforce.com Has Acquired French Startup EntropySoft.”) Entropy Soft is not a start up in my opinion. The company was set up in 2005 or so. According to the write up, Entropy Soft had about $3.5 million in funding. Details are limited. The question which the deal raises is, “What services will Salesforce introduce which acquires software to acquire diverse enterprise content?”
A list of Entropy Soft connectors is no longer available online. According to my files, Entropy Soft has more than 40 connectors. These include:
- Microsoft SharePoint
- IBM Lotus Quickplace
- Hummingbird DM
- EMC Documentum
Update: A list of Oracle connectors is at http://www.oracle.com/technetwork/search/oses/ses-connectors-178226.pdf
Stephen E Arnold, February 7, 2013
January 29, 2013
After the dress rehearsal for my weaponized information webinar, a couple of librarians and I were talking about the commercial database business. I narrowed the focus to the commercial outfits selling primary and secondary information to libraries and other professionals; namely, to the legal and health care sectors.
In a nutshell, the digital future does not look too bright for companies such as:
- Ebsco Electronic Publishing (everything but the kitchen sink coverage)
- Elsevier (scientific and technical with Fast Search in its background)
- ProQuest (everything but the kitchen sink coverage plus Dialog)
- Thomson Reuters (multiple disciplines, including financial real time info)
- Wolters Kluwer (mostly legal and medical and a truckload of individual brands)
I just reread “Why Acquisitions Fail: The Five Main Factors by Pearson Education. This outfit has a long and storied past. The irony of Pearson Education explaining the problems of making an acquisition work is interesting but not germane to the main points in the write up. the fact that this item was available to me without charge via the Internet is amusing to me as well. Here’s what the Pearson analyst suggests about the causes of failure:
Survey after survey has proclaimed that most acquisitions fail. Denzil Rankine’s Executive Briefing on Why Acquisitions Fail (FT Prentice Hall) examines why. There are five key factors, which we will examine below:
- Flawed business logic
- Flawed understanding of the new business
- Flawed deal management
- Flawed integration management
- Flawed corporate development
No argument from me. The business model for these firms has been built on selling “must have” information to markets who need the information to do their job. The reason for the stress on this group of companies is that the traditional customers are strapped for cash or have lower cost alternatives.
If one of these outfits buys a company, the likelihood that the acquisition will be a home run revenue success is low. These five companies are bottom-line oriented, so the acquisitions will have to perform. The idea of massive investment to realize the promise of the purchase is not in the game plan.
So big traditional commercial database companies have to find a way to work around the Pearson Education hurdles. Let me consider some of the options available to the Ebscos, Elseviers, ProQuests, Thomsons, and Wolters Kluwers of the world. (Yes, there are oligopolies in a number of other countries, not just the US and Western Europe.)
The Hail, Mary Deal
This is the option which makes investment bankers’ and deal brokers’ hearts go pitty patter. We know how that approach works.
Buy One Another
The idea is that no other outfit wants to buy commercial database companies. Ergo: These outfits buy one another in some combination. Good for the investment bankers but long term, the customers may not be able to cope with ever increasing prices. Librarians, lawyers, and accountants are not exactly in a GEICO made of money mode.
The Microsoft Dell Variant
The idea is that a third party like Google buys one or more commercial database companies and monetizes the content with ads. (I would lobby for this if I were attached to a giant money machine like the Google.)
I think that Thomson Reuters’ effort to get out of the health fraud business makes clear that the price offered kills the deals. Nevertheless, some of the commercial database publishers may be forced to chop off fingers and toes to keep the core alive. Highly probable path opine I.
Raise Prices and Innovate from Within
This option keeps the Board of Directors engaged. The reality is that such innovation goes nowhere. Ah, I am looking forward to annoyed vice presidents asserting, “I am innovative. We do innovate.” Okay, okay.
Big changes are coming for commercial database producers, access to curated content, and the quality of the commercial information. Lawyers are looking to cut costs. No good for Lexis and West. Librarians are under severe financial pressure. Accountants? Accountants don’t want to spend their own money.
Looks like the future is moving in directions different from what these traditional, commercial database producers are going. I suppose after a couple of decades of evolution, the arrival of the End of Times is tough to accept.
Disagree? Agree? Surprise me. Keep in mind that I don’t have a stake in these companies and find myself baffled by the management challenges each has created for itself.
Stephen E Arnold, January 29, 2013
Sponsored by Dumante.com
January 28, 2013
I read “Report: There Were 94 Tech M&A Deals in 2012 above $100M. Average Deal Value Was $717M.” Interesting stuff. Let’s assume that the data are accurate. Where do to search system buys outs fall.
Autonomy sold for $11 billion but that number may be soft due to the Hewlett Packard write down of $8.8 billion. I don’t have a dog in the fight, but the matter does suggest that some acquisition deals in search were overoptimistic.
Vivisimo sold to IBM for about $20 million. This is a fuzzy number which I heard about from poobahs at a trade show. If the number is accurate, Vivisimo sold to IBM for about the 2013 estimated revenues. However, Vivisimo is not a search company. Vivisimo was, at the time the deal was announced, was a big data company.
Brainware and ISYS Search Software sold for undisclosed amount. The amount Lexmark paid is not known. Lexmark did pay $280 million for Perceptive Software, so let’s assume that these transaction were less than $280 million. Close enough for horseshoes.
Oracle has not disclosed how much it paid for Endeca. Endeca, a privately held company, sold for about $1.1 billion. Again the only source I have is poach outputs.
- Search can yield some nice payouts
- Paying too much for search can trigger some interesting post purchase hand wringing
- Search is not really search. The reason people buy search companies is to get customers and solutions to specific problems like customer support or “meaning based computing.”
With some venture firms pumping money into search and content processing vendors, the likelihood of a big payday exists. On the other hand, search, like other software niches, poses some significant risks in my opinion.
Stephen E Arnold, January 28, 2013