The 2009 Market Leaders
Across the speech technology industry in 2008 and 2009, mergers and acquisitions, financial chaos, corporate belt-tightening, and customer apprehension have been the dominant trends. But despite the many negative pressures, common sentiment is that the industry is, first and foremost, not being hit as hard as most others, and second, that it’s going to emerge stronger, more nimble, and better prepared to take advantage of the pending economic upswing. That’s thanks in large part to the efforts of the vendors identified as this year’s Market Leaders. After dozens of analyst and consultant interviews, the editors of Speech Technology have put together these rankings—and while some of the winners, leaders, and vendor contenders might come as a surprise, no one should be shocked at the industry’s commitment to better itself, more evident this year than in many years past.
---- Categories and criteria: Speech Technology magazine’s Market Leader Awards names one winner, two leaders, and a vendor contender in each of six categories using a proprietary selection formula that involves input from leading industry analysts and consultants. The selection of these leaders was based on a composite score of the analysts’ and consultants’ ratings for each one in several key areas, including customer satisfaction, depth of functionality, company direction, accuracy, cost, and ease of use, using a weighted scale for each of these criterion based on their importance to current or potential customers. ----
In a year of copious mergers and acquisitions and severe economic downturn, the speech engine market saw the small pool of vendors offering proprietary speech engines competing to increase their mindshare and competitive footprints in an arena where the importance of a strong value proposition skyrocketed for many—if not all—companies and executives.
Among the most notable shifts in the marketplace were the acquisition of Siemens’ Voice Services Unit by SVOX and IBM’s licensing the bulk of its speech patent portfolio to Nuance Communications. Additionally, Loquendo made waves in 2009 by rounding out its services with automatic speech recognition, in addition to its text-to-speech. Nuance continued to show its strength, this time in the embedded market, through a deal that places its speech technology natively on Apple’s iPhone. Despite the recession, analysts see investments in speech solutions wavering but continuing to be strong due to speech’s inherently strong value proposition and return on investment.
After two years in the top spot, Loquendo slipped slightly this year in a close race with IBM. Once again putting up impressive numbers for accuracy, cost, customization and integration, and innovation, Loquendo—with its three decades of experience and an ever-increasing stable of languages (currently at 26, including U.S. and U.K. English; Mexican, Argentinean, Chilean, and Castillian Spanish; Brazilian and standard Portuguese; Mandarin Chinese; Catalan; Italian; Greek; Swedish; German; French; and Dutch) and voices (currently at 63)—boasts not only great technology, but a strong research angle that pushed the technology forward.
Juan Gilbert, a computer science and software engineering professor and director of the Human Centered Computing Lab at Clemson University, described the company, which has offices in Italy, France, Spain, Germany, and the United States, as doing speech research that is “on the cutting edge.”
Nuance Communications retained its lofty position in 2009, scoring an industry-leadership 4.6 for accuracy and 4.7 for customization and integration. But despite these impressive statistics, some analysts view the company as a “business-as-usual” provider offering extremely expensive solutions—a statement supported by Nuance’s low score in the cost category.
Nevertheless, Nuance, which was described by one analyst as the “800-pound gorilla in this space,” is certainly the leader in terms of market share and number of deployments, and will be for some time to come.
Gilbert agrees, calling Nuance “the leader by far in the field with respect to brand and just being out there. Nuance is everywhere.”
In 2008, IBM—a company that excels in product innovation, research, and development—fell off the leaderboard, winning the Vendor Contender Award behind the likes of industry heavyweights Loquendo and Nuance Communications. This year sees IBM returning to the forefront of the speech engine pack, leading a competitive field in several categories and scoring high marks for both accuracy and what one analyst described as “superb customer service.”
In 2009, IBM licensed its speech portfolio to Nuance—a move that increased both the company’s mindshare and financial balance sheet, but left analysts wondering about its next move. Since the dawn of speech technology, IBM has proved itself a strong player in the industry and will likely continue to do so in the future.
Microsoft scored highly this year, outscoring the likes of LumenVox and barely losing out to Nuance for the No. 3 spot. Among the company’s impressive numbers are a 4.75 rating in the cost category and a solid score of 4.0 for innovation. While some analysts think Microsoft has limited reach and market share, the company does offer its product on its ubiquitous operating system, and, as one analyst put it, “they integrate it very well,” proving that Microsoft is a company that will be worth watching in the speech engine space in 2010 and beyond.
Speech Self-Service Suite
In this year’s economy, proactive outbound contact center capabilities started to emerge as the next big growth driver in the interactive voice response and voice portal systems market, which prompted us to add that as a point of consideration in collecting scores for depth of functionality. But despite this emerging market, managing costs related to inbound customer contacts is—and will continue to be—the key impetus for the creation and deployment of self-service applications and the IVR and voice portal platforms on which they run.
Growing demand from customers for hosted and multimodal solutions also shaped the market this year. These developments prompted us to add the vendor's choice of delivery models as a component in our scoring criteria.
But despite the economic upheaval and market reaction that included some continued merger and acquisition activity, as well as changes in customer behavior, the market for traditional IVR/voice portal systems remains strong. Companies still want to simplify customer interactions and create a positive experience that allows users to receive important information or complete transactions, all while streamlining operations, controlling costs, and leveraging existing technologies.
After its acquisition of Intervoice last September, Convergys has advanced itself to the leaderboard as well this year. Less than 100 days after the companies joined forces, Convergys announced the availability of the Intervoice Voice Portal 6.0, an entirely new offering that leverages mobile technologies with outbound notifications and a host of other industry firsts.
Convergys/Intervoice scored slightly lower than Genesys in depth of functionality (achieving a score of 4.52), and even surpassed Genesys in the variety of its delivery models and ability to integrate. The company also scored high in customer satisfaction, evidenced by the fact that in the past 12 months, approximately 90 percent of its clients provided repeat business.
Genesys Telecommunications Laboratories, a newcomer to the Speech Technology leaderboard this year, is certainly not a newcomer in the speech self-service space. Not only are the company’s products gaining prominence in their own right, but they serve as the backbone of systems by powerhouses like West Interactive and AT&T. Analysts and consultants applaud Genesys, an Alcatel-Lucent company, for its strong partner network.
The company scored an industry-leading 4.74 on a five-point scale for its depth of functionality, and also rang up high marks in its ability to customize its solutions, system integration, and customer satisfaction. Its one downfall, though, was price. “Your buy-in is high,” commented Sheila McGee-Smith, founder of McGee-Smith Analytics.
Contact Solutions, a Reston, Va.-based firm, is a strong believer in continuous improvement of its products, coming out of nowhere this year to take the top spot following consultant and analyst voting. It also took the top spot in four out of the six key voting criteria: its ability to customize, integration, customer satisfaction, and cost. If the company has one weakness, it’s in the variety of its delivery models. When so many other vendors are offering hosted, on-premises, and hybrid solutions, Contact Solutions focuses entirely on the hosted voice portal space, though given the current economic situation, that might not necessarily be bad.
In another example of a company whose position in the market has been strengthened by recent mergers and acquisitions, it looks like Microsoft is finally starting to reap the rewards of its 2007 Tellme Networks acquisition. The company, also a newcomer in the Speech Technology awards pool, saw an expanded partner program, new and improved platform capabilities, greater platform performance, revamped research and development efforts, a greater push for on-demand, pay-as-you-go pricing that makes its offerings very cost-efficient, and one of the simplest licensing models in the industry.
In the voice biometrics market, the big drivers have always been customer satisfaction and enhanced security—two factors that are generally viewed as trade-offs. Increase security, and you risk angering customers with false negatives. Increase customer satisfaction by loosening security, and you risk false positives. Every year these drivers tend to fluctuate, and 2009 has been no exception. Because of the severe recession, the pendulum has swung toward customer satisfaction, with enterprises looking to retain customers, free up agent time, increase cost savings, and reduce call handling time.
Also because of the economic downturn this year, larger companies—those that don’t specialize only in voice biometrics—are doing better than smaller companies, many of which are struggling to compete. For years, analysts have been predicting that the voice biometrics market would take off, and once again it hasn’t—this time due to the recession. However, analysts remain optimistic, especially given the recent rise of hosted services within the speech security space and their ability to significantly reduce costs. They’re safely looking to the future and hedging their bets as to the fate of voice biometrics in 2010.
IBM, which offers its WebSphere Voice Server and uses two key technologies—speaker verification and conversation systems—to provide enhanced security for voice-based transactions, makes its first appearance on the voice security leaderboard this year, finishing just behind Nuance and scoring high marks for customer satisfaction and company expertise. Despite this impressive showing, some analysts questioned the future of IBM—one of the first companies to offer text-independent biometrics—in the security space. They pointed to the company’s recent sale of many of its technology licenses to Nuance as the main reason.
After a year-long absence from the leaderboard, 2007’s Vendor Contender, Nuance Communications, is back and posting some impressive numbers in the voice biometrics arena. The company—which offers mainly text-dependent solutions—ranked high in accuracy, company expertise, and ease of use, while scoring a 4.5 for customer satisfaction. According to analysts, Nuance—with its Caller Authentication Services, including the very popular Nuance Verifier, and some impressive deployments, including one with TD Waterhouse—has a strong speech background and boasts a very strong professional services team.
Last year’s Vendor Contender, PerSay, rocketed itself to the top of a very competitive leaderboard in 2009, impressing analysts and putting up some extraordinary numbers in all categories. The Israel-based voice biometrics provider scored an industry-leading 4.5 in accuracy, company expertise, and tied for the lead in customer satisfaction. Founded in 2000 as a spin-off of Verint Systems, PerSay provides its services—text-independent, text-dependent, and text-prompted engines—to the government, telecommunications, and financial services verticals, supporting some very high-profile customers, like Bell Canada, British Telecom, and Vodafone Turkey. One analyst praised PerSay for its “large breadth of products”—PerSay offers three different voice biometric engines, which gives it an advantage. Additionally, PerSay recently partnered with West Interactive—enabling the company to join some of its competitors in leveraging a hosted deployment model.
After two years of solid performance in the voice security category, Agnitio dropped off the leaderboard in 2009, despite solid scores for company expertise, ease of use, and customer satisfaction. The Madrid-based company, which offers text-independent biometrics, is very strong in the government sector but, according to one analyst, does “little in the enterprise space.” However, with a stable of impressive partners, including the national police force in Chile, Spain’s Defense Ministry, and Medical Management Technology Group in the United States, Agnitio is certainly not going anywhere, and will be a company to watch in 2010.
Having been available for several years now, the value proposition behind speech analytics solutions is finally gaining understanding in the corporate world. And although the technology is rich as a means of improving customer contact center operations, its adoption has been slow mainly because early solutions might have overreached—providing too many features and capabilities that forced prices out of the range of the typical buyer and heaped even more complexity on top of already hard-to-understand applications. But that is starting to change as applications become more streamlined, easier to use and understand, and easier to deploy and maintain. These next-generation solutions provide a mix of features and functions that demonstrate proven use cases and applicability to real-world business problems, and they can now quickly and easily adapt to the changing needs of businesses to solve real-world issues in near-real time. Add to that new pricing and delivery models that are making them more affordable, and one could say the speech analytics market is finally ready to explode.
Most current speech analytics technologies are the direct descendants of technological breakthroughs developed decades ago for use in the military and government sectors. As the market expanded to the business world, companies like Aurix, a phonetics-based technology provider located in the United Kingdom, had to change tactics and adjust their business models to accommodate a growing range of uses. Depth of functionality was one of Aurix’s strong suits, though the company also fared well among analysts and consultants in customer satisfaction and ease of installation. Though all of the vendors in this category got slammed on price, Aurix finished far better than most.
To say Verint Systems has been on a roll since its acquisition of Witness Actionable Solutions in 2007 would be an understatement. A perennial presence on the Speech Technology leaderboard, the combined company has seen steady growth and a rising market share following its integration of competing systems into the Impact 360 solution. During the past year, Verint improved the speech analytics within that solution, providing it with greater automated trend analysis capabilities, enhanced semantic indexes, new guided search and context visualization capabilities, and more. On top of that, it tailored the solution for small and midsize businesses, bringing speech analytics to the mainstream. A combination of phonetics and large-vocabulary conversational speech recognition (LVCSR) technologies provides industry-leading speed and accuracy, according to analysts and consultants, and its depth of functionality is also unsurpassed.
For the third year in a row, Nexidia captured the top spot in this year’s rankings, fueled again by its industry-leading positions in cost, ease of use, and ease of installation. But not one to rest on its laurels, the company was busy upgrading its phonetics-based Enterprise Speech Intelligence (ESI) application and Workbench development environment, expanding its customer service and professional services organizations, and launching a Center of Excellence to showcase best practices and the economic and business values of speech analytics. ESI enhancements include improved reporting, dashboards, call categorization, and speaker-separated analysis. Workbench upgrades include real-time analysis tools, automatic search enhancements, automatic audio classification, improved transcript synchronization, and beefed-up audio indexing.
Autonomy etalk returned to the Speech Technology ranks this year after a year’s absence, spurred by a strong showing in ease of use and installation and its wide range of solutions. The company captured attention recently by launching Autonomy Interwoven Multichannel Optimization, which combines Web content management, contact center, and advanced analytics capabilities into a single solution that enables organizations to automatically listen to, analyze, optimize, and act on all forms of customer interactions across multiple channels. That multimodal, multichannel vision will take the company far, analysts predict.
In the wake of the current recession, the capital-heavy professional services market has seen its share of grief.
Last year’s winner, Vicorp, which has a 28-year track record and was favored by analysts for its attention to customer service, this year voluntarily liquidated big pieces of its business, laid off a large part of its staff, and now operates with what some describe as “a skeleton team.” Meanwhile some other players, like Nortel Networks, languish in bankruptcy pending the outcome of Avaya's bid, or more modest bids to restructure.
There’s no doubt that the major shift in the economy has been reverberating through professional services for the past year. “Customer retention and cost savings are big drivers this year,” said one analyst. In light of that, our analysts’ and consultants’ scores have coalesced around solid, proven companies—all familiar names from our past two awards issues. Expect no upsets from upstarts entering the space this year.
SpeechCycle, though smaller than the other vendors featured, has returned as a leader in professional services after winning in 2007. This past year, the company released nRich Grammar Factory, a vendor-independent subscription service for building grammars, and two versions of its transformational services suite. The company scored fairly high across all criteria, only stumbling in cost and company direction. “They’re moderately expensive,” says one analyst, “but you get your money’s worth.”
In direction, SpeechCycle received mixed scores due to a recent shift in company focus. Others, however, felt that the company’s innovation more than made up for it, with one consultant remarking, “That’s the single strongest area: how innovative they’ve been, how good Roberto Pieraccini and others have been in coming up with new ideas for them.”
West Interactive, a familiar name from last year’s leaderboard, held steady this year, just barely being inched out for the big win. With its sizable speech services team and research group, it showed strong scores in company direction and the number and breadth of services it offers. Analysts were impressed with its gamut of services and particularly in its ability to execute them.
With remarkable consistency, Convergys/Intervoice received high marks in every category, scoring the highest in its ability to execute and company direction. Analysts and consultants referenced its assets and the experience of its team as the driving force behind their scores, particularly after Convergys’ acquisition of Intervoice last year.
“When you add up all their assets, they’re a pretty powerful company,” says one analyst.
“Doing this with partners is OK when you’re managing the whole process and at the end of the day your customer gets something,” she says, “but having internal capabilities gave the edge to Convergys.”
Last year, just a few months after the acquisition, Intervoice won a Vendor Contender mention in our awards issue. It was seemingly held back only by its customer satisfaction score, which was artificially low due to a lack of response. This year, however, with Convergys behind it, those customers seem to have found their voices.
At the end of July, Avaya put in a bid to buy the bulk of Nortel Networks’ Enterprise Solutions assets for $475 million. The unit is considered one of the most solid pieces in the Nortel portfolio, and by our own analysts’ and consultants’ scores, has an exceptional professional services team, particularly in speech.
While Avaya scored decent numbers in our rankings—just shy of placing on the leaderboard this year—Nortel scored near perfect in the number and breadth of services offered and scored very high in customer satisfaction and cost. In fact, were it not for its abysmal company direction numbers and more middling ratings in its ability to execute (both related to its filing for bankruptcy), it could have easily been the winner.
As things have shaped up though, should Avaya’s bid for Nortel’s assets be approved by the courts, it will find itself positioned to be a real force in the professional services area.
Mobile Voice Search
With more consumers using smartphones, a whole crop of applications have emerged almost overnight, hoping to cash in on the devices’ growing popularity. Those working in speech have been particularly excited by the development because phones, by their very nature, are a voice technology.
While excitement is high, with at least one trade show devoted to it, the market is still “only in its infancy,” according to one analyst, and “nascent,” according to another. Its value proposition is more aspirational than proven. Fundamental questions still linger about how to monetize and distribute not only voice search apps, but smartphone apps altogether. Does one place ads, for instance, or charge for use? And is there a viable way to get carriers and device manufacturers to pay for all of this?
Because many of the bigger challenges in the space remain unmet, scores were very close in this category, within fractions of a point. Our commentators were torn in some places, but all the same, general trends emerged that, at least for the time being, seem to trace the contours of this emerging landscape.
Operating under the auspices of the parent company that acquired it in March 2007, Tellme Networks has spearheaded Microsoft’s bid to make a place for itself in speech. The vendor is also at the center of Microsoft’s effort to become a player in the emerging mobile search market. At the start of 2009, Microsoft commanded only a slim 5 percent of the mobile market share, according to AccuraCast—nearly the same share it held the year before. This is only one-12th of what Google holds and one-third of Yahoo’s slice.
Many search players, in fact, have seen the mobile space as an area in which Google might be overtaken, or at least significantly challenged. When it comes to Tellme’s voice search offering, analysts and consultants ranked it well in multimodal integration and accuracy. “Their interface does a good job displaying the search results,” says one analyst. Looking forward, Tellme could emerge as a major force in voice search. Microsoft seems to be taking its investment in acquiring the company seriously, integrating Tellme’s search functionalities into the Windows Mobile 6.5 platform that is due for release in a few weeks. This could spell greater exposure and a real greenfield opportunity for Tellme and Microsoft’s voice search.
Powered by this year’s winning speech engine provider, IBM, Vlingo was barely inched out from the top ranking this year. The vendor scored consistently high marks for the number of devices it supports, ranked very high in its multimodal integration, and had the highest customer satisfaction ratings of all in the field. However, despite using IBM’s technology (and this might be related to the devices our independent commentators and reviewers have used it on), vlingo did not receive the top numbers in recognition, and the company suffered appreciably lower numbers in speed. Still, some analysts consistently liked Vlingo, citing its multimodal integration as one of its strongest assets, and preferred it despite its slower speed.
Almost everyone knows the search engine so popular it’s a verb, but with the advent of voice search, Google is starting to make a name for itself in speech. Google scored well in the number of devices supported—sizable with its Google Voice application and near limitless with its GOOG-411 service. It also earned high marks in multimodal integration and recognition accuracy (the engine that draws statistical data from Google’s vast storehouse of text searches). “Its accuracy grows over time. It has very nice adaptive model-making capabilities,” says one analyst. Above all, however, our reviewers noted its speed. Almost every analyst, consultant, and end user raved about Google Voice’s speed in returning results. “It’s quite fast,” notes one consultant.
Interestingly, the rankings in the mobile space seem to mirror the standings in the global Internet search arena. Google leads that pack with an 81.22 percent share, followed by Yahoo, with 9.21 percent, and Microsoft, which holds about a 6 percent share. These numbers indicate a bit of an upset from Microsoft, which just launched a new engine, called Bing, in June, and has managed to shave some hairs off Google’s hulking lead. Whether it will be able to do the same in mobile to advance to the top of our 2010 leaderboard will have to wait. Stay tuned.
Companies and Suppliers Mentioned