For Larry Mark, speech analytics is a lot like a Dremel tool. Initially created for simple hands-on projects, the Dremel has spawned thousands of other uses around the house—often for projects totally unrelated to its initial purpose. Mark, who is chief technology officer at SER Solutions, says speech analytics has evolved the same way. Once viewed as a spin-off of speech recognition used for just word spotting, speech analytics today is seen as a means to spot deep-seated business problems, many of which had previously gone unnoticed.
For example, broadband cable systems provider Comcast purchased CallMiner’s Speech Analytics Suite in hopes that the technology could help contact center agents better adhere to standards and policies. By recording and monitoring call content, the company sought to use the data extracted with speech analytics to develop better agent performance. But when the company looked at the actual data, it found that agent performance wasn’t really a problem; the real issue was a mix-up in the company’s billing procedures. Comcast subscribers were calling customer service at higher volumes because they appeared to have been billed twice. Though they were not, the mix-up generated an increase in call traffic. Comcast was able to address the problem and is now focusing on using the technology to pinpoint areas in which agents can promote service upgrades during customer calls.
How does a company use speech analytics to not only address initial concerns, but also to effect change within the enterprise? Therein lies the challenge. For speech analytics to work, a company has to put in a significant amount of work to analyze the data. And for the technology to produce a return on investment, an organization must make that data actionable; changes must be made enterprisewide. Most analysts and vendors agree that for speech analytics to take off, companies must begin to see the technology’s effectiveness outside the contact center.
According to Paul Stockford, president and chief analyst at Saddletree Research, companies must think outside the contact center box.
"For all the talk you hear in the market about how important the contact center is—and it’s true—the majority of companies still look at the call center as those people who answer phones and we pay $8 an hour," Stockford says. "There are a handful of companies that use the call center to their advantage, and they will immediately recognize the benefit of speech analytics."
Missionary Work and Rocket Fuel
In 2007, DMG Consulting published a 270-page report on the state of the speech analytics marketplace. "The 2007 Speech Analytics Market Report" made a strong prediction: that the speech analytics market would grow by 100 percent in 2007 and 2008, making it the fastest-growing contact center technology market segment.
At a time when products like workforce optimization (WFO) and unified communications (UC) attempt to stake their claims in improving contact center operations, speech analytics vendors are making the sell that their products can also change the way a business operates. But companies faced with IT budget cutbacks have perhaps been hesitant to invest a substantial sum of money in a still relatively new technology. It’s a big gamble, and some vendors realize this.
"Speech analytics is a great technology, but right now it has niche applications, and I think that what’s lacking is executive understanding and buy-in," notes John Benchoff, senior director of product marketing at SER. "It has the potential, but I don’t see it having the kind of rocket fuel right now to get there."
When it comes to software for the contact center market, speech analytics vendors and analysts are some of the most adamant and charismatic. And rightly so—while able to whip up impressive stats and technical terms, the market’s players have to be cheerleaders. Until speech analytics gets what Benchoff calls "rocket fuel," the stand-alone application means nil. And that is the biggest letdown.
While many organizations have purchased call recorders, few have utilized data gleaned from their customer interactions effectively. With millions of hours of conversations floating in contact centers worldwide, speech analytics vendors can see their software’s potential to extract actionable information. If only more companies would listen.
"There’s still a lot of missionary work to be done," Saddletree’s Stockford says. "We’re just starting to see some real traction for business analytics, and speech analytics is something that seems way out there to a lot of people still."
The laws of technology adoption cycles play a significant role in speech analytics’ growth. Most ideas operate along years-long curves before being accepted as an essential part of everyday business operations. That’s how Greg Borton, vice president of analytics at Nuance Communications, views speech analytics.
"No one knew what CRM meant in 1990, and look how it’s totally transformed our industry in 10 or 12 years," Borton states. "IP telephony was not heard of until 1996, and look how it’s transforming our worlds. These were real trends independent of the invention. Speech analytics is in year three or five of that pretty typical, 12-year curve of technologies."
What sparked CRM’s and IP telephony’s successes were not only more adoptions and established best practices, but also innovation and integration. Just because speech analytics is emerging in the marketplace, it doesn’t mean the vendors’ work is through. In a multimodal landscape, companies must find new ways to integrate their technology into existing applications, form partnerships with other contact center technology vendors, and work out the kinks between two very different speech analytics models: Large Vocabulary Conversational Speech Recognition (LVCSR) and phonetics (see the sidebar).
The vendor landscape remains full of competition, with companiescontinually announcing new partnerships and updated product releases.According to Daniel Hong, an analyst at Datamonitor, speech analyticsvendors can be broken down into three distinct categories:recording/quality management (NICE Systems, Verint, Autonomy, Utopy,and Voice Print International); stand-alone analytics (CallMiner,Nexidia, Utopy, and Aurix); and contact center infrastructure (SER).
Speech analytics will take several years to hit its stride for otherreasons, Hong maintains. "It will be another two years for the speechanalytics technologies to mature, and for newer and less complexversions to come out," he said in an email. "Standardization will makespeech technologies become more affordable and attractive to small andmid-size businesses and lead to more adoption in this market segment."
Cash Is Still King
While technological capabilities and methodology still act as the mainmotivational factor for companies when deciding which vendor to select,price has not left the picture. However, once organizations see howeffective speech analytics can be, they won’t care about the price tag,many vendors and analysts suggest.
"I don’t think cost has anything to do with [implementation rates],"Nuance’s Borton says. "If people saw the value, and they saw how theycan change the way the business works, I think they would be willing tospend 10 times more money than is being charged by vendors."
According to Stockford, however, the numbers today speak a bitdifferently. He estimates that revenue for speech analytics managedservices in 2006 was only $190,000 split among only three vendors thatoffered the service at the time. Further hampering adoption was areluctance by potential user companies to dramatically alterinfrastructures and buy additional servers needed to store all the dataand process it effectively. In addition, time and money must go towardcompiling a group to analyze the data in house or hiring someone fromthe vendor side who will do the legwork.
Jeff Gallino, chief technology officer at CallMiner, notes, however,that three pricing models exist to help an organization stay withinbudget: per-use, per-seat, or per-organization. Which one is chosenshould depend on a business’ size and what it hopes to gain from theanalytics’ findings. If a company records and mines 100 percent of itsaudio, Gallino says it’s best to buy on a per-seat or per-organizationbasis; if the data is being used for simpler, less time-consumingprocesses not done in real time, a per-use approach is a better fit.
While some companies offer speech analytics as an add-on to otherservices (such as WFO), stand-alone vendors like CallMiner say theirsingularity of purpose lets them bring bigger value to customers. "Weconsider ourselves highly functional, which means our prices are not ascheap as the people who use [speech analytics] as an add-on," Gallinosays. "That’s not our business model. The return on value thatcustomers are seeing—building their business cases in between six weeksto nine months—is what we’re selling."
Like other facets of the contact center market, however, many speechanalytics vendors are moving toward managed solutions as well. Usingexperts from within a vendor organization, companies can rent both thetechnology and brainpower that go into mining data.
Yochai Konig, chief technology officer at Utopy, says his companyoffers managed solutions, but that other issues can complicate thedecision to go with a hosted service provider. "We are offering [hostedsolutions], but the reality is that some verticals are very sensitiveto get the calls out of their network given the type of information inthe calls," he explains. "Some are more open to it, and obviously wehave safeguards. We have this [hosted] option for the customer, but wedon’t force it [on them]."
But while the hosted market is "not catching on like wildfire,"according to Stockford, some companies are cautiously experimentingwith speech analytics by investing smaller amounts of money initiallyand waiting for feedback before jumping in head-on. As more speechanalytics companies, however, show proven scalability, the market couldcatch on.
"What [some companies] are looking to do is try it out by spending$50,000 or $75,000," says Donna Fluss, president of DMG Consulting."What companies don’t want to do is make a quarter-million-dollarinvestment in something that is unknown."
Building Value
Stand-alone vendors like CallMiner firmly believe that the best way tooffer speech analytics is by keeping the focus on their service’s uservalue, but others within the market think a different approach isneeded for buy-in.
"What I’ve been hearing is that speech analytics will get rolled intoWFO," SER’s Benchoff states. "Vendors in speech analytics are lookingto bundle the apps as both differentiators and a value-addedapplication. It’s going to be difficult for stand-alone vendors tooffer a point solution in the marketplace."
Companies may also recognize that, though the contact center is stillthe most popular place for customer interaction, the Web is also anundeniable member of the customer service division. With youngergenerations gravitating toward self-service instead of human-to-humaninteractions, a company’s online customer service traffic has becomeincreasingly crucial to business operations.
"I think it will be necessary for speech and Web to partner," Stockfordsays. "Voice is still the primary means of customer service and willcontinue to be for several years, but in the long run there’s going tohave to be a drive to also start monitoring Web activity, lookingthrough to see where people are clicking to get their answers."
SER Solutions, which started in the telemarketing arena, notes thatspeech analytics should play a role in sales and marketingopportunities as well. The technology needs to be able to identify theparts of calls that agents can use to make cross-sell or upsellpitches. Speech analytics’ ability to produce results in real timecomes in handy when used as a sales tool.
"If agents can be told something to do when they have someone on thecall, that will either increase the volume of what they sell or thetotal sale on the call, or be given advice that can make a no-sale calla sale call; [that] is the type of app that makes real-time [speechanalytics] significant," SER’s Mark explains.
But the biggest goals in speech analytics (increased sales, smartermarketing, and improved business operations) will be hard to come bywithout partnerships. Vendors with partnerships bring to the table notonly their own capabilities, but their partners’ services as well.Stockford notes that from a competitive perspective, "any companythat’s not going to either partner or acquire technologies will notremain competitive."
The Debate Continues
The argument for speech analytics is, in some ways, far from over.Though case studies with large companies like Comcast have shown howextracted data can improve business outside the contact center,executive buy-in and sales to businesses with less than 100 seats arenot hitting the mark. Vendors and analysts agree that speech analyticsworks, but only if a company works speech analytics. That means morethan just pressing a button and walking away; to get the problemsolved, time, money, and manpower must be allocated. And, as morecompanies begin to view the contact center as a place for sales and notjust help, analytics could deliver on its promise of trulyunderstanding a customer.
In coming years, the market could also morph to more closely resembleother facets of the contact center technologies space. "There are lotsof players in the market, but there are different niches now," Nuance’sBorton says. "This market is going to differentiate into probably fiveor six submarkets in five years because that’s the typical pattern oftechnology."
CallMiner’s Gallino says the future of speech analytics lies primarilyin the hands of the customer. "Deep down, people are a bit intimidatedby the concept of speech analytics, the idea of automatinghuman-to-human interaction," he states. "If you look at the hype cyclefor speech analytics, we’re coming up the other side of the trough.Some of the promise is now starting to be delivered."
LVCSR versus PhoneticsThe two methods used for audio mining in speech analytics, Large Vocabulary Conversational Speech Recognition (LVCSR) and phonetic search, rely on different types of technology to extract data from recorded conversations. While both have proved reliable, vendors and analysts agree that sometimes one method can trump the other, depending on a company’s goals. A breakdown of the two methods follows.
LVCSR: Used in extracting specific words preprogrammed by a customer or application developer, LVCSR not only finds words, but also helps users grasp the context in which the word or phrase was uttered. While more CPU-intensive than phonetics, LVCSR is often more accurate.
"LVCSR is better than phonetics at detecting subtle word differences," says Larry Mark, chief technology officer at SER Solutions. "If you’re talking about a disclaimer where it matters—if I say, ‘Past performance is no guarantee of future returns’—that kind of difference in a longer phrase you won’t pick up in phonetics."
Phonetics: Rather than create grammars before implementation, as required by LVCSR applications, phonetic searches translate conversations and allow those analyzing data to pick up key words, phrases, or topics. Often less expensive to implement, phonetic search is typically used as a "first tier" form of data analysis—users mine data with phonetics, and then go back with LVCSR to refine the search. "Phonetics is still a great technology for lots of apps," Mark says. "It’s good for finding competitive mentions and understanding what agents are saying."
The Verdict: Vendors and analysts interviewed agree that speech analytics companies will need to offer a combination of both search methods to stay competitive. Rather than competing over which method is better, Donna Fluss, president of DMG Consulting, says vendors and users should see the two as complementary. "It’s unlikely that an application would run all conversations through both technologies," she states. "It’s more likely that an application would use one of the technologies and then refine the analysis with the other one.
Is Real Time Really Necessary?Vendors like CallMiner offer audio-mined data results on a real-time basis, giving users instant insight into problems that could have arisen within a contact center or on an enterprise-wide level. Of course, quick access means more money. We asked analyst Paul Stockford from Saddletree Research to weigh in.
Speech Technology: Is it compulsory for companies to have data produced in real time? When is it not necessary?
Stockford: We’re talking about a call center, not national security issues. Time is relevant, but it’s not the defining factor. There are advantages [to real time], but you take a service scenario where you take a five-day chunk, it gets processed into a text format, you can scan it, and that’s where you start gleaning things and asking questions you didn’t know you should have asked before you did the scan. You can go back and dig deeper and deeper into that single factor or group of factors as opposed to having to keep processing all the time. But if you need [the information] really fast, then real-time makes sense, but I don’t think there are that many call centers out there that need it immediately. I think it’s way too early to say whether one will be preferred over the other