Go BackHosted Speech Service sector
A Q&A with Bruce Eidsvik, Vice President of Sales, EMEA Voice Portals, Genesys Telecommunications
What is your projection for the growth trajectory of hosted speech services over the next few years?
Hosted speech services will experience a significant growth trajectory over the next few years, replacing existing legacy on-premise solutions and deploying new applications only possible in speech.
Do you offer speech automation both as on-premise and hosted applications?
Genesys Telecommunications offers on-premise solutions and hosted solutions through our partners.
Are there specific features or applications that lend themselves to an off-site approach? How about on-site?
Voice XML is particularly well suited for an off-site approach as it is a highly scaleable, flexible solutions with an open architecture. Organisations that deal with more sensitive applications, for example those involving high security protocols might prefer to have them placed on-site. However, in this instance, businesses can look to host only the application and still outsource the VoiceXML platform offsite.
Much is made of the fact that service providers generally have to keep up with the latest technology, removing that burden from their customers. So, what technology advancements have you made in the last 12-18 months?
The Genesys Voice platform provides full support for Genesys Customer Interaction Management integration. It also incorporates the latest in speech recognition, text-to-speech, along with call-steering and natural language applications. Genesys has released an updated version of CCXML fully conformant with the Draft 1.0 release and supports Video IVR capabilities. It has also made significant improvements and extensions to its reporting and analytics capabilities. In 2008, Genesys will be merging its VoiceGenie and GVP product lines into a single product. Genesys customers will have a seamless upgrade from either GVP or VoiceGenie to the combined product.
And as the market develops, especially at its current increasing rate, it is important for companies such as Genesys Telecommunications hosted partners to support and take advantage of such capabilities.
Can you compare (with actual numbers) the cost and deployment time differences between hosted and premises-based speech and IVR?
Premise-based speech and IVR projects generally have long lead times as it is necessary for such projects to get all the necessary resources in place. Hosted application deployments on the other hand take between 30 to 50 per cent less time to set up. In some cases, when the applications pre-exist, the service can be set up in days.
What are the criteria that you'd use to recommend that a particular application be speech-enabled, rather than touch tone-based?
A crucial application that needs to be speech-enabled is the customer front door call steering application. The customer is greeted with "What are you calling about?" - as it identifies the caller's query and suggests intelligent call-steering to the relevant department/agent or self service application. This stage is impossible to do effectively with touch tone.
Other applications that would benefit from being speech-enabled are 'Name & Address' identification, Authentication & Verification, and auto-attendant. There are some cases where touch-tone can be effective, entering numbers for example. Genesys recommends a proper balance of the two and always permits a fallback to DTMF (touchtone).
What kind of call volume warrants a premise-based system?
There is no definitive call volume that dictates when a premise-based system is warranted - it is really up to the company to decide. The common misconception though is that the bigger the deployment, the more sense it makes to host it. But in effect there is no minimum size here.
One determining factor to adopting a hosted service is the level of specialisation the company has, i.e. if its focus is centred on the specific product or service it is offering, and contact centre processes aren't its forte, then hosting these are definitely worthwhile.
Does the role of a speech service provider extend into any kind of analysis of the content of the interactions? With speech analytics, for example?
Most definitely! A good speech hosting company should undeniably provide analysis of the content of the interactions - especially with regards to skill sets around user experience. Speech service providers have the experience of deploying different applications in much higher volume than that of the enterprise and therefore it would be a shame for the latter not to take advantage of this expertise.
From a customer point of view, is speech recognition an acceptable substitute for IVR? Can customer happiness with it actually be measured?
Customer satisfaction can be measured simply with, for example, a series of three short questions at the end of a call that relate back to the experience the caller has just had. This is also timely as the person's query has just been dealt with and therefore the experience is still 'fresh'.
Can you perhaps tell us when hosting might not be the best option? In other words, are there circumstances in which premise-based tools are the best choice, maybe for reasons of security or privacy?
Deciding whether to adopt a premise-based solution or not revolves around a company's skill-set expertise. Businesses for example aren't advocates of outsourcing their applications when they have strong internal speech application skills.
Are there any speech "best practices" you can share?
Identifying a caller's intent and call-steering subsequently are two 'best practices' speech applications can cater for effectively - enhancing a company's customer satisfaction levels.
There is a lot of demand for speech out there, but we're still at an early stage. However, it is important for companies to realise that there is much they can gain from deploying such an application. Initial instances of this includes a caller ringing a central number - as opposed to having to sift through a long list of numbers to find which would deal with your specific query - and not only being directed to the right department, but also receiving the right service they rang in for.
For effective speech (or touch tone) applications the most important advice I can provide is constant tuning and performance assessment. I've seen the biggest gains in performance where the business has continually used the data collected from customer interactions to tweak the application. Also, being able to modify that application quickly when changes to the customer offered services change, anticipating the customer demand.
For more information: http://www.genesyslab.com/
