Personal Self-service
It is essential that the customer likes using the on-line self-service product. Self-service is only successful if your customer finds it a pleasant, easy and reliable way to communicate with your company. Using an on-line self-service product must be at least as good as contacting a member of your staff. Actually, it should be better, because self-service provides your customer with a solution to their problem without having to wait. Generally speaking, the current on-line self-service products are not very user-friendly. The customer has to trawl through a list of frequently asked questions before he finds the answer to his question. Even search functions are seldom satisfactory. Quite often, the customer receives just a summary of potential answers to his question.

Self-service can only be successful if the customer receives an immediate answer to their problem: the customer asks a question and gets an answer. Sometimes it is necessary to ask a question in return, in order to supply a more specific answer. The virtual assistant offers just such a customer-friendly self-service - an on-line personality who invites customers to ask questions.

Existing systems vs. Virtual Assistant
There is a big difference between the existing self-service system and the virtual assistant (see table below).

 

Automated FAQ	Virtual Assistant
Answers questions	Conducts  a dialogue
Question leads to several answers	Question leads to 1 answer
Suitable for answering a question	Suitebale for solving a problem
Impersonal	Human characteristics
Key words	Grammar, text patterns, context probability, historical details, customer profile

Comparison of a virtual assistant to existing systems

The most important difference between the two is that existing systems simply automate frequently asked questions whilst a virtual assistant has a dialogue with the customer and utilises the question’s content. The result is a self-service product that the customer will enjoy and will return to again and again.


Processes and technology
The virtual service assistant offers users of your site a simple way to unlock information. It replaces search machines or the endless FAQ lists on your site.

All this is possible by using a ‘script language’ that has been specially made to interpret natural language. Its basis is a language processor which has been developed to ‘understand’ written language, and a system that recognises the context of the question by finding the correct answer (a context parser which determines the syntactic structure of a sentence).


This script language is also used to record the questions and answers. An Information Pyramid makes sure that the virtual assistant can almost always answer a question. It is a structured method that builds up questions and answers in a database. In addition to general questions, specific questions are added that can answer multiple questions. This does depend though on the size of the project and your own preferences.

The advantage to this method is that the virtual assistant determines in advance which information to use, and then records this information efficiently. Thanks to the script language and the language processor, all potential questions about, say, the cost of a product are recorded just once. No matter how the user phrases the question, the virtual service assistant always gives the right answer.

The technology is simple to integrate with the existing systems of the organisation. This includes integration with the company’s databases such as CRM-systems but also with live employees through e-mail, chat or CallMeNow. It incorporates the integration of existing websites, too. It is simple to link the virtual assistant to the site as a plug-in, or to replace one frame of a page. Creating a link to existing pages is straightforward.

The Selfservice Company owns the Dialogue Manager technology behind the virtual assistant. We have developed the technology and we fully manage it. Close co-operation with clients and the modular build of the technology has enabled us to be flexible with our adaptations. That is why we can add functions that are specially adapted for specific projects.