CRM and CX

9 reasons for boosting your CRM with virtual agents (IVAs)

Virtualisation of customer services interaction was quoted by Time (March 2008) as being one of the ‘10 ideas which are changing the world‘. Pascal Levy-Garboua, Director, Business Development of, a leading-edge provider of IVA (intelligent virtual agents) technology, is giving us more insight into the future of customer relationship management in this post written on behalf of Visionary Marketing.

9 reasons for boosting your CRM with virtual agents (IVAs)

9 reasons for boosting your CRM with virtual agents (IVAs)
9 reasons for boosting your CRM with virtual agents (IVAs)
Intelligent virtual agents improve Customer Relationship Management (CRM)

Intelligent virtual agents are changing the face of online customer service and are fast becoming the main channel of customer support communication for generation Y. However, the domain of artificial intelligence and its presence on the Internet today is still relatively unfamiliar to many. “Intelligent virtual agent” has yet to become a widespread buzz word in the field of CRM, but it is full of potential.

So what exactly are virtual agents about, and what can they be used for?

IVAs are artificial intelligence programs which translate to autonomous, graphically embodied agents which appear in an online environment. Graphic design for each individual virtual agent may vary from video animation to a photo or an avatar, in an interactive 2D or 3D environment, in order to fit with a company’s brand image.

Intelligent virtual agents can be designed and built with the capacity to fulfill a multitude of functions on the Internet. They offer a large variety of online functional abilities which allows their missions to be accurately defined according to their role in the customer cycle. Be it for marketing purposes, help and advice with website navigation, sales functions, customer support or sales follow up, intelligent virtual agents are increasingly being sourced by companies to bring more personalized interaction to their CRM.

IVAs in the CRM cycle

IVA strategies for improving online CRM

  1. Immediate and ubiquitous real time access to answers. IVAs offer customers immediate and ubiquitous real time access to answers using a variety of functionalities. Customers are thus saved from frustrating time on hold with a customer service center or, as is the case with traditional search engines and FAQ explorers, from trawling through mountains of web content in order to find their answer.
  2. A personalized, tailored service. IVAs can mirror the in store experience from a customer standpoint. The ability to connect to company back office databases allows agents to retrieve personal information to provide the most tailored answer.
  3. Encouraging online self help and increasing web content ease of access. Online virtual agents encourage online self service and increase web content accessibility to even the Internet beginner. Helping e-customers efficiently zero in on the information they are looking for is consequential to CRM, helping to alleviate customer frustration when they cannot find a certain page, or certain information. Through facilitating online services discovery, virtual agents are becoming an invaluable tool in cross and up-selling as well as in the maintenance and reinforcing of a company’s brand strength and recognition.
  4. Proactive approach to customer issues. By simulating a text chat session with a live agent, encouraging customers to visit certain pages, and offering co-browsing, the Virtual Interactive agent is an efficient and low cost method of internet guidance, helping people perform tasks such as locating information, placing orders, or making reservations. Virtual agents are proactive in their approach to customers’ issues, identifying the customer problem and providing a personal, pertinent answer. Occupying sales and marketing roles, agents can be an effective viral marketing tool, acting to increase a customer contact base using proactive questions and features which push the customer to action. Market studies have shown an actual increase in customer satisfaction after IVA implementation, owing to the virtual agents’ ease of use and the pleasant customer experience they provide.
  5. Personalized human-like interaction. The humanized nature of virtual agents makes them easier for customers to approach. IVAs have a personality defined to compliment their professional mission, which allows them to convey a more bespoke service and add a human touch to online customer care. It has been observed that this humanized approach is reciprocated from the human end, where a reported 70% of visitors greet and say farewell to virtual agents. Indeed in a world where e-commerce is becoming a driving force, increasing competition will require e-businesses to humanize online interactions, mirroring the in-store experience with more tailored, engaging conversations.
  6. Supplementing outsourcing with virtual sourcing. IVAs allow companies with large customer service flows to supplement service center outsourcing with Virtual Sourcing. An agents’ ability to handle unlimited requests at any one time helps to take a dramatic load off traditional call centers, reducing the need to invest more time and money as businesses’ CRM needs grow.
  7. The most advanced technologies offer a predictive dimension that allows real time adaptation to customer behaviour. Contrary to previous generation “keyword-based” chatter bots, recently developed IVAs, such as those produced by VirtuOz, can evaluate a customer’s previous behaviour and trends to predict the course of an interaction. This allows the agent to adapt its strategy in real time to meet the clients’ needs and fulfilling a variety of missions simultaneously.
  8. Existing call center productivity optimized. IVAs all have well calculated exit strategies to ensure the customer is never left with their issue unsolved and has not wasted their time. Agents can be interfaced with other CRM tools such as webforms, callback systems and live chat, to render the customer service process more efficient. Agents automatically categorize customer issues, enabling them not only to transfer to the most relevant customer channel, but also to only transmit information pertinent to the customer’s problem. This allows for quicker resolution, therefore increasing existing customer service centres productivity levels. Comprehensive reporting tools, where all conversations are registered for monitoring, allow stringent quality control and ensure that agents are performing to the required standard, a quality which of course is near impossible with human customer care.
  9. Integration with other web based services. IVAs, such as the recent French development Skaaz, ( can be integrated with Instant Messaging services, blogs and social networking sites; such artificial intelligence technology may also soon appear in virtual worlds, such as Second Life, where AI software will be able to profit from interactions with humans and increase its knowledge base according to its experiences.

Several major players in the business world have already jumped onto the band-waggon of excitement surrounding IVAs and their ability to revolutionize online CRM. The French arm of eBay,, integrated Louise, an IVA made by European leaders VirtuOz, onto their site in 2007 in a bid to reduce the high volume of emails and calls received by their call centers. After only 2 months the volume of emails received by’s customer service department diminished significantly, while email qualification was increased, allowing the existing CS systems to experience a sharp increase in productivity. Communications giant Neuf-AOL, also in France, economizes 1 phone call per customer per year with their IVA, named Chloé.

Virtual Agents and the future

Contrary to public belief, the roots of the technology were planted several years ago. Simple keyword based agents such as Colloquis’ Encarta on MSN Messenger and Ikea’s agent by Artificial Solutions provided a broad technological base from which more flexible and powerful agents have stemmed. Having already developed the predictive dimension now associated with agents, European leaders in artificial intelligence, VirtuOz, are currently in the process of developing voice recognition for their IVAs, which will add yet another dimension to online customer support. The popularity of agents is ever growing amongst brand companies who are looking to revolutionize their online approach to CRM and profit from rapid ROI. Customers can already access agents across a wide range of channels, and when IVAs become even more accessible over mobile media channels (PDAs, cell phones), agents will surely be commonplace not only for business CRM strategies across the web, but for personal use also. IVAs will give users even greater and easier access to information than currently is the case.

Yann Gourvennec
Follow me
Latest posts by Yann Gourvennec (see all)

Yann Gourvennec

Yann Gourvennec created in 1996. He is a speaker and author of 6 books. In 2014 he went from intrapreneur to entrepreneur, when he created his digital marketing agency. ———————————————————— Yann Gourvennec a créé en 1996. Il est conférencier et auteur de 6 livres. En 2014, il est passé d'intrapreneur à entrepreneur en créant son agence de marketing numérique. More »

One Comment

  1. Short of being an outlet for Virtuoz (which, btw, is only a leading edge provider in their own version of it), this piece is missing the most important part of the problem – and the reason we stopped using these in all previous iterations.
    Virtual Agents are the latest iteration of Chat Bots, and both of them suffer from the problem of maintenance. As much as we would love to think that a computer can have any type of conversation with anyone (ala Star Trek, more or less), that is very far from the true. Virtual Agents need to have specific conversations scripted for them. Thesaurus (similar words), onthologies and taxonomies (phrases and categories), etc. need to be prepared ahead of time and fed into the virtual agent list of possibilities.
    You need to create the conversations, feed them in, and monitor for new words and phrases and be constantly updating the thesaurus, taxonomies, ontologies, etc. In addition, instances that are not programmed are not completed and must be escalated. The costs per transaction are very high in comparison to simple self-service systems (in some cases as high as 4x higher), and the amount of maintenance work makes them being forgotten very easy.
    Yes, IKEA works, so does British Rail – but the scope of conversations you can have with them is very limited (in the case of British Rail the rate is success is very high since they only measure conversations they are “allowed” to have, and even then they can only solve 60% of so) and they don’t know how to make exceptions.
    We have seen these have a resurgence every now and then in the market, usually by a new vendor who is trying to make a name for an NLP (linguistic) tool they developed. We have never seen any of them be financially feasible in the long-run due to maintenance costs.
    In all honesty, if their technology is so good they will be better off putting it to work in structuring unstructured data from the huge amounts collected in feedback and social monitoring.
    Sorry, did not mean to rant – but want to make sure people are not confused by 12-15 year old technology that has not future.

    1. Hi Esteban,
      Thanks for your long and substantiated comment. The aim wasn’t for this piece to be an outlet for Virtuoz but to let them express their views, and let readers react … which is what you just did. Actually I share some of your concerns what with virtual agents. Not just that I don’t like them much – that in itself has no value, it’s just personal feeling – but because it falls short of understanding how much near real life avatars can actually look unreal. I covered that point in another piece at and another reader from Sydney (Nik at Big Click Studios) rightfully pointed out that this phenomenon is known as the UNCANNY VALLEY. You will find all relevant material on that phenomenon in Wikipedia. I hope that this addresses your point, unless you have a personal issue with Virtuoz that I don’t know of but in that case there isn’t much I can do for you. After all, it’s not because there is such a thing as the uncanny valley that people shouldn’t try and make money out of selling virtual agents.

      1. No, no issue. I just get a little upset sometimes when vendors write posts or articles that defy common sense and good practices, and want to prevent people from reading it as a biased piece without a rebuttal.
        I hope it did not come combative, as that was not my intention — just trying to provide balance.
        A colleague of mine forwarded the link to me saying that it was a great solution for customer experience, so i felt compelled to expand the conversation.
        thanks for the reply, for hosting the discussion.

        1. Well… it did come a bit combative, but that’s a by-product of passion I suppose. No issue either though, I just want to ensure that everyone having something interesting to say can voice his claim. Thanks for stopping by, I will send you an invite so that you can join my LinkedIn club if you so desire.

  2. How-do-you-do, just wanted you to know I have added your site to my Google bookmarks because of your extraordinary blog layout. But in earnest, I think your site has one of the freshest theme I’ve came across. It really helps make reading your blog a lot simpler.

  3. How-do-you-do, just wanted you to know I have added your site to my Google bookmarks because of your extraordinary blog layout. But in earnest, I think your site has one of the freshest theme I’ve came across. It really helps make reading your blog a lot simpler.

  4. A late response on this article from me.
    Please have a look at, a list of all virtual agents in the world. As you will see, the quality of the chatbots varies extremely. As Esteban point out: maintenance (together with smart technology) is key. At The Selfservice Company, 70% of the 75 member staff is working on knowledge maintenance for our clients. The scope of the question recognition isn’t limited: more than 90% of the questions are answered correctly. By the way: The Netherlands is a front runner in virtual agents. The Selfservice Company alone already answers 40 million questions a year (on a population of 16.5 million).


Back to top button

Discover more from Marketing and Innovation

Subscribe now to keep reading and get access to the full archive.

Continue reading