Is A Virtual Agent Right For Your Web Site?

One of the questions I'm frequently asked by clients is if virtual agents are a good idea. Many of us have had frustrating interactive agent experiences over the years (recall Clippy, Microsoft’s animated paperclip that launched a thousand parodies).

Times have changed, and I think virtual agents are worth taking a look at. Today’s virtual agents can guide consumers through your Web site while answering questions effectively and conversationally. 

As virtual agent technology continues to become more sophisticated — features such as integrating with enterprise systems like shipping and delivery or CRM availability on mobile devices — virtual agents will continue to take on more complex customer service issues.

One of the benefits that I think is really compelling is that if a consumer escalates to live help, the transcript is pushed to the call rep, reducing call resolution times and sparing customers the annoyance of having to start from the beginning to explain their problem.

These features matter for many reasons. Here are two big ones.

  • Customers want immediate answers to their questions while they are on their purchase path.
  • Enhancing user experiences, deflecting customer service calls, and reducing call resolution times are important customer service priorities.

So how do you determine if a virtual agent is right for your organization? Conduct a customer service driver analysis.  (For more information on how to do this, Forrester clients can see my report “Selecting Online Customer Service Channels To Satisfy Customers And Reduce Costs.”

You're looking to see if:

  • You have a significant proportion of low complexity issues that go to live help (telephone, chat, email).
  • You have a notable proportion of customer service inquiries with repetitive answers.
  • Your abandonment rates are higher than they should be.
  • Traffic is coming to your help section from general search engines.

If you answer yes to any of these questions, a virtual agent is likely worth exploring. Here is a link to my report called “It’s Time Ro Give Virtual Agents Another Look,” which I hope you'll find useful.


Mobile Virtual Agents

Virtual Agents have been so-so for online engagement. Some are good. Others awful.

But where they seem to make the most sense will be in Mobile engagement?

No one seems to be doing this? But, if I'm out in a store and on a mobile web site, if I could text in a quick question and get a quick response, well, that would be like texting my kids. It would be really perfect.

Who's doing this?

Mobile virtual agents

Thanks for your comments. I agree - some virtual agents are very impressive but there are some awful ones too. The issue isn't entirely the virtual agent technology - the virtual agent can only pull out content in the knowledge base and an inferior knowledge base won't lend itself to a very good VA experience.

Virtual agents are on the brink of mobile. VirtuOz and Next IT announced support for virtual agents on mobile devices in earlier this year and they report that they have begun deployments that should be in production by the end of the year. So stayed tuned... there could be some interesting developments in this space soon.

One advantage of virtual

One advantage of virtual agents in mobile applications is that, for location-based scenarios, the knowledge base required is fairly confined. A contextually relevant engagement is more defined by the geography of the experience. Basically, bots can appear smarter. (If I'm in a store, the bot knows I'm in that specific store; that alone is a huge plus in terms of restricting a bots' required sphere of knowledge).

There are already products in the market. I saw Microsoft Tag tweeting about one this morning (apparently the first virtual agent that was resolved from one of their 2D tags). I'm trying to track these down since I'm looking for a job in AI and it seems like mobile and AI are meant for each other.