Edited: 7-17-2004

Originally written: 3-2-98
Subject: Thoughts on IPAs  (also wireless data)

I looked into  the area of Interactive Personal Assistants (IPAs).
(Sometimes called Intelligent Personal Agents.)

First, what is an IPA?
There is today's IPA and there is the IPA of the future....

Today's IPA is basically an electronic secretary. (e.g., Wildfire) You talk to the PC and it handles your correspondences for you. Other options may include keypads or keyboards instead of talking. Also, a phone, Personal Digital Assistant (PDA), or hand-held PC can be considered instead of a desktop or laptop PC.

The IPA of the future is a lot more sophisticated. A usage example is the travel agent feature. In Feb 1998, you tell your IPA that you want to go to Aruba sometime in May 1998 and to look for low fares...That's it. The IPA will do an extensive search for you (possibly via the Internet) and it may return with some options in a week or so... Even if you pick one of those options, it may still keep looking for a better deal until May 1998 is over. Also, after it "learns" your travel preferences, it may suggest trips in the future.

Business/Marketing View:

This section is based on the assumption that a successful product is a product that is ultimately used by consumers. However the are some business examples that don't fall into this category. For example, Hi-Fi stereo vendors add features to their products just so that they can claim more features than the competition. Some stereo customers will buy the products with the additional features even though they never use the features. Voice mail on cell phones appears to fall into this category for some customers. If IPAs fall into this category, then the following comments about ease-of-use and killer apps etc. do not apply. We can market hard-to-use products with rarely used applications and they will count just as much in the buy decision as easy-to-use products with widely used applications.

• IPA of the Future:
  Lucent's "Focus on the Future" (Vol.7 No1, Jan-Mar 1998) has an article called "Emerging Technologies: What's Ahead for 2001-2030". Periodically throughout the 1990s, George Washington University has called together a panel of prominent futurists, forecasters, and technical experts to predict when emerging technologies will actually emerge. The forecasts of the panelists tend to be within plus or minus three years of each other. In other words, they tend to agree with each other. The two forecasts w.r.t. IPAs are:
  • Computer Hardware (Personal Digital Assistants): Hand-held microcomputers are used by the majority of people to manage their work and personal affairs (year 2008)
  • Computer Software (Intelligent Agents): Knowbots, navigators, and other intelligent software agents routinely filter and retrieve information for users (year 2009)

  From the above, it appears that prominent use of the IPA of the future is about 10 years away. However, I don't have information on when initial use of the IPA of the future should start. Until I come across more info, I suspect that initial use will start in 5 yrs (2003-2004)
  

• Today's IPA:
  The February 1998 issue of "Computer Telephony" has two articles w.r.t. today's IPAs. "Personal Assistants: Wildfire & Webley...the secret lives of CT Slaves" and "Speech Recognition Auto Attendants: Operator, can you hear me?" It looks like there are already products on the market with Wildfire and Webley as the market leaders with over half a dozen smaller players chasing them. Also, I heard that even though the R&D community is excited about these types of services, Wildfire has been out there for years and has not made money. I would suggest that there be a strong business case before there is another  market entrant with a "me-too" product.
  
• Wireless Data:
  Wireless Data is also appearing to be slow to take off. The Nov 3, 1997 issue of Telephony has an article called "A Road Less Traveled". It mentions a sparsely populated Wireless Apps '97 conference that searched for "killer apps" but did not find any. It appears that the often mentioned stock quotes, sports scores, weather reports, and limited e-mail browsing are not enough to carry this service.

  From a consumer's point of view I agree. I have a personal interest in obtaining the above info (e.g., stock quotes). But there are other media I would prefer to get them from. When I am on the road, getting this info out of a cell phone is so awkward, I would prefer to wait until I have access to a PC (Home, work, hotel, limo, airport, airplane) and get the info from a browser.

  Can't wait to get stock quotes? As a stock holder myself, getting quotes is only a curiosity. I can look at market close info at the end of the day. However, being able to make a trade during the day is important. Today, I can place a standing order (limit order) to trade a stock at a specified price. It will execute whenever the stock reaches the specified price... even days or weeks later. In fact I did this once. I entered an order one morning and it executed later that day while I was in a meeting...no need for wireless data.

  Also, the targeted road warrior market is viewed as too small to carry this service. They may be the initial users, but the general public must get involved for the service to take off. What are some of the key factors to getting a popular wireless data service:

  • The biggest factor is having an EASY TO USE interface. From many sources, I hear that pushing buttons on a phone is only useful for making phone calls. Trying to do anything else with it becomes very difficult and therefore is not used by the vast majority of customers. If they don't use it, they may eventually cancel their service. This message was repeated at an IN/CTI conference I attended. I have also had the same experiences. I carry a cheat sheet for my VoiceMail commands. I also have to use the manual to program the name dialing into my cell phone...and I'm more technical than the average person! A visual (e.g., web browser) interface appears to be the only way to go. The Web itself did not become popular until browsers became prevalent in early 1995 even though the web has been around since 1991 and the internet has been around since 1969! A PalmPilot with a Netscape-like web browser interface may do the trick. By the way, passive services where the user does nothing (e.g., caller ID) seem to be the most used.
    
  • A reasonable flat-rate for billing may be an important factor...say $20-$30/month
    
  • Some truly compelling services for the general public (i.e., killer apps) appear to be a key factor.

  Today, the biggest player in wireless data appears to be Unwired Planet. They supply servers and microbrowsers. They are also getting industry support for their protocol, WAP. 

Summary:

• Tomorrow's IPAs- More market research is needed to pinpoint when the market and this technology will emerge. If it is really 5 years away, it may be an appropriate project to start working today...
• Today's IPAs- Wildfire is out there. A business case is needed to justify funding a  a "me-too" product
• Wireless Data-  Unwired Planet has a big lead in the industry and in the market. Also, the wireless data market appears to be soft, lacking killer applications and lacking easy to use products. Again, technical questions appear to take a back seat to the market questions mentioned.

 Anthony Clark