John Betz was been with Microsoft's MapPoint Unit almost from the very beginning.He was one of the visionaries for the MapPoint product and now gives us his thoughts on the company's direction on location technology in product development.Editor-in-Chief Joe Francica sat down to talk with Mr.Betz at the CTIA conference in Atlanta on March 22.
Joe Francica (JF): What is the strategy for the MapPoint Business Unit and the
long-term outlook for where Microsoft will develop location technology
for applications?
John Betz (JB): At the highest level, the company views location as a strategic
asset worthy of investment over the years.And that investment obviously
started on the consumer CD ROM side and now we want to bring that same
core technology to business productivity, from a certain standpoint,
recognizing that location is best seen or is most valuable as an ingredient
in other applications.We've got the "religion" about .NET and Web Services;
you've seen us come out with MapPoint Web Services.Now, I think what you
are seeing is the proliferation of mobile devices; mobile computing is
taking off.GPS chipsets and other location acquisition technology is taking
off.Now, we want to combine that investment that we have in client side
software with Web Services.So if you can bring those two assets together,
you have what is called "informed mobility."
JF: There seems to be more happening at the "enterprise level"
than at the consumer level.Many applications depend on location technology.
What's the next step from a "business intelligence" standpoint from Microsoft's
point of view?
JB: I think there is a existing recognition, where the ROI (return
on investment) works, and I can equip my mobile assets with radios and
GPS chipsets; people have done that.Very expensive; I've got to license
the data; I've got to bring in some GIS expertise.That's really the value
proposition behind the Location Server - lowering those barriers to entry;
single API (application programming interface); everybody has handsets.
You can "location enable" those applications that warrant it and it's a
much lower level of investment.The net effect will be location in the
next tier down of enterprise applications.You'll also see it go the other
way.So with smart devices, you will have mobile productivity scenarios
where it can integrate with (Microsoft) Outlook and make contact lists
and know where customers are nearby.Workers will have that "mobile" experience.
JF: Do you see the company using the MapPoint Business Unit to
help spatially enable "back office" applications in Finance and Customer
Relationship Management?
JB: We have done some of that in the desktop product, like the
MapPoint CD.There are other scenarios that don't involve mapping at all
and I'll give you one that is pretty easy to recognize the value for and
that is expense reporting.So, corporations could save millions of dollars
every year in just having a user say, "OK, I drove from my office in Redmond
to SEATAC (Seattle-Tacoma Airport), and I think it was about ten miles,"
but instead have it calculate out the miles exactly.So its a non obvious
use, but it is one where you could walk into a CFO and say, "You know what,
I could save you a bunch of money."
JF: MapPoint is now in its third revision.Was MapPoint conceived
as an end-user desktop business client only, or was the thinking that it
would seed a market for application development and, now with an open API,
become an established development platform?
JB: You know, it's interesting, I wrote the vision document
for MapPoint CD ROM and it took us about three versions for us to get to
the vision as it was originally described.And I think initially our target
was just the interactive user and let them map data in an easy way from
Excel, etc.What was important for us to do in the second version, was
to open the object model up so developers could drive applications.So,
to answer your question, the initial target was office user, but we always
had a mindset towards, just like the other Office applications, exposing
the functionality so developers drive it with VBA.
JF: To me, there was always a funny dichotomy in the product.
It had all of the "bells and whistles" of a desktop mapping product; beat
a lot of them, certainly from a price point level.And when I saw it come
out I thought it was a good product, but knew it would never be a competitor
to MapInfo or ArcView.But then you started to add functionality like "drive-time"
and I thought that was a "higher-end" application that really targets the
business customer; not typical for just a desktop mapping product.What
was the reasoning for including this type of functionality?
JB: That was one where we listened to customers.It is always
a trade off between what people want to do and what we could accomplish
in any given product.So, isochrons was one that we knew how to do, and
we thought it was a horizontal enough scenario that a lot of different
customers would want to accomplish, whether it would be direct mail or
any type of catchment of customers.We thought the "proximity to customers"
would be one (application) that the interactive customer would actually
find that valuable.
JF: Is there any inclination to go deeper with spatial analysis
and querying capabilities within the product?
JB: In the MapPoint CD ROM product we rely on external data
sources.We have an internal format that once you load your data in we
can do those spatial kinds of operations but we don't have this ad hoc
query engine that would allow you to ask that kind of question.What you
really want is spatial grammar added to SQL to answer those questions.I
don't see us doing that because it is a pretty "hard core" specialized
function.
JF: So, if you don't do it in MapPoint, will you allow MapPoint
to query Microsoft SQL Server?
JB: What you really want are the spatial extensions to SQL Server.
JF: OK, let me ask that...Is there going to be spatial extensions
to SQL Server or the ability to add user-defined operations?
JB: Yes, that's the stated direction (to enable user defined
operations).Within the MapPoint group, we have had that on the radar for
some time.It's a market where Oracle shows dominance in.The SQL team
is tracking it, and know about it and look to us as the spatial experts
of the company to help them solve that problem.It has never made the "cut
line." They have a bunch of work to do for "Yukon" (Microsoft code name
for the next version of SQL Server), and spatial is really interesting,
they'd like to do it, but not in this release.
JF: Thanks, these have been some good answers...I appreciate your openness and your time.