It was just about 18 months ago that business intelligence (BI) players started
talking about the mid-market -- the small- and medium-sized enterprise (SME)
segment that industry watchers say is poorly served by many commercial,
off-the-shelf (COTS) BI packages.
With everyone from then-independent players Business Objects SA, Cognos Inc.,
and Hyperion Solutions Inc. -- along with heavyweights such as IBM Corp.,
Microsoft Corp., and SAP AG -- expressing interest, this latest incarnation of
mid-market concern looked like it might actually take (see
http://www.javareport.com/article.aspx?id=20182 ).
So, a year and a half later, how are they doing? Has the BI industry succeeded
in breaching a mid-market nut that's proven to be historically difficult to
crack?
More importantly, how have BI players responded to the specific needs of
mid-market customers? Are they delivering solutions that, in one way or another,
make it easier for resource-strapped mid-market customers to take a BI plunge?
It's starting to look like it. At last month's TDWI Spring World Conference in
Chicago, for example, a surprising number of BI players wanted to talk about the
mid-market. These weren't just start-up or traditional SME players, either.
These were BI giants such as SAS Institute Inc., whose midmarket presence has
hitherto been confined to researchers or statisticians in SME organizations.
Also eager to talk about the SME market was respected heavyweight (and fellow BI
independent) Information Builders Inc. (IBI), which -- though it has had its
share of midmarket success -- is more frequently associated with huge reporting
deployments in financial services, insurance, and other verticals.
Even data warehouse (DW) appliance players, which have typically focused on the
Big Data crowd (e.g., retailers, telecoms, financial services), are talking
about reaching out to mid-market customers.
"Up to this point, we've been selling speed and feeds to technologists. Someone
has a data warehouse that's performing very badly. They're investing lots of
money in performance tuning. Now they're saying, 'You know, we've come as far as
we're going to by adopting this approach.' Nowadays, most people are looking for
a business solution. They don't just want to throw technology at a
general problem. They want specific solutions designed to solve
specific business problems," says Tim Young, vice-president of marketing
for Netezza.
"At the moment, we're retooling our operations to start focusing more on
specific industries. We'll be focusing on specific verticals, and not just in
the high-end, where we've [traditionally] played. We think there's a real
opportunity in the sort of medium enterprise [segment]," Young explains.
He cites his company's vertical-specific efforts with the Netezza Developer
Network (NDN) that it launched in September of 2007. The goal of the NDN is to
enable software developers to incorporate their own programmatic logic into
Netezza's snippet processing units (SPU). SPUs do the data processing
heavy-lifting for Netezza's Performance Server (NPS) appliances.
What Netezza is doing with NDN appeals to SME customers, Young maintains,
because it gives them a measure of control over their BI and data warehousing
(DW) implementations and let's them more easily adapt NPS to suit their own --
often very specific -- requirements.
"We have about 70 companies that have signed up [for NDN] already, and they're
building applications … that range from all sides of the spectrum," Young
explains. "They're doing application-side processing, [where they're] basically
embedding [application logic] into the Netezza SPU, so as the data streams off
the [hard] disc, all of this analysis happens at this kind of raw data level,
such that the only thing that gets passed back to the client is basically the
results of a data set," he continues.
Mid-market customers are particularly deaf to products that are long on
technology prescriptions and short on business specifics, Young notes. As a
result, he hopes, Netezza's NDN push -- with its SDK and open APIs -- will help
prime the mid-market for its NPS appliances.
"We can deliver [to customers] these business-oriented applications, where the
value of those applications is absolutely crystal clear. It's for specific
applications -- things like re-rating or geo-spatial-type simulations -- so
[customers] know what going to get before they buy," he observes.
The Control Conundrum
Another thing that Netezza's approach gives customers is control, and as
Software Labs Inc., a developer of xFusion Studio and xFusion WebDbServer (data
integration, or DI, middleware), can tell you, mid-market customers crave
control.
"They're much more concerned about it than they are in the enterprise," says
president and CEO Pradeep Tapadiya. Software Labs should know: it caters to SME
customers. Contrary to what you might think, Tapadiya argues, mid-market
customers want more flexibility than large enterprises do.
"[SMEs] are quite different. They'd like less hand-holding, more control. From
my experience, what happens is that [if I'm a mid-market customer], even if I
outsource -- and more frequently, they are looking to outsource -- I always want
to feel that I'm in control, because if things go out of control with this
outsourcing, you need to know [that you can take it over yourself]," Tapadiya
explains. "What [midmarket customers] want is to make sure that they're not
dependent on any company."
Tapadiya doesn't buy the claim that SME customers are comparatively
resource-strapped, at least when it comes to IT talent. In many cases, he points
out, SME customers are getting by largely on the game-breaking talents of their
in-house IT staffs -- performing, for example, highly complex data integration
(DI) extractions, transformations, and loadings by means of programmatic SQL.
They're doing this, he argues, because COTS ETL tools have simply been too
expensive, too bloated, (or too training-intensive) to justify their costs. As a
result, he says, sometimes hotshot SQL programmers are literally stitching
together top-to-bottom DI programs.
There's a flip side to this coin, of course. Although the claim that all
mid-market companies are resource-strapped clearly isn't true, the fact remains
that plenty of SME firms could use a helping hand, at least when it comes to an
esoteric practice such as data integration.
What mid-market customers typically want are solutions that let them cater to
either in-house talent (e.g., SQL experts) or to critical stakeholders (e.g.,
novice business users) -- or both. In this sense, Tapadiya concedes, they're a
lot like large enterprise shops. However, rather than lagging behind the
enterprise market in this regard, he argues, the mid-market has actually been in
the avant-garde. Some of the usability features that are now cropping up in the
Big Enterprise ETL tools -- e.g., drag-and-drop UIs and a more collaborative
user experience (to better involve business stakeholders) -- have been
part-and-parcel of Software Labs' mid-market DI product for a long time now,
Tapadiya maintains.
"If you look at all of the functions that we have within our product, somebody
who's highly technical, they can probably do everything in SQL Server writing
complex queries. For many [SMEs], that's what they're doing," he contends. "What
companies have been saying is that they also need some simpler functions so that
they can involve [their business stakeholders] in the process, so we've tried to
focus on making [our tool] as simple to use as possible. We have a two hour
training program. At the end of the two hours, they're ready to go."
Software Labs patterns xFusion Studio on a very familiar paradigm: the Microsoft
Office suite. The xFusion interface isn't a replica of the Office UI, Tapadiya
concedes, but it does shoot for a similar user experience, with Office-like
toolbars, menu items, and other features. The idea, he says, is that once
customers start using xFusion, they get quickly up to speed -- in part because
they're assimilating a kind of portable knowledge from their other UI
experiences.
"We wanted to deliver the same kind of look and feel so that people know what to
expect, so if you open an existing [xFusion] document, or you save it, it's
really very similar to what you're familiar with [in Office]," Tapadiya
contends. "That's been one of our goals [in the mid-market] all along: to make
sure that people feel comfortable using the product. That's something that only
recently you're starting to see in [DI tools aimed at large enterprise
customers]."
A Different Kind of Customer
Software Labs likes to say that it's ahead of the curve, so to speak, with its
long-time emphasis on mid-market BI, but BI and DI vendors who've gone through a
come-to-mid-market transformation find that the economics of that segment --
which are very different from the large enterprise space -- appeal to them.
Take Michael Corcoran, chief marketing officer with IBI. He seems all but
smitten with mid-market buyers.
Such customers typically eschew much of the thrusting and parrying that large
enterprise customers -- accustomed, for example, to evaluating a fixed number of
vendors for an RFP -- typically employ, Corcoran points out. "The midmarket
companies are just a pleasure to do business with. There's no nonsense. There's
no evaluating-technology-for-the-sake-of-evaluating-technology. They just don't
have the resources. They don't have evaluation teams like that," he comments.
"What we hear, instead, is, 'Hey, I've got a problem, and if you can prove to me
that you can solve it, we'll do business. It's a much shorter cycle."
There are other benefits, too, according to Corcoran -- benefits that serve both
client and vendor alike. "You tend to get very high-level buy-in from the
executive team, you're certainly going to meet the CIO very early in the cycle,
and sometimes you're able to meet with all of the stakeholders at one time," he
indicates.
"There's such buy-in to the problem [i.e., why do BI] itself. There's such
awareness, from top to bottom, in mid-sized companies," Corcoran continues,
contrasting this awareness with what he describes as a kind of "segmentation" in
the large enterprise. "When you find problems in a large enterprise, you're
going to find it at a business level-unit. You're going to have to somehow map
that back [throughout the enterprise]. It's just a much longer, harder process
to do."