It's that time of year where we bid goodbye to
what we did not like about last year and look forward to our best hopes and
wishes for this coming year. As I look back on the past year in BI&A, here
is what I'd like to see in 2016:
1.
Some real standards
the software vendors will respect
I'll admit I've
wanted this for a long time, but I can still hope. We are now in the big-data
driven third generation of BI&A technology. The first was relational databases.
Then, the industry settled on SQL as the standard query language and that
facilitated a whole industry with interoperable query tools, ETL tools,
database platforms, and a generation of expert professionals. The second was
multi-dimensional technology. In this case, we never even saw a standard for
query languages aside from some weak attempts to extend SQL. Now 'No SQL' is
the emerging non-standard. Metadata standards? Dream on. The one credible
attempt at it, the Common Warehouse Metamodel in the 90's, never really caught
despite support from several major vendors. Those same vendors eventually
decided that proprietary solutions could become de-facto standards if they
acquired enough market share through buying out smaller competitors. So
metadata standardization is achievable, but mostly with single vendor
solutions. This brings me to my second wish.
2.
The next consolidation
wave
As is always the
case during a technology consolidation, you get a wave of startups with new
technology and established players trying to reposition older technology in a
fight to win over the early adopters in the market and some love from the
industry analysts. In time, the weaker players drop out of sight and the
stronger ones get swallowed up by the big enterprise players looking to buy
technology and market share. It will be a little different this time as SAAS
deployments will allow more of the upstarts to thrive on their own. Some may
end up pushing out older players in the process. For me though, there are just
too many incomplete solutions out there right now and the consolidation wave
cannot come fast enough.
3. Fewer Big Data wannabees
Another consequence
of this technology shift is the emergence of a huge number of resumes on the
market promising a depth of knowledge and experience in big data technology
stacks that is more hype than substance and becomes clear 10 minutes into an
interview session. I’d rather work with seasoned BI&A professionals that
have mastered the basics of software engineering, project management,
requirements development, data governance, and testing who have demonstrated
the ability to learn and adapt to new technology quickly.
4.
More agile BI&A teams and less Agile
methodology zealotry
BI&A pros have
known for decades that agility is mandatory in our work and waterfall methods
do not work. Requirements are generally not completely known in advance and
only revealed through prototyping and iterative development. Value should be
delivered on an ongoing basis. On the other hand, some of the main tenets of
the now-revered Agile movement do not work all that well in the BI&A space.
We do not develop structured applications as much as we strive to create
environments where our users can create their own applications. This limits the
development of workable user stories in advance. The evolution of our platforms
over time and need for stable development and support teams does not lend
itself well to the scrum concept as it is typically advocated. Yes, we need to
be agile (small A) but not necessarily Agile (capital A.)
Business
Intelligence and Analytics has never been more recognized as vital to success
in business, government science and education. Our tools and technology are
better than ever. Those wishes have come true. Now, I wish for all who read this a happy and
successful 2016.