Article reviews how to deal with the deluge of new technological options and the aspects of a strategy for quick, high quality implementations of emerging technologies. Based on company success stories, article lays out what will work.
201207 Tech Decisions: 5 Keys to Fast Successful New Deployments.pdf
1. 5 Keys to Fast, Successful New Tech
Deployments
By Steven Callahan, Robert E. Nolan (mailto:)
July 19, 2012
The acceleration in new technologies and associated business needs is creating an almost
plaguelike spread in an accelerating use of pilots. The relative calm –or at least the
managed chaos –of recent years' large scale implementations, conversions, integrations
and in or outsourcing projects has recently been broken by the increasingly loud buzzing of
social media, predictive analytics, big data, social mining, mobile technology and cloud
computing "pilots". Unfortunately, unlike the governance and oversight that typically
comes with large projects, this flurry of pilots is occurring under mixed circumstances.
A conversion or integration or outsourcing clearly takes several yearsÍľ putting up a
Facebook page, building a mobile app, announcing a Twitter account for servicing, moving
select simple processes to the cloud, or contracting with experts to discover social data for
claims adjudication all can occur in a relative blink of an eye. And in most cases these
"blinks" occur very inexpensively, putting them below the radar. Structured methodologies
and project management offices can be confounded trying to address the combination of
volume, diversity, and velocity of new project demands. Market pressures amplify the
demands for immediacy while proven simplicity erodes the barriers of control. For those
who have been in the industry long enough to remember, the inevitably rapid onslaught of
structural and operational change is a diversified equivalent to the introduction of the
business PC — but at an extremely abbreviated run rate.
Customer service is critical, and technology adoption inevitable. The challenge rests with
management of the many concurrent efforts. While many companies are approaching
moving these to market under the umbrella of pilots, the shortterm characteristics of these
technologies require a more diligent approach. By following straightforward guidelines,
greater success with less pain can be achieved even at faster rates.
•Even if a miniaturized version is used, make sure there is a steering committee,
executive sponsor, business and technology partners, and a well communicated
purpose and scope. The other important element, particularly when working with pilots
and proofs of concept, is to timebox the project with a relatively short horizon. Absent the
parameters of a timebox, these smaller projects can run forever as they multiply with
retries and new technologies. Discipline and rigor should be used early to set a firm
timebox for the project to work against.
2. •Determine early if it is a pilot, a proof of concept, or a controlled rollout. This single
decision, often missed, has major downstream implications on the eventual outcome.
Rollbacks do not work with controlled rollouts, and are difficult to apply to pilots that are
not tightly managedÍľ that means some major function or customer block is now resident on
a system with no place to go. Knowing early if there will be a rollback plan and if so what it
will be helps tremendously. Deciding what to do with the 120 customers on the broken
system is best not done at the last minute.
•Spend a brainstorm session or two early in the process discussing end of project
"whatifs". What if it doesn't scale? What if it doesn't do everything desired? What if the
vendor experts disappear? What if it works but the customers don't want to use it? If you
know what you might do, you can incorporate hooks along the way that will make doing it
easier and faster.
•Here is a difficult one. If at all possible, go through the various transactions and
features that are planned and decide which ones are showstoppers, which ones are
needed but can come later into production, and which ones are just "nice to have." It is
often difficult to create this list early given the project usually is, by definition, only the
"must have" items. And yet projects that start with everything in the showstopper
category that run into issues moving to production suddenly are able to reprioritize some
features and functions as not required for production. More often than not, the
reprioritization is an end of project rushÍľ however, if it can be done up front it will be a
calmer and more reasoned process. Worse case, the project runs its course, the timebox is
exceeded, and there are still showstoppers.
What to do?
1. If there is a rollback strategy, use it. Yes, money was spent, but if the overall assessment
is less than ideal, consider it a learning experience, and exercise the rollback. Use the
experience to approach a new project on a more informed basis versus patching the
broken one.
2. No rollback is more often the case than not. The team must take a disciplined look at
every defect, missing feature, or absent function. Not only must each be reprioritized, but
the defects must be rated by criticality, impact, time to fix, and alternative (manual or
another system for example).
3. The elements of criticality, impact, time to fix, and alternatives have to be combined into
a resourced plan to move the project to a minimal level of acceptability as defined by the
sponsor and business. Achieving this baseline, regardless of how low compared to initial
plans, is critical to gaining the necessary foothold for next steps.
4. Handson, extremely detailed project management focused on progress to plan, new
defects, prioritizing and reprioritizing as things change, negotiating with the business, and
brainstorming alternative means to solving each issue is all required at this point.
3. •Upon reaching the initial milestone, whether for a pilot, controlled rollout, or proof of
concept, another often missed but key step is the development of a project review
document. For some projects, less than 5 pages may sufficeÍľ for others, it may need to be a
bit longer. The important point is to capture the successes, shortcomings, remaining known
issues, updated cost and scale data, and next steps with a rough prioritized timeline. It is
also important to get the sponsor and business involved to learn what else is functionally
needed and the desired next steps.
Much of this likely sounds familiar in one form or another. Most companies have matured
considerably over the last few decades since the influx of PC's and Infocenters and all the
accoutrements of the time. Yet with that maturity have come a number of changes in the
marketplace and with the workforce. Service expectations in the field and with customers
combined with continued challenging economic conditions are together putting immense
pressure on executives and CIO's to leverage any possible relevant and applicable
technology. The industry is more technologically dependent than 10 years ago, has growth
and margin pressures, and is facing a veritable flood of advances in tools, platforms, and
even environments like virtual worlds and social networks. Slippage and errors are
inevitable in such a fastpaced and challenging situation –they also seem to be less easily
forgiven. Anything that can be done to bring coherence and rationale to the process of
meeting the demands for rapid adoption and deployment should be shared, discussed, and
leveraged wherever possible for mutual benefit.
Think of it this way: in the end, Denzel Washington did indeed stop the train in the movie
"Unstoppable"Íľ however, I would not have wanted to have been a passenger! Avoid the
wild ride with tools, techniques, and practices that help the industry continuously improve
technologies adoption rate and speed of deployment.
About the Author: Steven M. Callahan, CMC is a practice director for the Robert E. Nolan
Company, a management consulting firm specializing in the insurance industry. He can be
reached at Steve_Callahan@RENolan.com.