Can you feel it? The Twittersphere is spinning faster on its axis, anticipating the next Forrester TweetJam, Wednesday, December 15 at noon/1 p.m. U.S. Eastern Time. Our topic will be “Advance Your Analytic Strategies.” We’ll have a Forrester tweeting-squad that includes yours truly plus any or all of the following Forrester colleagues: Rob Karel, Boris Evelson, Clay Richardson, Gene Leganza, Noel Yuhanna, Holger Kisker, Leslie Owens, Suresh Vittal, William Frascarelli, David Frankland, Joe Stanhope, Zach Hofer-Shall, and Henry Peyret.

Clearly, advanced analytics is an important theme and huge space that sprawls across many Forrester analysts’ focus areas.  What I’m about to present are the humble opinions of one Forrester analyst — c’est moi — for whom this is the heart and soul of his focus going forward. These thoughts, presented under each of the proposed TweetJam questions, will give you a foretaste of what I’ll tweet on that session in just a few weeks:

What exactly is and isn’t advanced analytics?
KOBIELUS RESPONSE: In a Forrester report from last year, Leslie, Boris, and I scoped it out as follows: “Any solution that supports the identification of meaningful patterns and correlations among variables in complex, structured and unstructured, historical and potential future data sets for the purposes of predicting future events and assessing the attractiveness of various courses of action. Advanced analytics typically incorporate such functionality as data mining, descriptive modeling, econometrics, forecasting, operations research, optimization, predictive modeling, simulation, statistics, and text analytics.

My current thinking on this topic is: “man, that’s wordy!” Let’s net this out. What is advanced analytics? It refers to approaches for fusing intelligence of the past, presen, and future. I like to refer to this as the “synoptic” view of decision time horizons.

What is advanced analytics not? It’s not basic analytics — in other words, it’s not traditional BI, which is all about gleaning intelligence from the historical data.
As such, this graphic, which I presented at a DAMA Philadelphia chapter earlier this month, goes to the heart of how we’re scoping advanced analytics down to the product segment level:

What are the chief business applications of advanced analytics?

KOBIELUS RESPONSE: If you look at the graphic above, you’d be inclined to flip this question around and ask: What aren’t?
If you simply stick to the core of predictive analytics and data mining (PA/DM) — per my Forrester Wave from earlier this year, you can point to diverse applications in front-office business processes (marketing, sales, customer service, etc.) and back-office functions (finance, operations,  etc.).

If you also consider multivariate statistical analysis and constraint-based optimization, which are the core of operations research, it’s clear that advanced analytics is pivotal to modern business practices in risk mitigation, scheduling, transportation, logistics, network optimization, algorithmic trading, and beyond).
If we include all the approaches often lumped under text mining or content analytics, it’s clear that this is at the heart of e-discovery, enterprise search, and social-media listening platforms.

If we factor complex event processing (CEP)into the mix of advanced analytics, we must consider business activity monitoring (BAM) and real-time dashboarding as key business applications for this technology.
And if you’ve been reading my published output recently, you’ve noticed that social network analysis — a.k.a. behavioral analytics – is coming on strong as an approach for fine-grained customer segmentation and next-best-offer targeting.

Here’s a graphic I’ve been using recently to show how advanced analytics technologies are converging as enablers for the new vision of “next best action everywhere” across all business processes:

What’s the overlap between advanced analytics and our BI, data warehousing, and data governance/quality initiatives
KOBIELUS RESPONSE: Clearly, traditional BI anchors the “past” (a.k.a. historical intelligence leg of the advanced analytics stool). You can't have a comprehensive 360-degree synoptic time view without an enterprise data warehouse, which is inexorably virtualizing into a distributed, cloud-based, complex-content infrastructure that persists historical data, supports CEP and other low-latency ingest/query/delivery, and executes predictive models inline through in-database analytics. And if you are not loading cleansed, standardized information into that synoptic time view — per your data governance/quality and master data management initiatives — then your view is distorted, misleading, or just plain wrong. In that case, your “advanced” analytics are merely highfalutin algorithm-packed falsehoods.

Is advanced analytics ready to roll out to all information workers, or is it still the province of a priesthood of data mining specialists?
KOBIELUS RESPONSE: The observation I put forth in a blogpost earlier this year still stands: The core problem with today’s advanced analytics offerings is that many of them are power tools, not solutions suitable for the mass business market. Their core user base consists of statisticians, mathematicians, and other highly educated analytics professionals.
No one denies that traditional advanced analytic tools are the analytical bedrock of mission-critical applications in diverse industries. But I still challenge you to point to a single case study where they are used directly by the CEO, senior executives, or any other casual user, rather than indirectly through being embedded in some custom application. Over the coming several years, they will take up residence in many vendors’ BI solution stacks, but still be employed primarily by power users, rather than by casual users, for whom canned historical reports are sufficient. Most power users will primarily require lightweight advanced analytics and pre-integrated models integrated into their BI visualization layers.

How should process intelligence impact your analytic strategies?
KOBIELUS RESPONSE: Per another blogpost from earlier this year, process intelligence is all about identifying what works and doesn’t work in your current workflows, both human and automated. It’s a key focus for us here at Forrester. Organizations must deploy tools for tracking, exploring, analyzing, anticipating, and optimizing all processes, hopefully in a continuous, real-time fashion. It’s clear that this must include, at the very least, business activity monitoring tools, which roll up key process metrics into visual BI-style dashboards for operational process managers. Likewise, historical process metrics should be available to the business analysts who design and optimize workflows. And each user should have access to whatever current key performance indicators are relevant to the roles they perform within one or more processes.

Can advanced analytics help us with our social media strategies?
KOBIELUS RESPONSE: You better believe it! In yet another blogpost from this year, I spelled out how advanced analytics can figure into your social media strategies, in at least four ways:

  1. Monitoring social media for marketing, customer, brand, and competitive intelligence: Social media analytics leverage BI and advanced analytics tools — reporting, dashboarding, visualization, search, predictions, text mining, etc. — to find patterns of awareness, sentiment, and propensity among current and potenetial customers, as surfaced up from social media such as Twitter and Facebook.
  2. Mining for patterns of influence and expertise: Social network analysis is advanced analytics that is specifically focused on identifying and forecasting connections, relationships, and influence among individuals and groups. It mines transactions, interactions, and other behavioral information that may be sourced from social media, and/or just as often from CRM, billing, and other internal systems.
  3. Integration of real-time social intelligence into internal processes: Social media monitoring is real-time analytics that uses CEP to acquire, filter, and display issues, exceptions, and other events surfaced from social media, so that alerts can be forwarded, workflows triggered, and response loops set up in internal operations.
  4. Infuse social-style interface into your BI user experience: Social intelligence refers to the trend toward incorporation of social network style interaction models — such as those associated with Facebook and wikis — into the BI user experience.

But those are just the questions we’ve scoped out for the TweetJam. These are all broad best-practice scoping questions. Rest assured that I’ll also tweet, where appropriate, on the most important developments this past year in advanced analytics, with predictive analysis as its core. Here’s what I see as most significant:

  • Continued emergence of enterprise-grade Hadoop solutions as the core of the future cloud-based platforms for advanced analytics.
  • Development of the market for analytic solution appliances that incorporate several key features for advanced analytics: massively parallel EDW appliance, in-database analytics and data management function processing, embedded statistical libraries, pre-built logical domain models, and integrated modeling and mining tools.
  • Integration of advanced analytics into core BI platforms with user-friendly, visual, wizard-driven, tools for quick, exploratory predictive modeling, forecasting, and what-if analysis by non-technical business users.
  • Convergence of predictive analytics, data mining, content analytics, and CEP in integrated tools geared to real-time social media analytics.
  • Emergence of CRM and other line-of-business applications that support continuously optimized “next-best action” business processes through embedding of predictive models, orchestration engines, business rules engines, and CEP agility.

In terms of the principal advanced analytics trends I see in the coming year, above and beyond deepening and adoption of the above-bulleted developments, the following stand out:

  • All-in-memory, massively parallel analytic architectures will begin to gain a foothold in complex EDW environments in support of real-time elastic analytics
  • Further crystallization of a market for general-purpose “recommendation engines” that, operating inline to EDWs, CEP environments, and BPM platforms, enable “next-best action” approaches to emerge from today’s application siloes
  • Incorporation of social network analysis functionality into a wider range of front-office business processes to enable fine-tuned behavioral-based customer segmentation to drive CRM optimization

We invite you all to join in the TweetJam a few weeks from now. Remember: the hashtag is #dmjam.