New Year's Resolution: Pilot innovative analytics

With the new year coming, companies are thinking fresh regarding innovative programs they will be running. Examples may be:

* multichannel marketing,
* digital interactive sales aids on iPads, perhaps with segmentation,
* social media presences, or
* online portals or communities.

With these efforts and the appropriate measurement planning, will come innovative data sources for your companies. Your new year's resolution can be: let's explore analytics for these new data sources, learn how to analyze, and discover what trends there may be.

The right analytics partner can help in translating your business objectives into analytics frameworks. Your partner or staff can also determine the approriate exploratory analytics software. The image posted here is one example of a data visualization and dashboarding tool called Spotfire, of which I have found valuable during yeras of use for detecting trends and bookmarking insights. In addition, I remain a frequent user of Netbase for social media, SAS products for data cleaning and statistics, and Angoss data mining software. There are more tools, a great resource to find them is KD Nuggets.

So, don't wait, make it your resolution to dive in and mine those behaviors.

High Stakes Business Analytics Across Industries

This Wednesday I had the pleasure of participating in the first Aberdeen Business Analytics Summit in Boston. It was a fascinating collection of business and I.T. speakers across industries: banking, packaged goods, semiconductors, auto, health, etc. These business leaders have critical jobs like forecasting customer demand, optimizing the supply chain, or monitoring efficiency and profitability at retail locations across the country. Having access to real time, insightful analytics is critical to performing these jobs.

I participated in a panel on how better alignment can be achieved between "business users" and the "I.T." community. There was a surprising statistic that Aberdeen found which as 57% of business user do not see the value of business analytics. What shocked me was that... the business users should be defining and driving the value in the first place!

At times it can seem like two very different environments: the world of data and technology on one hand, and the world of marketing, sales, and supply chain on the other. I've been fortunate in my career at General Electric, Pfizer, and even now at The CementBloc, to be part of a "bridge" department that spans both worlds.

However, in these days of convergence, the more each of us can understand the language and motivations of our collaborators, the better our companies will be.

I would love to hear comments from readers on your perspective!

Long term consistency: harder than it looks

There has been lots of attention in the USA this past week spent on Derek Jeter of the Yankees getting his 3,000th hit. This is a measure of an expert artisan performing well at his draft for over 15 years, a model of consistency. Even as the game has evolved around him, and some competitors changed.

What does consistency mean in analytics? It means continually providing valuable consulting and insights, even as the promotional channels and the hardware platforms evolve. Even as the clients become more cost conscious than ever, and even as the marketplace for analytic services becomes more competitive. Consistency also means staying ahead of the curve on the prevailing technology trends,and the newest data sources and software tools.

Yet, just like a base hit to the opposite field, some principles of analytics endure: synthesize the right data, reduce the dimensionality, show the insights that are directly linked to business decisions and actions.

promotional planning on rainy days

Sometimes rainy days come unexpectedly, and one must rapidly determine how to respond. This weekend was a great lesson for me and thousands of others in Blacksburg Virgina. At the Virginia Tech graduation, the initial ceremony in the large football field was curtailed as lightning and rain approached. However, spirits remained high. Some wondered if the ceremony should have been moved indoors, meaning more safety but far fewer attendees. Anyway,the next day grads got diplomas, the sun stayed around for photos, and then the rain continued off and on. At a barbecue were wetness made it seemed lost, a group of us pressed on and served up a happy lunch repast in celebration.

In business, rainy days can mean a quarter or two of disappointing product sales, and businesses should keep two things in mind:

* dig deeply into your analytics dashboards as to what the root causes are. Which customer segments are performing better than others, and in which channels? Which promotions and media are more effective than others.

* plan for efficiency in promotion, and invest where the highest sales leverage is.

* use quality and six sigma frameworks to trim the inefficiencies. However, by all means, keep investing in what has been working.

Social Dashboards of Web Dashboards? Cut to the business

Website analytics has been an area I have been practicing for over 10 years now, back when this was an active software area of many startup companies (with colorful names like WebsideStory) outdueling each other for the most strking visualizations and most insightful path analyses. Nowadays, I have first found it surprising to see the rise in use of the free but spartan Google Analytics over the more mature and visually robust Webtrends and Omniture (now part of Adobe site catalyst). Those two have gobbled up many of the aforementioned startups.

Lately this month I have stumbled across new developments: software that take Google Analytics output and create new dashboards. One is Unilyzer software for dashboarding social media trends out of your Google Analytics output. The other are actual Google Analytics post-processing apps. I'll bet there are more that post-process in this way. I myself often take "G.A." output and visualize in my favorite dashboard prototyping tool, Spotfire. Tell me ones you know!

How to learn about the best choices for dashboard design, cost aside?
Much has been written over the past two decades about effective display of visual information in dashboards.
The work of Edward Tufte is foundational, and there are excellent texts like
Information Dashboard Design by Stephen Few, reviewed in UX Design.
One can even read about dashboards in other original areas for inspiration: see this
blogger on automobile dashboards

However, don't forget the main purpose of your dashboard should be to drive business decision making. Therefore be business focused. Create the dimensions that matter to your application domain and to your clients. All data miners say that data processing to get clean,insightful KPIs is up to 80 percent of the project. Do not push that off by merely re-displaying the raw feeds. Make sure you get right to the drivers of your business.

Interactive dashboards for RM programs

Want to track your relationship marketing programs periodically, and make the results insightful and impactful? Do not just count responses in some tired tables or bar charts.

Here are some tips:

* Have a monthly (or weekly) forecast of estimated responses, broken out by sub-segment and promotional source as well. Track results versus the forecast.

* Use a visual, interactive dashboard to show results, not only high level vs. forecast, but drill down by segment, or by promotional source. Try to use drill down to explain why your campaign is ahead of or behind forecast.

* Show insights, interesting trends, and outliers right on the dahboard, and hyperlink them. Focus your attention on highlighting the insights, not a methodical stack of reports.

Putting this all together, here is an example, implemented in the Spotfire thin-client web player environment. Click around, interact, try the bookmarked insight, and let me know your thoughts.