We talk a lot about the virtues of data, but there is one important caveat – you have to have good data.  Sounds easy, right?

While I’m not going to write a dissertation here on how to set up databases or program data collection forms, I do hope to provide some high level thoughts on the process of data collection (not just the tools):

• Determine what data needs to be collected. Start with the end in mind – what do you hope to gain from the data?  Also consider what is possible – what is feasible to collect?
• Create a system. This one of the biggest areas I have personally seen gone astray.  You can have all the databases, software, and forms, but if the data doesn’t get in there right, or consistently, the whole database breaks down.  Develop a simple process that everyone can follow.  Then enforce it.  Where possible, make it automated or at least integrate with other processes.
• Share success. It can be hard for people collecting, entering, and sorting the data to understand how they are helping the big picture.  Keep them informed with how their efforts have helped and the benefits gained.
• Review and update regularly. Something not working?  Need additional/different information?  Check your data and processes regularly to ensure you get the most out of your data.

What else has been helpful in getting the right data?

Enter Pivot.

As the Economist recently described in a special report, the amount of data we are creating is immense.  So much so that we can’t even store it all, let alone make sense of it.  Tools like Pivot, and many others, are designed to help us create true connections and insight.

I walked over to the post office across the street to mail it, and was surprised to see a table there that was manned by a temporary U.S. Census worker.  Above the table was a sign that said, “Census Questions Answered Here”.  I can never resist a census worker, so I wandered over and asked her what type of questions she was fielding.

None yet, I was told, because the census forms had not yet been delivered.  They were being proactive, I guess, so I won’t second-guess that.  So I asked what types of questions they were expecting.  Atop the basic questions about content and how to answer certain questions, the friendly census worker showed me her various language forms, 60 or more, with instructions in languages ranging from Somali to Polish to Creole, and said that language was a likely issue.  So she was prepared for Problem #1, which was making sure that people could understand the forms.

Problem #2 was access.  This was a small town with significant “informal housing” by an immigrant workforce, so she mentioned that many residents live in informal housing such as garages, spare rooms, and other homes without formal addresses.  Making sure that all of these people receive census forms was a potential problem, so much so that they were actually hand-delivering the forms rather than mailing them.  So Problem #2 was making sure that everyone had access to a form, which is another reason to have a physical presence at the post office.

And finally, she mentioned the big picture.  Why are people receiving these forms?  Why should they fill them out?  The census worker was ready and prepared to talk to people about all the reasons to fill out a census form.

Of course, they were doing the same thing that we at Corona do, but on a larger scale.  Anyone can throw questions up on survey monkey or zoomerang and generate numbers.  The key, though, and the hard part, is to generate accurate numbers and relevant numbers that truly answer your questions.   On surveys, you do this by making sure that you’re reaching a representative portion of the population, by making sure they understand what they’re being asked, and by maximizing their likelihood of responding.  On a census, you do this by making sure that you reach everybody, by making sure they understand what they’re being asked, and by maximizing their likelihood of responding.  And you do this via good distribution and by explaining the value of what you’re doing.

Kudoes to the people at the Census Bureau for their hard work in making sure that the government – and Corona Insights – has good census data to analyze for our clients.

What do we do with Census data once it is released?  Check out our data analytics practice to find out some of the ways we utilize census data.

Baseball and statistics are inseparable.  There seems to be a stat for every aspect of the game from the classic batting average, ERA, and strikeouts, to the more complex “sabermetrics” BABIP, BsR, and EqA.  Compiling the voluminous data to create these statistics is a monumental task, but making use of the numbers is a greater challenge.

Bloomberg, known for their financial analysis, is hoping to use their expertise to help major league baseball teams.  Bloomberg is focusing on determining trends in order to predict future performance.  Major league teams already compile much of the same information, but the advantage of the Bloomberg system is the speed at which teams can access the data and having one repository for all stats.  The data is not limited for distribution to major league teams; Bloomberg also has a similar product available to fantasy baseball players.

One more way analytics can provide a competitive advantage.  In this case, literally.  The three of us just hope our team, the St. Louis  Cardinals, are making full use of it.

Some of the trends we noted last year – economic pressures, instantaneous research – will continue and become even more pronounced. Here we offer some additional thought specifically on market research and analytics.

• As companies seek answers to their important questions, they will first turn to the data they already have. Using data analytics to squeeze every last drop out of their ever growing data will be the most cost effective way for many to get their answers. (see a related Business Week article here)
• Related to the economy and overall cost cutting/justifying, the need to have demonstrable results will be even more important. In research, this will mean producing insights that drive profitable action. For broader strategy, this will mean setting (or keeping) a course that drives the organization forward.
• While we noted it last year, the need for instant feedback is growing even stronger (one sign of this is the major search engines indexing social media sites like Twitter). While the need has always been there, newer technologies continue to make getting it even easier. Mining online conversations and any “buzz” will become commonplace for all companies.
• As companies and organizations seek ever better information to inform their decisions, surveys, especially using online panels, are becoming more and more popular for their speed and cost effectiveness. However, as we continue to ask people to take an ever greater number of surveys, the cost of straining survey takers is becoming greater – greater burnout, lower data quality, and professional respondents, to name a few. (see our other posts on the topic here and here)
• One reason for the popularity of monitoring social media and online surveys is their speed. Real-time feedback through quick measurements is crucial for altering strategy and achieving demonstrable results.
• All the data in the world is useless unless it’s properly conveyed. From the quick overview of key metrics using data “dashboards” to telling the deeper story, clients and end users will demand to get the most out of their research. Deciding how to convey findings, tell the story, and do so efficiently will be crucial. (See a related article on Quirks about how to get those stories from respondents…Article ID 20091206 free registration required)
• Similarly, online communities will continue to gain momentum. These communities build off the common experience of online networks (i.e. Facebook) and allow for richer information to be gathered over a longer period of time than a one-time focus group or survey can provide.
• Finally, will location based services such as Foursquare and Gowalla create new opportunities for research? Combined with mobile research, the ability for short surveys, instant feedback, all in a relevant environment may provide one more tool in the researcher’s toolbox.

These are just a few key ways that new tools and the changing nature of respondents are impacting how you find your answers.  Do you see other changes taking shape?

Tomorrow, we’ll have part 2 of our trends post on strategy and the social sector.

But for all the fun we have with numbers (not a typo), we always enjoy hearing the many other ways numbers impact our lives. What’s Benford’s Law? Do you have an Erdos number?

Check out the recent Radiolab episode, Numbers. Audio excerpt from the episode below.

Every result, process, and assumption of the study was scrutinized.  Each stakeholder group in the City reviewed the report with their own lens, using the data to test their own assumptions and often to bolster their own positions..  During this study – and every other study we do at Corona – maintaining a neutral position is extremely important.  Any bias, real or perceived, would ruin the integrity of the study and the integrity of Corona.  During the entire research process, we had to ensure we kept our neutral position, that our methodology was rock solid and, just as important, that we could explain it in a way that would understood by any inquiring group.

Overall, the results were well-received and all interested parties welcomed having the right answers and insights that could help move the review process forward.

A controversial issue puts pressure on decision makers and they need the best information to make decisions.  By staying neutral and maintaining rigorous methodologies, Corona provides this information.

My initial thought was that Starbucks would close the 600 lowest performing stores. Easy. But how would “lowest performing” be defined? And why 600 stores? I had a discussion, fittingly over a cup of coffee, with fellow analyst, Dave, about the factors we might look at: total customers, revenue (and profit) per customer, and some other typical business measures.

Our conversation then meandered to more complicated factors, such as the amount of pedestrian traffic around the stores; proximity to other Starbucks; whether closing one store would cause another location to become too crowded; ease of entry and exit (and whether there is a drive through); and rent/lease terms in various locations, among others. I quickly decided my initial thought of just looking at the numbers ignored the many interdependent variables that must be taken into account and the change in consumer behavior that would occur if a particular location were to close.

Their final decision about which stores should close, and how many will close, must be based on several of these variables weighted by their order of importance. Simply identifying low performing stores in a vacuum ignores the interplay between the features of each store and its environment as well as the complex interactions between locations, especially given the close proximity of many Starbucks locations (there are no fewer than 10 Starbucks locations within a mile of Corona’s office). As a side note, these factors would probably be the same variables used in determining whether and where to open a new store.

While I don’t think I’ll do this analysis in my free time (unless Starbucks would like to hire us to do so), it would be one complex – and yes, fun – optimization model to build. Then I would be sure to win that office pool.