It’s no wonder that Deb Hartman and Sandra Kelly of 3M found that back in 2004 that marketing departments in the company were all doing their own DIY research using free or cheap tools. In fact, the marketers LOVED being able to do research on the penny. They didn’t have any money to use on more sophisticated software and really, they didn’t see the point of it much either. As they put it in an amusing slide “The Monkey was running amok at 3M”.

The problem was, with all this decentralized research going on by DIYers decisions were being made off bad or skewed research. Bad wording, poor sampling, and biased questions were just the start of the problems.

So in 2004, a group of 3Mers got together and decided to create an internal company to provide services to the 3M divisions. As with most 3M divisions they had to start with a business plan and get buy in — and their beginnings were modest, they started with two people simply programming peoples’ surveys and helping them pick the right questions. Between 2004 and 2008 they added 2 more people, reporting services and other services like conjoint and sample selection.

The model they used, because they are not a 3M profit center (this is all internal money) was to bill their internal clients as little as possible. Just enough to cover their own costs and the costs of their software. They shared their costs with us which was interesting:

Vision Critical Survey Software: $35,700
Sawtooth Software: $15,000
SPSS: $,5800
Contract Employees: $360,000
Total: ~$416,500

All I can say is wow. That’s quite a bit of value for very little at a company the size of 3M! Also, wow, they are spending too much on software — If you guys are reading this: you really should check out our dedicated-level accounts!

As a closing comment — they have saved the company an estimated $1.7 million by forming this internal company when compared to external vendors. That’s really impressive! Good job guys!

The Pareto principle also applies to survey research. When planning your survey project, plan on expending 20% of your resources on the survey execution (data collection) and 80% of your resources acting of the information gathered (follow-up actions).

Let me explain using an example:

I recently completed a survey project for the Taos Municipal School District (TMSD) in Taos, New Mexico. A year ago the school board formed a subcommittee with a charter to make recommendations for improving communication within the district.

The committee quickly decided they needed to do a survey in order to learn what problems existed and what solutions may work best. The need to do a survey project led them to me. My first task was to set expectations on two fronts:

1. The scope of communication within a school district is huge.
2. The survey project is not an end. Early in the design the focus needs to be on what is going to be done once the survey is complete.

After nearly a year with the “volunteer” committee meeting once a week, we had completed three surveys (one each to parents, staff and the community) and collected over 1,000 responses.

TMSD spent $12,500 and the volunteer time on the data collection and analysis. Does it seem reasonable that they might spend $50,000 and many many committee hours to implement solutions and improve communications? Let me help with some more information. The results showed that the problems and solutions were different for different schools within the district. It was clear communication was not going to improve unless changes were enacted; website upgrading, school improvement teams (high school, middle school and elementary schools), a district liaison to the community and a stronger relationship with the town’s newspaper were just some of the recommendations.

Will the information be used and changes enacted? There was a glimmer of hope right at the start: The study results were presented at a School Board meeting in the city council chambers. After seeing the results and hearing the recommendations, the board took action on one recommendation right there on the spot. The board voted on and passed a resolution to make “progress on improving communications” a standing agenda item for their monthly meetings. (In this case, the cycle time from results presentation to specific ACTION broke my old record!)

Do yourself and your customers a favor. Plan at the start to expend time and money on the implementation side of your next survey research project.