(This was pulled from a Client email I just sent where we were discussing if linear regression was the best analysis technique for a banking brand’s awareness research. Use the comments area below to tell me if you agree.)

I wrote…

No matter what analysis you do, it’s all just a tool. There are very few statistics that are worth anything outside of basic human judgment and insight. So long as you follow the rules and don’t manipulate the data or the people reading it, you can’t really go wrong.

I try to design data that will allow a group’s wisdom to come to the surface, spark an educated debate, or deliver confidence to move in a direction that indicates progress. Sometimes I’ll take a strong position just to try and get people who know more than me to debate the conclusion.

This is where things happen. It’s the apathy and insecurity that kills me. There is too much mountainous reporting that no one reads other than the author.

I’ve long ago let go of “the book of truth.” Maybe it’s a lazy way out and I really just need to go to more conferences and read more of the trade journals.

What do you think?

Their goal was to identify spending patterns among residents versus tourists and segment this into categories that defined the key drivers of tax revenue from non-residents.  The public officials needed to accurately attribute tax revenue to different spending pools to evaluate budgeting guidelines for state-sponsored programs.

The challenge was how to measure when so many variables were in play.

By using the methods outlined in the two previous articles that make up this market research consultants’ secrets series, we were able to cross reference a number of variables to link those activities which generated incremental in-state spending, identify where state promotional dollars were best allocated, and advise on future spending that would generated the greatest return for state residence in incremental tax revenue.

The question of how to measure marketing performance may be daunting on the surface but anytime you can measure the delivery and result variables over time you will have the basis for a predictive model that can be used to maximize future performance.

Measure Marketing Case Study #1: Cell Phone Manufacturer

A major cell phone manufacturer wanted to understand the dynamics around sound quality for a new pre-market cell phone. Portland Marketing Analytics organized a control/test beta trial among a representative consumer base to look at usage patterns and call quality.

As a result, particular industrial sites were identified as containing materials that interrupted the quality of service (not the phone’s architecture) thus clearing the product for general market release.

Measure Marketing Case Study #2: Credit Card Provider

A credit card provider focused on consumer affinity cards needed to streamline it’s call center reporting.   At any given time the business had 2,400 customer service reps across 200+ teams and 7 call centers taking calls from US consumers.  The organization used this delivery and results positioning to understand the performance of their customer service function.

This laid the foundation for rep scorecard measurement and exception based reporting that allowed management to improve performance and streamline training.

Measure Marketing Case Study #3: Marketing Agency

An agency focused on the delivery of TV, Radio, and Print advertising organized their marketing execution around these delivery and results measures and was able to accurately project the performance of a marketing program while the marketing was still in the planning stages.

This allowed modifications to be made to the overall design that would maximize marketing performance.

…

Portland Marketing Analytics is a team of market research consultants who specialize in creating custom measurement solutions that evaluate and predict marketing outcomes, product trial performance, and consumer behavior. Read below for some more information about how we can help you. Or use the comments box at the bottom of this page.

Your goal once you have collected your “delivery” and “results” data is to organize that data in such a way as to make it actionable.  (Note, this method also works well if you’re working with data that was originally collected for another purpose but is being re-purposed for a new need.)

Just organize the variables into the two groupings, “Delivery” and “Results.”

Nine times out of ten, your delivery data will be something you have control over.  By “control” I mean those aspects of your business that you can change pretty much at will.  When thinking about how to measure marketing results you need to keep in mind the things that define performance.  Last week we gave four examples of delivery data:

• Staffing
• Packaging
• Marketing
• Price

As a business manager, you or someone in your company has control over these four items (and the hundreds more that define your business operations).  For example, you can choose how heavily to staff the check-out counter, you decide the volume packaging segments, you know what type of TV buys or PPC online advertising you’ve bought, or when a discount is offered.

It would be fair to expect that each of these items will have some impact on your results measures.  For example, you’d expect staffing price to impact sales volume.

Here’s the Market Research Consultants’ Trick

The trick is to test these relationships (using the right statistical methods) to see at what levels in a delivery variable are associated with a meaningful change in a results variable.

For example, let’s say you measure a 10% increase in sales volume when you dropped the price by 10% in January.  In fact, you see roughly the same change in sales every time you drop the price by 10%.  Moving forward, you could therefore increase sales by 10% anytime you wanted to… just drop the price by 10%.  This finding is therefore “actionable.”

This is an obvious example.  It’s common for sales volume to go up when price goes down.  And for the most part, people don’t tend to consider the obvious as “insightful”.  However, there does exist a multitude of this same type of example in your business that is not so obvious.  These might include things such as:

• The relationship between packaging and seasonal sales fluctuation.
• The color of your “add to cart” button and how many people opt-out of the shopping cart.
• The placement of your 800# on a flier and the related call volume to your call center.

The goal is to hunt around your data to find these relationships.  And then to deliver them to the people who need to know in a way that they can then make positive business decisions.  This is the type of work we do at Portland Marketing Analytics every day.

Next week we’ll provide some case studies describing how other agencies and brands have used this process to make insights actionable.

The last couple of years, the field of business analytics has been ultra-focused on “consumer insights”; the art of understanding consumers at a level which reveals the underlying motivations behind purchasing decisions.

A lot of products and services have come to market and report to deliver “insights”.  However, even with the best tools now in hand, the process of applying these insights to real business problems is often no easier.

A Market Research Consultant’s Secret

How do market research consultants make insights actionable? Marketing professionals spend so much time, money, and effort to create a data flow, but so often don’t know what to do with it. And, in many cases, if you’re asking these questions once the data has been gathered and reported, than you may be too late.

This article is the first in a series of three that will explore how to measure marketing results from the angle of making facts and insights actionable.

The secret to making insights actionable is in how your organize the data collection process.

Often the task at hand is really, “how to measure strategy” and strategy starts with the deconstruction of things. All business data can be grouped into two categories that together define any business strategy.

1. Results – or that which we’re trying to impact or change (i.e., call volume, sales, coupon redemption, loyalty, revenue, widgets produced, etc.)
2. Delivery – or that which impacts the things we want to change (i.e., staffing, packaging, marketing, price, etc.)

So we have information that is “results” and information that is “delivery” by nature. If you were a statistician you’d call these dependent and independent variables. Another way to look at these are as cause (delivery) and affect (results).

When your information is organized in these two categories you’ll find that you’ve laid the foundation for converting data into actionable insights. Next week I’ll write about this process further and suggest some methods for organizing and reporting on the resulting data.

I’ll wrap up this mini-series with a set of case studies on how how others have tackled the question of how to measure marketing results and made insights actionable.

What do you think?  Do you organize your data with the end in mind?  Use the comment area below to tell us your technique or ask a question.

Are You Looking For a Market Research Consultant?

Portland Marketing Analytics is here to help. We’re marketing research consultants who talk to marketing professionals like yourself every day to solve measurement related challenges.