An Analytical Approach to Market Research

Published by Tara Piazza, EVP, Fulcrum Analytics on January 26th, 2017

While the marketing industry has been steadily shifting toward a data-driven new reality, the marriage of observed data and custom-collected survey data has been adopted more slowly than might be expected.  This is often due to a segregation of duties between research, analytics, and technology staff.  In the future, these disciplines are predicted to merge closer together.  Fulcrum’s capabilities in data science and big data management, along with its market research practice, offers a unique and powerful set of services that blend the three disciplines for comprehensive customer understanding.

The Market Research Mission Statement

Fulcrum’s mission is to combine, synthesize, and activate the power of customer level data for our clients.  Fulcrum’s expertise in big data and advanced statistical analysis further bolsters the predictive power of custom quantitative and qualitative market research.  Typically, our research results are combined with publicly available and/or client-proprietary data to yield a much richer analytical insight than any of these data sources can provide alone. Fulcrum’s data scientists then use statistical models to apply research analysis results into a prediction of future activity, or an algorithm for what-if scenario testing.

For clients who utilize Fulcrum primarily for our data science and big data functions, our market research group is often deployed to provide context for patterns observed in the statistical analysis of large data sets.  Our research functions as a dialogue with consumers to guide businesses on what to do next. Often, test and learn scenarios are then deployed to further inform business decisions.

Our DNA:  Cutting Edge Analytics and Market Research

Our market research function has a long history of combining cutting edge data collection practices with academically recognized best practices in an effort to further this mission.  Fulcrum Analytics was founded as a market research company in 1993 and had much success in the early years of the commercial Internet as Cyber Dialogue, Inc.  From the start, Fulcrum gathered customer insights using cutting edge market research data collection technology.  In 2000, Fulcrum formed its data science group to decipher the patterns of data using more sophisticated statistical processes than traditionally utilized in marketing research.

Today, the language of analytics is ubiquitous in marketing, though not all marketers or analytical consulting firms perform the same detailed level of data science as Fulcrum.  Fulcrum continues to push to the edge of new horizons by folding in the latest technology, programming languages, and data gathering concepts into our analytical service offerings, while also maintaining the rigor of long-studied best practices in marketing research to minimize error and maximize data accuracy.

Combining Big Data and Small Data

Within the market research industry, the concept of big data is just beginning to enter the lexicon and is still unproven as a widespread means of gathering accurate insights, often due to the expense and effort to harness disparate data sources.  Fulcrum has developed a highly secure lab environment for clients to use on a SaaS basis.  The Fulcrum Agile Analytics Lab integrates proprietary and public data sources to allow for the proving out of test use cases prior to making investments in big data infrastructure and software.  Our technical experts quickly and efficiently structure these labs and use case scenarios for clients to run by themselves, and to provide support for sandbox experimentation.

The marketing research function at Fulcrum further overlays onto the big data insights the consumer’s sentiments relating to the observed behavior.  Some call this the collection and analysis of “small data” to compliment the big data.  Observation of behavioral data tells us what has happened and allows us to form theories as to why it happened, but only the end customers themselves can provide the real context to decision making patterns.  For example consider the following:

Change is Good, but how to get Customers to Change?  Let’s consider a scenario where big data analysis has confirmed, through pulling together auto insurance claims data and NOAA weather data, the particular types of weather in particular geographies that result in the most expensive weather-related accident patterns.  The insurer would like to reduce these accidents, and decides to test a program where the insured customers might get better rates if they agree to not drive in the particularly dangerous weather for their location – possibly even tied to alerts which might be sent by the insurer to the insured based on NOAA forecasts.  But the insurer wants to know will customers be willing to give up their driving freedom in exchange for lower rates?  How much discount would be needed to impact the customer behavior?  How could the insurer get maximum compliance?  The answers to the questions are gathered through a custom market research study with customers prior to an expensive test market experiment.

Combining New Data and Historical Data

In some companies, market research and data science exist as separate divisions, where data science is utilized solely to forecast the future based on historical data, which may be sales data, site visit information, loyalty card activity, campaign response, or third party data as some examples. However, this leaves untapped the rich patterns that could yield detailed insights into consumer behavior and preferences when combined with market research data.   For example:

Assessing Effectiveness of Advertising Spend. An analytical approach to assessing advertising spend effectiveness could involve capturing all of the media buys, campaigns, messaging, audiences and other variables of the investments a company has made historically.  The resulting analysis could yield a media mix model and process for planning and spending in each market.  A market research data feed rounds out the intelligence identifying the most impactful language to deploy for messaging; determining the optimal imagery to elicit the desired brand associations and calls to action; and possibly providing guidance and feasibility metrics of strategic partnerships for bundled promotional products in keeping with the core messaging.

Fulcrum’s Market Research Fills the Gaps when Needed Data Does not Exist.  Some example of non-existent data are the appeal or ideal pricing of new products; understanding reasons for attrition; knowing customers’ purchase journey; profiles of a customer base to support new product development or variable content via segmentation; and competitive insights/share of wallet.

While the availability of free survey tools makes it easy to administer surveys, many hidden challenges remain to generating reliable insights.  Avoiding pitfalls requires a proper understanding of market research practices such as question wording, survey length, incentives, data collection methodology, and sampling.  Surveys executed without training and experience can result in gathering biased, misleading, or inconclusive information. Think of do-it-yourself surveys the same way you would think of home repairs or treatment of an illness – while there are many websites and items for sale to enable a DIY approach, when there is a lot at stake, they should be done by a professional.

For more information, contact us today.

BACK TO TOP