In 2009, YCG developed consumer segmentation for EMI Music. EMI now assesses opportunities by the size of the segment and the value that would be realized by moving those individuals forward in the life cycle, from awareness through to acquisition. In addition, the segmentation helps EMI communicate with its consumers by providing insights into artist, product and channel preferences.

In a 2012 assessment of EMI’s segmentation, Millward Brown concluded: “EMI has a good, strong segmentation that is effectively used throughout the business and is underpinned by strong analytics, a large amount of data and exceptional knowledge of the industry.”

YCG replicated the consumer segmentation strategy for Harper Collins Publishing in 2011.

In 2011, Dr. Yuhas received a grant from Johns Hopkins University as the principal investigator on a comparative effectiveness study. The study focused on Medicare beneficiaries diagnosed with Amyotrophic Lateral Sclerosis (ALS), also known as Lou Gehrig's disease.

The objective was to analyze the diagnostic and treatment information on Medicare claims to identify potential off-label impacts of drugs on survival rates. In the first phase, he identified candidate drugs based on statistical associations between their presence and longer survival times. In the second phase, he quantified the impact and computed the significance of these associations. Several classes of drugs were found to be associated with longer survival times among ALS patients.

In 2013, Dr. Yuhas received a second grant to continue this research.

Eli Il Yong Lee, executive director of Center for Civic Policy, New Mexico, shares his experiences with YCG:

We were intrigued about modeling but had never done it before and frankly, were a little terrified about the process. We also wanted to create a model that hadn’t been done before – a model for socially responsible voters, not a model for just one issue or characteristic.

Ben and Robin were exceptional. They put real thought into how to go about developing a socially responsible model and walked us through the process with great patience and clarity.

The results speak for themselves – our targeting was precise and effective. We saved tens of thousands of dollars because our phoners and doorknockers got yeses at a rate three or four times greater than with previous targeting methods.

Similarly, our radio and direct mail instantly became more cost-effective – for the same amount of money, we could now target many more people because we knew from first-hand experience that almost everyone we were reaching was in our target audience.

The modeling not only paid for itself within two months, but it is now saving us money each month in voter contact.

YCG worked with Planned Parenthood of Greater Iowa (PPGI) to improve the effectiveness of their outreach programs.

By leveraging existing polling data, they were able to develop models in a cost effective means. PPGI used these models to target their canvass operation, phone banking and house parties.

Chief public affairs officer Julie Stauch: “Volunteers reported that they had significantly fewer negative calls and in many instances, people were very friendly and receptive to our message. We also had significantly stronger supportive responses. As a result, we were able to add over 500 people to our contact list, in addition to identifying over 1,300 voters for our candidates.”

During the economic crisis, Minnesota voters overwhelmingly voted for a constitutional amendment to increase their taxes to invest in clean water, habitat conservation, parks, trails and the arts. It was the largest state-level conservation initiative in U.S. history, increasing the investment in clean water, natural areas, cultural legacy, and parks and trails by about $290 million a year for 25 years.

Working with Robin Pressman, YCG developed models to identify amendment supporters as well as likely activists and donors for Conservation Minnesota. Conservation Minnesota used these models to target key swing audiences with direct communication. The campaign also used the models to identify supporters who were more likely to skip the ballot question. Other models predicted which constituents would open email communications, allowing Conservation Minnesota to expand its email at less than half the previous costs.

“Key targeting of voters and message development was critical to the success of this initiative,” said Conservation Minnesota director Paul Austin. “The microtargeting developed for the campaign will pay dividends in years to come as constituent groups continue to use the data to build their memberships, target constituents on key legislative issues, and focus legislators on the issues their constituents care about.”

YCG has extensive experience working with financial services companies. Using data-driven solutions, we have helped our clients:

• Lower the cost of account acquisition by designing products with specific appeals and by better targeting solicitations.
• Minimize credit losses by using predicted risk to manage lines of credit, fee assessment and collection strategies.
• Increase cross-sales and customer retention by predicting preferences based on transactional data.

YCG combines polling data with enhanced voter files to build models that predict voter behavior. Once built, these models can help you choose individuals who are likely to have the behavior in which you are interested. These models help campaigns find supporters beyond traditional constituencies and identify softness within the base.

While polling often aims to find messages that have wide appeal to a broad class of voters, YCG aims to find specific messages that will appeal to specific individuals so that campaigns can better focus and prioritize voter communications.

For a political campaign, traditional targeting identifies individuals by party registration or precincts by their historical vote. For example, they would target those precincts where over 65% had voted for their candidate’s party in the past.  In Virginia, there is no party registration and in 2005, there were not enough 65+% precincts. To address this unique challenge, YCG developed models to predict the probability of support and turnout for each individual on the Virginia voter file. Individuals could be targeted anywhere in the state with greater precision.

In 2008, YCG predicted the early vote for the Obama campaign using models that predicted the individual probabilities of support for each voter. Earlier in the same cycle, YCG help the Clinton campaign target voter contacts during the Iowa caucus and led the boiler room team predicting the final vote based on sampled returns from targeted precincts.