StopPress
StopPress + Ipsos

Linkage analysis: data's hidden stories

Ipsos' Leo Cremonezl and Jonathan Dodd talk transforming your customer data into real financial and business benefits.

By Leo Cremonezl and Jonathan Dodd, Ipsos | October 17, 2017 | Sponsored content

In the age of big data there’s no shortage of stats and figures, and more often than not these confuse more than clarify. But that needn’t be the case, now that linkage analysis combines data sets to extract more tangible and actionable insights. 

Take advertising for example - every advertisement brings with it its own set of data and key performance indicators (KPIs) to judge it by, but what success means is different for every business. For data to prove itself useful, it must be transformed into real financial and business metrics.

To do this, linkage analysis combines different sources of data to uncover important relationships between variables. By combining survey data and client data, linkage analysis can demonstrate how an advertisement affects a brand in terms of measures such as brand recognition, consideration, customer satisfaction, business performance, ROI, etc. For instance, if a fast food chain wanted to know how their most recent ad campaign had affected the sale of cheeseburgers they might use linkage analysis to compare sales figures with brand recognition in their ads.

Statistical modelling can then be used to explore the relationships between tracker data and organisational performance KPIs, to develop an integrated picture of the drivers of business performance so that clients can focus on the areas that have the greatest ROI.

This kind of predictive analysis poses the question: if you increase X then what happens to Y? Or, returning to our fast food chain, if you increase investment in an advertisement by X thousand dollars then Y percent more people will recognise your burger on their walk down the street. These predictive models allow clients to learn the optimum moment where they can achieve the most gain for the least possible expense.

Do you want fries with that analysis?

So, back to the fast food, supposing you own a fast food restaurant and you have decided to do some linkage modelling to investigate how different aspects of your business work together, such as ad spend and brand recognition or sales figures.

Two important graphs help us to understand the relationship between KPIs of interest: the scatter plot and the quartile chart. Our example graphs illustrate a strong positive correlation between fast food brand recognition and investment in each advertisement. At a glance, it demonstrates that advertisements with higher investments tend to experience higher levels of brand recognition.

But does it change things?

We saw above that the correlation analysis revealed a connection between ad spend and brand recognition, but correlation does not always denote causality. This is where predictive modelling adds additional value to this analysis.

Predictive modelling analysis can quantify these relationships so that researchers can not only interpret if there is a relationship between factors X and Y, but how much. For instance, if the fast food chain spends $1,000 more on advertising then they can expect Y percent more recognition. Quantifying these relationships allows researchers to find the optimum moment between factors to provide effective results.

Creating a story behind the data which drives strategic investment decision-making

With the use of linkage modelling, it is possible to understand and quantify the relationships between different KPIs from different data sources. The insights inform marketers about where to invest their efforts to stand the best chance of delivering high performance.

Every data set has a hidden story, and rather than delivering raw data that provides few tangible insights, linkage modelling allows a narrative to be drawn from the data, interpreted to deliver tailored and actionable insights, revealing the stories behind the data rather than just a page of figures.

Actionable insights are the key to unlocking the potential of data and using it to drive performance. Linkage modelling is the ideal solution for extracting the story behind the data quickly and cost-effectively.

This story is part of a content partnership with Ipsos. 
 

This is a community discussion forum. Comment is free but please respect our rules:

  1. Don’t be abusive or use sweary type words
  2. Don’t break the law: libel, slander and defamatory comments are forbidden
  3. Don’t resort to name-calling, mean-spiritedness, or slagging off
  4. Don’t pretend to be someone else.

If we find you doing these things, your comments will be edited without recourse and you may be asked to go away and reconsider your actions.
We respect the right to free speech and anonymous comments. Don’t abuse the privilege.

features

TVNZ-NZ Marketing Awards 2017: Lion's journey to reignite the beer market

Faced with a market in free-fall, a decade of declining beer volumes and an iconic brand fast losing relevance, Lion was in need of creating ...

Results for
Topics
Jobs
About

StopPress provides essential industry news and intelligence, updated daily. And the digital newsletter delivers the latest news to your inbox twice a week — for free!

©2009–2015 Tangible Media. All rights reserved.
Use of this site constitutes acceptance of our Privacy policy.

Advertise

Contact Vernene Medcalf at +64 21 628 200 to advertise in StopPress.

View Media Kit