• Text Hover

Advanced

Analytics

Our analytics team led by Karsten Shaw provides expert advice and consultation

on the design/analysis of advanced analytical methodologies and allows us to provide

a seamless and integrated solution with our data collection services.

Our aim is to ensure that the statistical outputs we provide will be in plain English and in a commercially focussed way, ensuring maximum levels of insight are generated, not statistical jargon.

 

Range of Techniques

Please click on the below to view background information on these techniques and to understand the business questions they answer:

 

“We have a long standing relationship with Populus Data Solutions based on trust that they always deliver exceptional service levels. The addition of an in-house statistics team has strengthened this relationship and given us new ways to work together. The feedback from our client to our report and the simulator was extremely positive.”

 

Richard Barton
RED C Research, London

Driver modelling 

Questions it answers:

 

– What are the drivers of customer NPS/ Satisfaction/ complaint resolution?

– What drives customers to return to my brand?

– What are the drivers of employee engagement?

– Which areas should I prioritise to improve customer satisfaction?

– What would happen to satisfaction is one service aspect was improved?

– What are the most important features for my brand, but are underperforming?

 

 

Background:


Driver modelling uses varied range of techniques, from correlation to regression models or prediction trees (CHAID) in order to investigate the relationship between brand attributes and performance. This analysis is a powerful tool which helps brands describe, diagnose or predict their most important KPIs and is packaged in a simple easy-to-understand deliverable.

Segmentation analysis

Questions it answers:

 

– Which customers should I target to grow my market share?

– How do I ensure that relevant messages go to different groups of customers?

– How can I categorise my customers/ employees based on their attitudes?

– How do I identify customers which are most loyal or most at risk?

 

 

Background:

 

Segmentation analysis uses a varied range of techniques, from factor analysis to various forms of cluster analysis in order to divide the market of potential customers into groups, or segments. The segments created are composed of customers that respond similarly to marketing strategies and share similar traits such as attitudes, interests or location. This analysis is a great tool for understanding your customers and knowing how to best target them.

Conjoint analysis

Questions it answers:

 

– What is the optimal mix of my products or service portfolio?

– How will new products perform compared to existing ones?

– What are the gaps in my existing product/ service portfolio?

– What would the effect of changing the price to my products be?

– What is the willingness to pay for additional features?

– What product strategies should I use for different customer segments?

 

 

Background:

 

Conjoint analysis is a statistical technique used to determine how customers value different attributes that make up an individual product or service. The objective of conjoint analysis is to determine what combination of a limited number of attributes is most influential on customer’s likelihood to purchase the product. Conjoint is a great tool for testing current products against competitors, or against future own products.

MaxDiff

Questions it answers:

 

– What are the most appealing features of my product/service?

– How appealing would a new feature of my product/service be?

– How effective are different messages that I communicate?

– If I phrase the same benefit in different ways, which would be most effective?

– What are the most important priorities for improvement?

– Which attributes are a priority in an absolute, not just a relative sense?

 

 

Background:

 

MaxDiff, short for Maximum Difference Scaling, is an analysis based on choice, where customers engage in a trade-off exercise regarding different pairs of product attributes or features. This analysis allows us to determine how customers evaluate all possible pairs of attributes within the displayed set and choose the one that reflects the maximum preference or importance to them. The outcome of this analysis allows for an absolute ranking of importance, preference, relevance, etc. of attributes.

TURF

Questions it answers:

 

– What combination of product features will appeal to the broadest range of customers?

– What is the incremental customer reach if I add a particular product feature?

– Which combinations of messages will attract the widest possible audiences?

– What is the optimal combination of messages for the greatest customer/audience reach?

 

 

Background:

 

TURF analysis, short for Total Unduplicated Reach and Frequency, is a type of analysis used in order to find the best mixture of features or messages that will reach the widest possible audience or customers. TURF is a simple tool but can be very powerful and should be used to compliment other measures of gauging appeal in order to understand how attributes work together.

Media modelling and Econometrics

Questions it answers:

 

– How effective is my advertising campaign in driving KPIs?

– Do I have the right mix of channels in my campaign?

– What is the reach of different channels?

– What is the ROI of each of my channels?

– What is the incremental reach of the channels?

 

 

Background:

 

Media Effectiveness modelling is not a single technique, rather it refers to a range of approaches developed by Populus which aim to measure the impact of advertising and marketing activity. These approaches use aggregate information about media spend along with aggregate reads of KPIs to detect trends and be able to determine the impact of multiple channels on those KPIs. Other information can also be incorporated such as economic data and product level data such as pricing.

Perceptual mapping (Correspondence maps)

Questions it answers:

 

– What are the distinguishing features about my product and where are they placed in the market?

– What areas does my product underperform/ over perform in, compared to the market?

– Where do my competitors over perform/underperform?

– Are there any gaps in the market?

– What are the opportunities for my product in the market?

 

 

Background:

 

Correspondence Mapping is a data reduction technique, allowing simple interpretation of the market trends by using visual two-dimensional quadrant mapping. This analysis aims to identify and describe links or ‘correspondence’ between different statements and brands, which can then be mapped.

Pricing models

Questions it answers:

 

– What is the price elasticity for my current products/services?

– How much can I charge new products/services?

– At which point my products/services begin to look expensive but customers will still but them?

– At which point my products/services begin to look to expensive for my customers to buy them?

 

 

Background:

 

Pricing models comprise different techniques suitable for new or existing products or services. For new products, pricing scenarios can be pre-defined by the client or acquired using a Conjoint design. However, simpler techniques such as Gabor Granger and Van Westendorp are also used to understand fundamental responses to different prices of products and services.  We align this to likelihood of purchase of customers and compare with product demand and revenue estimates in order to determine the optimum price of product delivery, whilst ensuring maximum revenue.  

Data fusion

Questions it answers:

 

– How does customer satisfaction/recommendation link to financial outcomes?

– How does employee tenure influence our current KPIs?

– Is there a different in customer satisfaction after a product/service had been changed?

– Is there a relationship between staff training and customer outcomes?

 

 

Background:

 

Data fusion is not a statistical technique, rather it refers to a range of different ways in which we can align internal company data with survey metrics, in order to provide a more comprehensive picture of the results. Internal data sources that we can work with include financial data, transactional data, HR data, training data.  It could also include things such as dates for store refurbishment or service changes.  These are then fused with information from survey data.