Advanced Analytics

Our in-house Advanced Analytics team is always on hand to support the design and analysis of any quantitative research project we conduct.

Overseen by Dr David Pearmain (whose PhD is in stated preference methods), their expertise will help you to discover the meaning in your data.

Advanced Analytics team areas of expertise

More about the techniques we use

SegmentationSegmentation analysis classifies a population into distinct groups of similar individuals according to survey responses.

We use several techniques to achieve this, including k-means, latent class analysis, normative methods, hierarchical segmentation, and trees.

To help you action your segmentation, we create a profile of each segment, and can usually append the segments to your database.

CHAIDCHAID for CHi squared Automatic Interaction Detection. It's 'tree technique' which is really effective at finding all the significant drivers of a particular outcome, for example, both how quickly a call is answered and writing a follow-up email may have a big impact on customer satisfaction.

The results allow you to target impovement on the drivers which will have the greatest effect.

Driver analysisKey Driver Analysis uses statistical techniques to ‘model’ or ‘explain’ causal relationships in data, for example, 'If we improve the staff attitude, how will that impact recommendation?' It is used to help prioritise areas for improvements which will have the greatest impact on a key variable.

The techniques we use include correlation, linear/ridge/logistic regression, Kruskal’s analysis of relative importance, and Shapley values.

Conjoint and Discrete Choice ModellingConjoint is a powerful tool for understanding how consumer choices are made as a function of the combined features (including brand) and the price of a product or service.

It can be used for a wide range of purposes including understanding the appeal for different package options, guiding new product development and portfolio management, and predicting whether customers would pay more if more features were added.

Simulators and toolsWe can turn a range of advanced statistical techniques into 'tools' that you can use whenever you need.

These include sophisticated simulators based on conjoint and Discrete Choice Model utility data which can help you to understand the market and feed into your decision-making about the marketing mix portfolio for a brand/product.

Switching matricesSwitching matrices show where customers move across a category when prices are changed.

They show sensitivity to price, brand loyalty and switching behaviour, and answer questions such as, 'Do customers stick with the same brand?', 'At what price point do they switch?', and 'To what do they switch?'

MaxDiffMaxDiff (maximum differentiation) is an effective tool for identifying the product claims and features that are most important to consumers. It offers the benefits of both ranking and rating scales because it determines the ‘distance’ between items as well as their rank order.

It's easy to administer and it really comes into its own when there is a large set of attributes (too complex for respondents to do) or if the attributes are very wordy.

TURFTURF (Total Unduplicated Reach and Frequency) is used to discover the combination of items which gain the broadest ‘reach’. It's rarely a case of picking the ‘best’ individually performing items from a set as these will often ‘overlap’.

This technique is ideal whenever there is a lot of choice, and is most commonly used to help manufacturers decide on flavours to launch. It also enables ‘Basket Analysis’.

Brand MappingBrand maps display a large amount of information in a simple format. They illustrate the complex relationship between variables such as brands, performance rating of these brands across a number of attributes, and market segments.
They tend to stimulate discussion and strategic thinking, and are often popular with senior management.

Willingness to payWillingness to Pay is a type of Discrete Choice Modelling which derives monetary values from attribute importance measures. Examples include the value of time savings in transport, of outage limitations on electricity networks, or of additional services in the gas network.

It is commonly used in public sector investment studies, including as part of a ‘Cost Benefit Analysis’ (CBA) to inform a business plan.

Price sensitivity meterThe Price Sensitivity Meter (also known as the van Westendorp Pricing Model) measures consumers’ perceptions of the value of a product or service.

It's a quick and affordable way to identify the price landscape although it doesn't accommodate competition, assumes respondents know the market, and does not directly measure likelihood to buy.

Gabor Granger and contingency valuationGabor Granger allows an optimum price-point to be calculated for a product or bundle of attributes based on the price level which would generate the greatest revenue potential.

Contingency Valuation is a variation on Gabor Granger. It produces a similar price curve, but expresses the average amount respondents are willing to pay in terms of an upper and lower boundary rather than a single value.

Data fusionData fusion or triangulation is the process of taking data from multiple sources and joining them together on some criteria so that analysis can draw on different sources.

This validates the insight that are gained but also increases the breadth of understanding.

Social media integrationCompanies can tap into the rich amount of social media data available by fusing it with primary research data and sales data to give a multi-dimensional view of the market.

Social media data can be used to understand changes in the market, changes in sales and can be used to drive sales and/or brand perception.

TrainingWe run courses in any analytical technique at any level and can design courses to fit your specific needs.

Bespoke SPSS training courses and 1-1 tuition are also available.

Meet the Advanced Analytics team

Dr David Pearmain20200723145842

Advanced Analytics

More about the techniques we use

Meet the Advanced Analytics team

Dr David Pearmain

Steve Whennell

Acsah Naripaty