Advanced Analytics /

Below you can find out about our analysis

NTF has a very strong commitment to innovation. NTF’s innovation includes: algorithmic development and application; embedded analytics (development of robust analytic algorithms embedded in devices performing real time applications); benefits evaluation frameworks and methodologies (measurement, modelling and realisation of most impactful benefits), simulation and latent class to understand drivers of behaviour and attitude.

Prior to NTF’s involvement, many clients had been calculating relative importance using Shapley regression. While this approach is consistent with industry practice, it has a known weakness in terms of its lack of discrimination. Over the years NTF has developed a suite of different algorithms for the measurement of relative importance. Through combining different measures, our clients are better able to isolate aspects of customer service which have the greatest impact on customer satisfaction, thereby enabling a focus of managerial attention and investment on the factors which will lead to customer KPIs being met.

Since its inception in 1996, NTF has delivered customer insights based on sophisticated analytic modelling together with qualitative and quantitative research. It is highly innovative in its development and use of Analytics.

We have built multivariate predictive and classification models to address a wide range of issues across the Retail, Financial Services, Utilities and other sectors.

Techniques have included Random Forests, Quantile Regression Forests, Conditional Inference Trees, Conditional Inference Forests, Recursive Partition Trees, Generalised Linear Models (especially binary and multinomial logit) as well as k-means clustering, hierarchical clustering and Partitioning Around Medoids using Euclidean, Mahanoblis and Gower distance measures. We also have experience with Principal Component and Factor Analysis.

These techniques have variously been used for example to classify or segment customers, predict product take up, estimate Willingness to Pay and identify satisfaction drivers

For major service industry and both State and Federal Government clients, including Queensland Health, Department of Human Services, Commonwealth bank, Telstra, Australia Post and Cisco Systems.

Additionally, NTF has attracted global FMCG clients (e.g Coca Cola, Kraft, Sanitarium) who utilise NTF’s advanced retail analytics and have benchmarked our outputs favourably against global best practice: “The NTF brand price models are the best I’ve seen anywhere in the Coca Cola system, anywhere in the world”(Mary Minnick, Former Global President Marketing, Strategy & Innovation, The Coca Cola Corporation).

We have conducted modelling of the Shanghai financial and real estate market for Singapore’s GIC that resulted in a model that predicted within 3% accuracy for 5 years and is still being used as a training tool.

We assist our clients with three key strategic objectives:

  • Increasing customer satisfaction and advocacy: NTF creates driver tree models to understand which aspects and parts of the customer interaction are most important in determining satisfaction.
  • Accelerating the simplification of customer processes to improve productivity and reducing costs by modelling cost-to-serve and other operational data, so that we can simulate scenarios and optimise outcomes in terms of cost minimisation as well as customer satisfaction.
  • Assisting to improve digital performance amongst business customers (particularly SMEs).
  • Media spend optimisation. NTF has experience working with clients to optimise marketing expenditures by media channel, product area and geography.