Our Services

Janus Analytics provides a range of bespoke customer analytics solutions that allow you to implement sophisticated marketing science that improves customer relationships.

Analytics Strategy

The analytics strategy maps out the process that the business should take to achieve their business intelligence goals.

Janus Analytics adopts the industry standard practice for analytics development, known as the Cross-Industry Process for Data Mining (CRISP-DM). This process provides a framework to solving problems unique to each of our clients. Our analytics strategy aims to answer the following questions

  • What does success look like for your business?
  • What data needs to be collected?
  • Why does it need to be collected?
  • How is the data going to be used?

In addition to this methodology, Janus Analytics overlays industry experience and domain expertise to the analytics development that is specific to the industry in which each client operates.

Primary Research

The Infrastructure - CoreData has the infrastructure to complete end to end research operations from sample sourcing, recruitment management, data collection, analysis, reporting and implementation consultancy.

The process - The research design includes critical components that must be sequentially addressed to ensure we understand the state of play, develop a market map and generate robust novel primary research data to drive a comprehensive segmentation solution. They include stakeholder consultation, qualitative and quantitative research methods.

Qualitative research - Building on the perspective gained from stakeholder consultation and systems analysis we can then conduct exploratory qualitative research on the market designed to uncover in-depth insights critical to developing a psychographic needs-based model. This process allows us to go beyond the superficial insights and to truly understand how customers see the world, assess value, consume information and make decisions.

This important formative stage of research then informs the architecture of subsequent investigations and ensures we keep grounded in the reality of consumer perceptions and behaviour. This allows us to develop additional novel hypothesis and establish critical areas for robust testing in the subsequent quantitative research phase.

Quantitative research - Having established these critical components and developed some a prior hypothesis about the model, we then need to test this and incorporate ad-hoc analysis in a more robust way. This is achieved with quantitative research that allows statistical analysis on sufficiently sized and appropriate samples so that findings can be confidently inferred and a solid business case built. This also drives accurate market sizing work and predictive model development.

We also link our more complex ‘hard-to-capture’ psychographic and behavioural dimensions to available first or third party data proxies for system implementation. Based on this empirical evidence a final model solution is developed in close consultation with our clients

Database Analysis

  • BAU operational data
  • EDM and website interactions
  • CRM data

Qualitative Research

  • Focus Groups
  • Interviews
  • Ethnographic Research

Quantitative Research

  • Online survey
  • Website polling
  • Third party product holding
  • Behavioural Data

Predictive Analytics

Predictive models allow the marketing department to focus efforts on marketing to customers most likely to complete a certain action. Supervised machine learning techniques are utilised to predict key outcomes.

People Problems not Data Problems - For all of the promise of machine learning and big data, machines still lack empathy. Attempting to solve people problems with an analytics solution is fraught with danger. We ensure our models have a strong overlay of practical business operation considerations and common sense!

Want to know who to talk to first and how? - Predictive modelling essentially provides you with lead scoring for all your customer for cross sell amenability to various offers, relationship escalation, defection risk, CLV and goal outcome achievement. After selecting your best leads (or most in need), the segmentation model then overlays across this to provide you with actionable intelligence on how best to approach , frame and communicate with them one a one to one basis.

The Training Set is Key - Many issues have come from improper use of machine learning or predictive analytics. These errors have been caused by biased training sets (i.e. junk in, junk out). Designing an unbiased training set is vital in any analytics solution. Our experience ensures you avoid the pitfalls.

Database Implementation - Through finding correlations between regularly collected variables and psychographic information collected through primary research, Janus Analytics is able to implement the segmentation system on to the client database. We use multiple data sources in our modelling to ensure a complete picture of your clients, using both primary research, database variables and secondary data to truly paint a picture of your entire client base.

Supervised Learning

Unsupervised Learning

Reinforcement Learning

Monte-Carlo Simulation

Psychographic Segmentation

We get past the “average consumer” to deeply understand the different motivations and emotional needs that make your customers unique. Segmentation is not voodoo analytics. It is simply the mathematical process of dividing people into groups that are relevantly different, and then profiling the groups. The key is keeping focus on how the model will be used.

Meet the tribes - Segments are the tribes that live within your customer community and they all have their own important differences in perspective and behaviour. These idiosyncrasies need to be understood intimately, not just who they are and how they behave but how they see the world, how they think, what motivates them, what they truly value and what kind of communications and relationships they prefer.

In order to answer these critical research questions a model must pull together many different types of variables – socio-demographic, psychographic, geographic and behavioural.

Boiling it down - A good segmentation solution builds on what is essentially very complex interactions of many factors until it can boil these down into the essential dimensional driving forces. Conceptually these are then highly comprehensible by everyone in the organisation and maximize utility in implementation.

The end game is the ability to drive clear prescriptions on how to understand these groups, best approach product and service offerings, marketing, communications and relationship management for each of the unique segments within the customer base.

An effective segmentation model ultimately needs to transcend its fundamental algorithmic mechanics and actually come to life.

Database Enhancement

Database Enhancement is an optional add on to our analytics solution, Janus. Usually Database Enhancement is employed after data collection issues have been identified from systems analysis. Database Enhancement uses a range of techniques in order to quickly fill the gaps in a database, as well as implement new collection techniques moving forward. We work with your Information systems team to collect the required data as necessary.

Improved Data Modelling

Most businesses have the means of collecting the data they need. We get your data to the right people in order to make better business decisions.

Leverage Our Data

Our propriety database, 'Athena', contains more than 100 million data points on more than 100 thousand Australians. We can utilise our data to contextualise your data.

Third Party Databases

Sometimes you need data that someone else has. Janus has a number of third party data partners that can help get the data you need.

Digital Breadcrumbs

Our team utilises advanged data collection techniques to identify the digital footprint of individuals on the web and find data points about your customers.

Implementation & Maintenance

Implementation and Maintenance are the final stages of our Janus solution. It is our belief that these stages are the most valuable, as they truly bring the research to life, allowing you to operationalise the insights. It is vital that any segmentation system and predictive modelling system are used by the right people, at the right time, to get the full utility from the models.

We do not believe that any analytics system should be a “set and forget” solution. We believe in constant evolution of your analytics system just as your business constantly evolves. We offer an extended hours help centre and an account manager that you can access during business hours. Our team of data scientists and software developers are committed to creating the best experience possible, dealing with your data issues so you don’t have to.

We are constantly looking for new ways to improve our service, and issuing new features and functions to the front and back end of our solutions.