Advanced marketing & decision science.
Helping clients get more from Bayesian modelling, Marketing Mix Modelling, machine learning and AI.
Specialists in Bayesian modelling, causal inference, simulation and decision science.
contact@dsanalytics.co.uk
“Their team smoothly integrated Bayesian statistics, simplifying our understanding of LTV and turning a complex issue into actionable insights.
Their commitment to collaboration was evident, fostering a dynamic exchange of ideas. This, coupled with their technical acumen, ensured a streamlined project progression.”
-
Marketing Mix Modelling
Modern, fast, cost-effective. Bayesian models designed around the needs of decision-makers.
-
Retention and Lifetime Value
Find and keep your highest value customers.
-
Causal Modelling and Experiments
Understand the true incremental impact of your actions.
-
Business Simulations and Optimisation
Make better decisions by understanding the full impact of actions on KPIs.
Measuring customer lifetime value
[Bayesian modelling], [data visualisation], [python]Client: A US-based audiobook and podcast service.
Brief: To improve the accuracy of their customer lifetime value predictions in order to optimise marketing investments and product development.
Solution: Working closely with their analytics team, we developed a custom built python package. This includes models that accurately predict retention behaviour across their subscriber base and a scenario planning tool for simulating the impact of changes in marketing tactic, price and other variables.
Impact: Both finance and marketing are able to use the models to make better investment decisions. They have confidence and trust in the models and we were able to successfully handover future development to their in-house data science team.
Predicting the causal impact of retail outlet openings on sales
[Bayesian modelling], [causal analysis], [predictive modelling], [R]Client: A global furniture retailer.
Brief: Build models and tools that help them optimise their store opening strategy.
Solution: A python package that measures and predicts the causal effect of opening a new outlet on the existing store network. A web-based simulation tool for mapping the store network and visualising the impact of openings and closures.
Impact: Established trust in the performance of the models and allowed the in-house data science team to take over the maintenance of the models. Provide ongoing mentoring and support to their team.
Marketing mix modelling
[Bayesian modelling], [data visualisation], [R]Client: A UK-based fashion retailer.
Brief: To develop and deploy and modelling framework that measures the impact of marketing activity on their business on a weekly frequency and enables them to scenario plan and optimise.
Solution: We developed and deployed an R package and shiny application. This uses Bayesian marketing mix models to accurately measure the impact of marketing and other drivers of sales. The web dashboard allows the team to generate media plans and identify optimal budget allocations.
Impact: The business has trust in the models and we continue to work closely with them to develop and improve the functionality.
Simulating patient flows through a hospital
[Monte Carlo simulation], [python]In collaboration with Manchester University Hospital Trust, we built a tool that simulates the movement of patients through a hospital. This was used to design ward configurations to minimise in-hospital spread of the COVID during the pandemic.
We provide data science training and support for some of the world's most innovative companies.
Data Science Training & Mentoring
Our training courses are tailored to our clients' needs.
We cover practical analysis in R and python, as well statistics, Bayesian modelling and machine learning.
We provide ongoing support for data analytics teams, including data strategy, helping with the hiring process, helping transition from older technologies to R and python and data science mentoring.
“DS analytics continue to prove invaluable to our research activity. Our recent publication on rapid testing for COVID-19 was greatly improved by their data visualisation approaches and we continue our productive collaborations in the areas of acute genomics and combinatorial pooling. The company is responsive, efficient and highly engaged.”
— Professor Bill Newman. Professor of Translational Genomics. University of Manchester
“DS Analytics helped with one of our most high profile strategic projects that used complex bayesian analytics algorithms. We were very impressed with the technical depth and clarity that they brought into the project and it was an absolute pleasure working with them. Highly recommended.”
— Ayan Kar, Head of Analytics EMEA, Essence
“DS Analytics were immensely helpful getting our product team up and running with R and data science. Would definitely recommend them.”
— Magda Piatkowska, Head of Data Science Solutions at BBC News
“The R training ran for us by DS Analytics has been perfectly tailored to the needs of our team. They designed a bespoke curriculum in well laid out modules. It was clearly presented and pitched at the right level. Feedback so far from the team is extremely positive and we’ll be continuing to use them to take us further on our R journey.”
— Jane Christian, Business Science Director at MediaCom UK
