Case Study

Maru Syngro: Predicting future revenues based on customer satisfaction data

Posted on August 29, 2017

Company background

Maru/Syngro specialises in customer experience technology, collecting customer feedback and translating this data into insight and action for businesses. Maru/Syngro, founded in 2004 by CEO Keith Schorah and based in Livingston, ensures its clients are able to act on customer insight from more than 80 countries, and in 30 different languages.

Maru/Syngro’s core product, Syngro Eye, responds to these data-led insights, thereby driving clients’ profitable actions and operational performance.

Project detail

Maru/Syngro set out to develop a new suite of forecasting tools to integrate within its existing product offering. It wanted to develop an advanced predictive data analytics capability which could identify as and when its clients’ customers were likely to spend both less and more. The idea being to provide clients with further detail and root cause analysis of customer behaviour.

Maru/Syngro sought the support of The Data Lab to deliver the project. As a result, the project was collaboratively funded; The Data Lab contributing £54,127 and Maru/Syngro contributing £47,580.

The Data Lab facilitated a partnership with the University of Strathclyde, whose expertise and facilities were used to find the best method to deliver the project. Notably, the project was unusual as it worked across university departments, with collaboration between the university’s Management Science, and Computer and Information Science, faculties.

The project results, recorded in a report by the University of Strathclyde, are being developed and integrated into Syngro Eye, supported by its current data management and visualisation capabilities.

The solution

To bring the project to life, Maru/Syngro needed to overlay the existing customer satisfaction data, namely customer satisfaction feedback, share of wallet, net promoter score and textual sentiment scores, with further financial and operational data. This would provide greater context for decision making and ensure the ability to forecast future financial KPIs.

The project involved the testing of the new technologies and functionality by collecting and analysing data from one of Maru/Syngro’s largest clients.

Technical Detail

The cross-department collaboration within the Strathclyde University Maru/Syngro's project team ensured a range of expertise was relied on to interrogate the best method – i.e. using test machine learning models, univariate time series models or multivariate time series models. This was an innovative approach as, instead of developing new models as is the normal practice for such projects, it tested current models to react to the issue.

The team found that the best tool for the job, of predicting future revenue based on financial KPIs, was a Random Forrest model, from the machine lending model group. This was due to its ability to utilise the entire data set simultaneously, ensuring it was able to predict financial returns even when not using customer score data.

When implementing new analytical tools, a significant challenge is always integrating them into an organisation’s current operations. To ensure that the new methods could be implemented without experiencing data loss in another area, it was recognised that updating the model parameters could not realistically be done in real time. To address this Maru/Syngro ran the proposed modelling for all KPIs and parameters of the model once a month, when new financial data was uploaded onto the system. Once this is done the new parameters can then be stored and accessed immediately as the user interacts with the system.

As the use of machine learning models is not common practice in customer satisfaction analytics, Maru/Syngro’s application is an internal innovation. Moreover, the combination of predictive analysis, visualisation and workflow tools is a commercial innovation.

What does the future look like

The end product, Syngro Eye, made possible by the collaboration with The Data Lab, Maru/Syngro and the University of Strathclyde, has predictive analytics included within the product’s capability, which also includes self-service reporting, analytics and action driving.

The new capability ensures Maru/Syngro clients are able to use customer loyalty data to prioritise internal resources. The net result flows to the bottom line by enabling clients to retain customers through addressing areas that would previously have been missed and potentially could have led to customer defection.


I would like to thank The Data Lab for helping to facilitate this project and getting it off the ground. The Data Lab is very important in helping data driven companies adopt new techniques and functionality, and plays an important part of the technology landscape in Scotland.

- Keith Schorah, CEO of Maru/Syngro


Mailing List


To recieve updates from The Data Lab please enter your details below.