Working with clients to find optimal investment strategies for maximising the value of a portfolio – where value is often a composite of measures which needs to satisfy the perspectives of multiple stakeholders. Much of our work has looked at optimal investment strategies for tangible assets – water company infrastructure, for example.
Predicting the likelihood of a given outcome from a potential change. We can develop and apply models that forecast how likely a given level of impact is to happen. This reveals not only the most likely outcome but also the probability of an extreme impact, whether good or bad.
Using our analytical and modelling skills to identify benefit metrics, forecast the trajectory for achieving these, provide baselines and devise methods to attribute performance changes to the associated change programme. We also build custom benefits reporting systems and link these to data sources.
Developing metrics and methods for measuring these, to provide meaningful, timely and practical information to stakeholders. We often develop prototype reporting systems to allow performance measurement to be refined before being embedded into existing management information systems.
Providing analysis to support all aspects of business planning – forecasting future demand, using simulations to identify the resource requirements to meet demand, assessing the costs and benefits of proposed process changes and developing the measurement systems to track progress against plans.
Using statistical techniques to forecast future performance or demand. We use existing data to understand how a business reacts to changes and use this to forecast what to expect given the likely drivers in the future. Examples have included forecasting sales of food products, forecasting customer satisfaction scores and forecasting demand for the services of a national charity.
Building datasets and analysing these to understand what drives the observed behaviour. We make sure that the datasets are accurate and representative. We use a range of statistical techniques to understand what factors drive the changes in the data and develop reports and systems that let the customers understand and make use of these insights.
Using graphical formats to present results in a meaningful, insightful and readily comprehensible way. This might involve selecting the right chart format, or using geographical formats, or presenting groups of charts that illustrate different aspects of the analysis.
Applying our skills and experience in building and delivering models and analysis to a high level of quality. We are often asked to review/test third party models and are also asked to review in-house quality assurance procedures for models and analysis
Ensuring that analysis is based on good quality data. We can clean data, we can convert data to provide consistency, we can join datasets to create rich sources, we can sample from data to create representative subsets. The results are only as good as the data upon which they are based.
Helping to appraise potential policy options. We can help identify policies – by providing insight into the root causes of problems. We can model complex policy impacts to understand which policies will deliver the required outcomes within the constraints on action. We can analyse feedback from policy consultations. We can help to build consensus on the right policy to implement across diverse stakeholders.
Understanding the detailed impact of a planned change. We can build models and simulations that evaluate the likely impact of a planned change, often where complex interactions in the system make it hard to foresee the full implications of a change. We can use What-If analysis to assess the outcomes for a given scenario. We can also forecast how likely it is that a given level of impact is felt.
Working with data to provide deeper insight or rapid analysis, allowing clients to react more effectively. We understand both how to collect and clean datasets and also how to analyse and present the results, giving meaningful, powerful information.
Helping to make complex decisions. We can use structured methods to understand the key criteria for making a decision and synthesise the readily quantifiable (costs, performance, etc) and the qualitative factors (perceptions of trust, political factors, etc) to form a single framework for making a decision. We can facilitate the process from developing the framework for assessing the options through to rating the options and reaching a consensus decision.
Quantifying the benefits of a project or programme. We can use our analytical skills and experience to forecast the expected benefits – improved performance, cost savings, revenue increases – for a project, providing a key element of the overall cost benefit analysis. We can also quantify the risks of the project and resources required to deliver it.
Predicting the impact of changes to HR policy and how best to deploy those resources. We can develop models to forecast the impact on staff costs, turnover and gaps arising from changes to pay, recruitment or pension policy. We can build models to calculate how many staff (by grade) are required to deliver a revised service/process.
Making rational and auditable procurement decisions. We can use structured methods to allow disparate stakeholders to agree how a decision is to be made and to provide a framework that combines the readily quantifiable factors to be combined with the more qualitative. The process is objective, builds consensus and provides a full audit trail of the route to the decision.
Collecting data and views in a meaningful way. We can help to design questionnaires, develop representative samples, run the surveys, analyse the results and publish the results. We can ensure that the outputs are meaningful, useful and insightful.
We are skilled and experienced in applying statistical methods to gain insight and understanding from datasets. We have an indepth knowledge of the techniques of illustrative statistics (e.g. regression, time-series and ANOVA) and hypothesis testing (t-tests, Chi-squared tests and correlation analysis). We can ensure that the confidence in statistical results is understood and can apply methods to structure data to improve that confidence. We also understand the importance of presenting the results in a meaningful and accessible way.
We have experience at capturing data directly, or designing systems for data collection. We collect ‘soft data’ / expert opinion through surveys, structured interviews, and workshops; and ‘hard’ data from source systems, data warehouses and published reports. Further we have experience combining soft & hard data to inform decision making.
Organisations frequently now find themselves swamped by data but starved for information. We have experience sifting through large volumes of ‘messy’ data, manipulating it into a form amenable to analysis and assessing its quality and utility.
We have significant experience creating tools to solve optimisation problems. Whether the challenge is to set employee rotas, constrained by complex rules on working hours, or choosing the set of capital investments which will provide your business with the maximum value for money, we use mathematical techniques to structure and solve them. These can be integrated into your existing systems, to support your key decision makers.
Analysis of historical performance provides insights into what drives that performance, but the real power comes with the ability to use this understanding to predict expected future performance. We can offer a variety of approaches from the purely statistical through to the use of complex models. We help clients to understand what is most likely to happen and how likely deviations from this are, supporting both planning and risk management.
Decision-making is often complex, with multiple stakeholders (and perspectives), many criteria (some readily quantifiable, others less so) and disparate options to consider. We are experienced at structuring these complex decisions, building consensus on what factors to consider and their relative importance and rating the many options – and can facilitate the end-to-end process, providing an audit trail to show how the decision was made.
Representing data visually can be an extremely effective tool for communication and persuasion. Whether it’s a simple time series chart showing seasonality and growth over time, or a sophisticated force-directed graph to show the frequency of co-working between staff, we have the experience to help you present it. Our favourite out-of-the-box tools for data visualisation is Tableau; our most frequently used is Excel; and our most powerful is d3js.
Understanding geographical aspects of data often brings added depth to the overall picture, but on occasions travel time is the key aspect. We can make use of the wealth of travel time information now available to bring this important dimension to analyses.
For large programmes of work it is potentially difficult to quantify all the costs and benefits – but it is essential to do so in order to present a meaningful cost benefit analysis. We can help to develop methods (and source the data to feed into these) to quantify a wide range of costs and benefits.
Finding the nugget of gold buried deep within a mass of data takes a lot of expertise. We are knowledgeable and experienced in using the standard tools to collect/access and manipulate data, but add the value of knowing how to identify the value of what is found. Our statistical knowledge and skills in data visualisation allow us to find the insights – and quantify confidence in these – that simple analysis by rote cannot deliver.
Pattern recognition is an important tool for both identifying spotting correlations in data that provide fresh perspectives on cause and effect and for image processing. We can support both applications. We use statistical techniques to establish correlations in large multi-variate datasets, using other statistical techniques to understand the strength of the linkage. We have developed techniques for matching complex patterns drawn from images, bringing additional value from data collected for other purposes.
Data science is the core of what we offer. It covers a wide range of methods, tools and techniques, but is at the heart of what we do. Our emphasis is on using statistical, logical and mathematical techniques to manipulate and analyse data and to present the results (rather than on the specific hardware and software to capture and hold the data). We believe that this is where the true value is added – a mass of data and a set of tools are of little use if you don’t have the knowledge and experience to know how to use them.
Spreadsheets offer a quick, powerful and highly affordable means of analysing data and presenting results. We are power-users of spreadsheet modelling and have countless years of experience. We routinely extend spreadsheet models with custom code (e.g. VBA code added to an Excel workbook) and have built highly complex models as well as simple one-sheet models. Our added value comes from our insistence on proper design, thorough testing and documentation – plus the ability to recommend when a spreadsheet model isn’t the best solution. We regularly help our clients design and build models and are often asked to act as independent testers of 3rd party models.
Whether it’s optimising capital investment using linear programming, market segmentation using k-means clustering, or pattern recognition using neural networks. We have experience in the practical application of machine learning techniques to solve the problems in your business.
Surveys provide an excellent way to collect data and opinions, but a badly designed or analysed survey can provide misleading findings. We are highly experienced in the design of surveys (both in terms of the questions to ask and the selection of the sample of people to take part), the running of surveys (particularly on-line surveys) and the analysis of the results. We can help a client with one aspect of a survey or run the whole survey and reporting cycle on their behalf – making sure that the results are meaningful and useful.
We can offer tailored training and support to your in-house analysts on Excel / VBA, Business Intelligence reporting, or any of the tools & techniques we employ. Whether delivered as a part of a larger project, or to meet a stand-alone need, we can help you greatly increase your analytical team’s productivity.
Some of the tools we use
Our favourite tool for Business Intelligence and easy data visualisation
Our go-to relational database
Powerful discrete-event simulation software
We use R for data processing and statistical analysis
For web interfaces and bespoke data visualisation
Our preferred tool for bespoke, responsive, interactive data visualisation
Ubiquitous ``swiss-army knife`` of analytics
For turning expert opinion into quantified actionable insight
For programming Web APIs, data wrangling and machine learning
Graph databases are best when your data is highly networked