Resource / Knowledge Base
From raw data to visual analytics, we're experts in helping organisations draw conclusions from data
We work as a team, making use of our wide range of technical expertise to ensure the quality and accuracy of data, then process, design and present it in ways that help in…
- Discovering useful information
- Informing conclusions
- Supporting decision-making
Understanding the problems you want to solve is the first and most important phase in your data analytics project. We use an agile approach to building solutions, starting with you, your stakeholders and key users to gain a good understanding of your organisational needs, what processes need to be improved and what the end goals are. Once we have clear objectives we will work with you to plan timelines, targets and plot key performance indicators before starting on your solution.
Getting Hold of Your Data
Whether your data is stored in databases, static files, online services (e.g. Pipedrive, SalesForce, Google Analytics) …… we can certainly help you extract and combine it. We can also help guide you on the huge amount of freely available data which may compliment your datasets to give you even more insights.
Exploring, Cleaning and Enriching Your Data
This is the stage most people hate, but we love! It’s the digital equivalent of polishing a rough diamond, slowly revealing its beauty before our very eyes. Whether your data is cloud based or on-premises we can work with you to extract that data, even building our own DLLs and APIs to facilitate the process and incorporate into your tailored solution. There isn’t a single project we’ve come across that hasn’t needed a lot of cleaning, normalising and combining, or that can’t be significantly enriched when combining with external data. Our approach involves building routines on an Agile basis which we can run, time and time again, refining as exceptions are identified and building new logic back into the evolving solution before narrowing down to essential features (such as time-based features).
Building Helpful Visualisations
Rather than scoping out a design from the outset & then being rigid in its application, in keeping with our Agile approach we initially provide quick proof-of-concept designs, often with several approaches trialed for feedback. This helps to generate discussion about preferred styling, the art of the possible, as well as often presenting approaches or hidden truths not previously considered. From this starting point we will then coalesce the initial sketches into production ready deliverables, run UAT & extensive QC’ing before publishing to all stakeholders. No design needs to be set in stone and we encourage constant feedback from all levels of stakeholders so we can adapt to new requests or changing situations, as well as build on finalised views to deliver more in-depth analytics and enhance viewer insights. To learn more please read Data Visualisation, or any of our Case Studies.
Machine learning algorithms can help you go a step further into getting insights and predicting future trends. By working with clustering algorithms (aka unsupervised), you can build models to uncover trends in the data that were not distinguishable in graphs and straight statistics. These create groups of similar events (or clusters) and more or less explicitly express what feature is decisive in these results. The next step would be to dive deeper and predict future trends with supervised algorithms. By analysing historical data, it is possible to discern features that have impacted past trends, and use them to extrapolate predictions. More than just gaining knowledge, this final step can lead to building entirely new products and processes.
Refining and Iterating
Because we’re careful to segment each stage of your data analysis project, making changes anywhere along the chain is a relatively straight forward process and safeguards against future “unknowns”. This can be exceptional helpful in dealing with new exceptions, changes in trends & novel data sources – or indeed changes in your analytical requirements – as we can adapt any pertinent module without needing a comprehensive redesign of the system as a whole. Moreover, we will sit down with clients periodically to discuss any potential changes, so we can proactively adapt before any crucial crunch events.