Analysis is, of course, a significant part of everyone’s lives at both Geordie Consulting and Geordie Intelligence. However, it is a topic that covers a vast area. There are two types of analysis. Although helpful, they are split into three areas often considered in terms of maturity. What is not beneficial in this is the implication that as you “mature”, you can almost ignore what came before. The reality is that this is a journey of growth. Let us take some time and talk about it from a growing point of view.

No matter what we do, we must understand how well we do it. That is the cornerstone of the foundation of Analytics — Descriptive Analysis. It is about “How many transactions occurred?” or “What was our monthly Gross and Net Profit?” The purpose is to “Describe” what has occurred in your organisation. It explores and describes the actuals. Have you worked out the challenge that Descriptive Analytics have? The challenge they face is that they lack context. When we talk about context, we mean targets. If a business makes £1,000,000 in sales in a month but targets 2,000,000, it achieves 50% of the target; meanwhile, if the target is 500,000, then we achieve 200% of the target. There is a marked difference in how that will be received in the Board Room. So, how can you define targets and ensure they are correct for our organisation? We can all agree that having the Board arbitrarily decide what our business targets, or should there be some logic to the process? I am sure that there is not a business owner out there who would love to be able to come up with arbitrary targets and then have the business achieve them. Still, in a world not fuelled by sunshine and lollypops, we need to apply a modicum of science to the process – I say it this way because there is always the overlay of “We want to expand in that region” or “We are walking away from this product”.

We typically start with a known period of Descriptive Analytics and then apply a modifier and project that in the future. So for example, “In Q2 of 2023, we made £100,000 in revenue in the UK, we want to increase our market share by 25% in 2025, so our target for 2025 is £500,000 (100,000 4 1.25)”. So, we have the basics of what goes into the next two types of Analysis, which are inherently future-focused. Keep the £500,000 target in mind. Our board seeks confirmation that it is valid. Enter Analysis type two, Predictive Analytics. Predictive analytics looks at the actual performance (Descriptive Analytics) and, based on the month-over-month performance from January 2020 to Today, projects that in 2025, revenue will be £450,000 with an x% confidence (normally 80%+). A lot of maths goes into this way of producing projections, as seasonal variations are typically considered. Over time, these models can become highly complex, but ultimately, they are using past performance to predict future performance; then, in the future, we can compare actual performance to predicted and use that to modify the model and so further enhance our predictions. If a single sentence could sum up an industry sector, that was it. This is the basis of Machine Learning and the mechanism of Generative AI.

The evolution of Analytics in Generative AI and Machine Learning is simple: Prescriptive Analytics. If we go back over our data and look at sections of our Descriptive Data Archive while overlying it with external factors, weather, interest rates, Unemployment %, etc,… we identify where there are correlations and build up and test historical data against historical predictions. The result is a statement: “To achieve £500,000 in revenue in 2025, the Bank of England will need to keep interest rates below x; given the trend for more adverse weather, you should also expand in the following product areas….” This is the result of Prescriptive Analytics, a set of internal and external actions that must be undertaken to achieve the desired results. It may even go as far as Prescribe doing something completely different to achieve a maximum benefit, so £1m in revenue. The ability of Machine Learning to replay scenarios thousands of times in a short time with different variables gives it its allure; however, remember that the reality is that both Predictive and Prescriptive Analytics are fantasy, and they can both be invalidated in an instant. That does not invalidate them, but it should remind us to be careful. Projections and Prescriptive plans are, at their core, the same as a board-defined target, granted we may assign them more “confidence”. Ultimately, they will come back to the quality and capability you manage to implement within your Descriptive Analytics models. The ability and efficiency of your future models and plans are maximised by the stability of the foundations you build.

Geordie Consulting uses the Microsoft Power Platform to provide businesses with agile, cost-effective data solutions. We specialise in integrating data across business applications into a single, manageable platform. Our services include enabling business users to develop custom reports using Power BI and Excel, leveraging Semantic Models for consistent data analysis, and enhancing user self-sufficiency through Copilot. We aim to ensure your data agility to support a data-driven business. We provide comprehensive documentation and support to maximise the business value provided by your data. Our insight unlocks your insights.

Default sample caption text

Contact Us:

email: office@geordieconsulting.co.uk

Tel: +44 191 432 7330