Cutting through the hype - how to use advanced analytics to do practical things today webinar
0 likes528 views
The document discusses the challenges faced by capital programs, particularly their high failure rates due to complexity and data management issues. It highlights the Oakland Intelligent Forecasting platform as a solution that leverages artificial intelligence for improved project forecasting, data integration, and better decision-making. Key benefits include significant ROI, enhanced reporting, and the ability to predict and mitigate risks effectively.
Cutting through the hype - how to use advanced analytics to do practical things today webinar
- 1. Intelligent Forecasting A practical guide to finding the value in your data Cut through the hype of project analytics with a real-world solution demonstrated to reduce risks, increase confidence and improve outcomes in major capital programmes.
- 2. Introduction Capital programmes are prone to failure, partly due to their complexity and often long gestation, and can quickly become the subjects of critical reports. Crossrail is a great example. This 73-mile railway under London is running more than two years late and will cost over £2bn more than the original £15.9 billion estimate – with further setbacks looking likely. In its 2019 annual report on UK major projects, the IPA said its overall confidence in the delivery of projects had been steadily falling since 2013: only 22 from 133 were rated as ‘probable’ or ‘likely’ to achieve their aims and objectives, and to do so on time and on budget. It is little wonder such ‘megaprojects’ are often in the news and subject to constant political and public scrutiny. It is, however, important to acknowledge that these programmes are incredibly technically challenging and major project achievements can be obscured by news of cost overruns and delays. Long timescales, some running into decades, mean it is inevitable that plans will change. In 2009, HS2 was initially intended to run from London to Birmingham; by 2017 the plan had been expanded to include lines running much further north. These hugely multifaceted engineering challenges also demand intricate supply chains and funding arrangements, adding to complexity and creating a vast number of moving parts. And every part of the delivery structure also creates data – these increasingly massive volumes of data have been challenging to integrate, and it has been difficult to extract value from them. However, the latest information technology can revolutionise the way major programmes are delivered and help to deliver more for less. Leading businesses are taking advantage of tools such as the Oakland Group’s Intelligent Forecasting Platform to continue their journeys towards digital transformation. 2 3
- 3. So, what is the answer? The Oakland Intelligent Forecasting platform is a scalable platform which services both reporting and analytics within a single environment. Harnessing powerful artificial intelligence (AI) and Machine Learning to help organisations make better decisions, it transforms reporting and predicative capabilities. What can be achieved? • Vastly improved project forecasting - over 50% more effective than project managers at spotting expensive failures up to 9 months out • Enhanced board and management level reporting • Improved process compliance • Targeted assurance activity • Better data quality • Strong Return on Investment: on average the Oakland platform delivered a 10:1 ROI in the first year, with improvements in years two and three 4 5 Intelligent Forecasting - A Practical Guide
- 4. Complex capital programmes face significant project failure rates, Bent Flyberg, an expert in project management at Oxford University Business School, estimates that 9 out of 10 projects go over budget. Why do you need Intelligent Forecasting? The Oakland Group are experts in complex operational issues. We understand running major projects is difficult. Most major projects have complicated structures and complex information architectures with a whole raft of different source systems; integrating their data can seem like an impossible task. Our Intelligent Forecasting platform helps to transfer project learning more efficiently. It enables you to predict where and when risks are likely to happen. Intelligent Forecasting - A Practical Guide 6 7
- 5. Understanding the pain of capital programme delivery Data is nothing new - businesses have been producing and storing data long before the ‘Big Data’ explosion. However, typical data flows are complex, extracting data from large enterprise systems can be difficult, and replicating data to the cloud where it can be utilised more effectively has been problematic. Common challenges include: • How to improve forecasting, being able to spot both the underspends and overspends? • Knowing where to target your manual assurance and audit activities? • Extracting additional value from data and spotting areas of poor process and data compliance? • And finally, how to unlock deeper insights – perhaps through advanced predictive analytics, machine learning and artificial intelligent tools – to drive the organisation forward? 8 We recognise the challenges which come with capital programmes. There is no silver bullet; no single solution can do everything we need. Intelligent Forecasting - A Practical Guide You need to: Understand the key business decisions that will benefit from being more data-centric - for example: • Areas where stakeholders are sceptical • Areas where there are constant ‘surprises’ Challenge why simple questions are hard to answer • Do you require excessive ‘calls for data’? Why? • Would you get the same answers if you asked the questions twice? • Would you get the same answers if you attempted to reconstruct the results from a few months ago? Understand where reporting is being ‘manipulated’ to make it work. • Why is the source data not sufficient? • Is there organisational data missing (i.e. project to programme mapping)? • Does the business connect the same errors in the data every month? • Are key aspects of the data held ‘off-system’? Identify appropriate project controls • How does your lead time direct assurance and validation activities? • Experience is vital, but could you justify why some projects get checked while others don’t? With our help, companies can integrate, access, interrogate and understand their own data instantly. They no longer rely on monthly reports that provide only a fraction of the true picture. They can also extrapolate their data to predict more accurately what might happen in the future. Major programmes struggle to collect clear and consistent data. Sometimes data is also under-utilised or not used effectively. 9 Legacy Architecture ApplicationsSeparate Source Systems Disjointed Reporting Ad hoc analytics Current issues • Manual Integration of information • Parallel, inconsistent and excessive cleansing activity. • Hard to combine data across platforms - analytics often restricted to limited datasets Siloed reporting Over-reliance on ‘experts’ vs data- driven decisions Difficulties in maintaining the corporate memory
- 6. The Oakland Intelligent Forecasting Platform A scalable architecture that services both reporting and analytics within a single environment 10 Intelligent Forecasting - A Practical Guide How the platform works 11 Identify and integrate data Data sources need to be identified and some level of standardisation agreed. We begin with the data - our Intelligent Forecasting platform integrates data from different data sources: spreadsheets, Oracle databases, PostgreSQL databases, SAP and even Blob storage. We integrate these data sources into our cloud environment - Microsoft Azure or AWS – and then apply Databricks, a core part of our Intelligent Forecasting platform. 2 Set objectives We understand your business projects are similar, but not identical. The starting points must be your business and project objectives: what activities do you want to undertake and with what outcomes? 1 Intelligent Forecasting Platform VisualisationIngestion Oakland 4D Data Model Advanced Analytics Intelligent Forecasting Data Quality
- 7. Analyse and predict Using our ‘4D’ data model we can carry out Intelligent Forecasting, making predictions using machine learning and artificial intelligence. We can carry out data quality assurance to ensure we are meeting internal processes and compliance checks, And we can undertake advanced analytics to identify the issues causing projects to underperform. Visualisation is a vitally important way to engage teams and generally help reduce manual intervention. We can then export the data into any front-end platform such as Power BI, Tableau and Looker. But it doesn’t stop here. While we can start to predict future risks, we must also be able to implement changes needed to bring the project back on track once areas of concern have been highlighted. This involves looking at different aspects of quality and governance, and at implementing process change. Not everyone in the organisation may be happy at having a light shone on their areas of responsibility. It is essential, therefore, to create a culture where project managers feel confident to challenge the way things are being done. The Intelligent Forecasting platform provides a powerful tool linking system outcomes to project level benefits, and giving teams confidence that changes will improve outcomes. 3 4 Ingest and expand data Having aggregated lots of different data sources in a variety of different forms, ingestion is a key step ensuring we have collected all the data needed. Then our unique Oakland 4D data model creates the ability to look at any time period across a project or portfolio. This allows project managers to view the data from any point in history. It is this ‘4th dimensional view’ that also drives our Intelligent Forecasting platform. 12 13 Intelligent Forecasting - A Practical Guide
- 8. What is the impact of Intelligent Forecasting? Intelligent Forecasting provides not only consistently objective results but also more accurate predictions on the current progress of a project. This insight is not just focused on short-term predictions, it can highlight potential risks as far out as 6-12 months’ time. • We have identified which projects don’t hold the right data, are missing milestones, or are missing other critical information. • We have intercepted bad practices that were previously undiscovered (for example, people updating forecasts after actuals). • We found open projects that should have been closed and which carried substantial amounts of risk contingency. In short: there are huge benefits to be gained from helping your organisation become more data-centric and from incorporating Intelligent Forecasting tools and thinking into day-to-day operations. 14 Intelligent Forecasting - A Practical Guide At a 9-month range, PMs had traditionally identified about 40% of the projects that were on track to go wrong at the end of the year. Using the Oakland Intelligent Forecasting platform this risk was more accurately predicted at over 70%. 70% 15
- 9. 16 Intelligent Forecasting - A Practical Guide Q & A My software vendor says they can do this There are some greats tools out there but many applications focus mainly on presentation and reporting. Our platform utilises and integrates historic data to analyse, to provide vital insights, and, critically, to make predictions. My data is poor quality The processes of data ingestion and integration (which is often 70% of the work) helps improve the quality of data enabling us to extract the insights you need. How do we integrate your team with our team? We work very much in collaboration with a client’s team, running technical workshops and helping to upskill where necessary. How long would a typical project take? This could be anything from 3 months up to a year - depending on the complexity of the project and any subsequent ‘spin offs’ 17
- 10. 18 What sort of questions can Intelligent Forecasting answer? Our Intelligent Forecasting platform helps answer detailed questions that were traditionally difficult to answer: • What is the value of risks compared to the approved cost envelope? • Have activities been closed correctly in our data? • How close to deadlines do people wait before changing forecasts • Do expected historic risks map to actuals at the portfolio level? • How often do PMs roll the previous period’s spend forecast into actuals? • Is frequency of updates a sign of control or of panic? • Which projects have mandatory milestones? The platform is configured to deliver the answers you need and can be fully customised to your organisational and project requirements How do I get started? Digital transformation is a journey, not something that can be achieved overnight. We suggest starting small with an initial proof of concept. This will help demonstrate the art of the possible and begin to win buy-in from your stakeholders. Intelligent Forecasting - A Practical Guide Initial POC System Integration and Architecture ‘Production’ (Support and Develop) Who have Oakland worked with? 19
- 11. Construction errors cost the UK industry billions every single year. According to the Get it Right Initiative, annual spend due to errors is estimated to be around seven times the total annual profit of the UK construction industry. 20 To organise a demonstration of the Oakland Intelligent Forecasting platform please contact: intelligentforecasting@theoaklandgroup.co.uk 21 Intelligent Forecasting - A Practical Guide
- 12. www.theoaklandgroup.co.uk 1 Aire Street, Leeds LS1 4PR 0113 234 1944