History of Big Data
The term ‘big data’ has been around for decades, and has grown considerably from its small beginnings to the hot topic it is today. Forbes gave a potted history of it back in 2013, noting that the concept (under various different names) started back in libraries in the 1940s.1 However, the big data analytics market did not really come into its own until more recently, when mainstream companies, particularly across the retail sector, began to get on board. The UK’s largest food retailer, Tesco, is a ‘big data’ pioneer, being one of the first major supermarket chains to track customer activity through its loyalty system. As cited by Forbes: “Applying cutting edge analytics and the most upto-date data is the supermarket’s answer to dealing with obstacles ranging from evolving customer behaviour, to facing up to newer competitors.”2 Historically quantitative data was difficult to manage, model and mine, but with the ability to ‘machine learn’ via increased computer processing power, this has now changed.
The competitive advantage gained from data analytics, combined with these advances in machine learning, the advent of faster broadband speeds and affordable, secure cloud storage has led to increasing numbers of organisations across multiple sectors to manage and extrapolate value from their data. We have seen a shift from organisations fighting hard to purge their data to save storage costs, to today, where they strive to maximise it – doing everything they can to store, analyse and utilise their data to better understand their business, the sector they operate within, and their customers. At the end of the day, every organisation produces data, and increasing numbers have become savvy to the benefits it can offer. But this educational piece has taken time – organisations simply don’t know what to do with the data they have. We are also seeing a rise in the number of systems being launched to manage data mining activities; a trend that we will see more of this year and into the future, especially in industries relatively new (and arguably a bit late) to the big data game. Enter the commercial finance market…
Commercial Finance and Data Analytics - Why?
“Commercial finance has a lot to gain from using data science. Delving into data and deriving trends will provide lenders with a much greater understanding of their clients and potential clients
“These limiting factors mean that within the commercial finance sector, big data analytics has largely been a series of theoretical and proof of concept exercises”
There is no doubt that the commercial finance industry is playing catch up when it comes to making the most of its data. Historically, we have consumed large amounts of data, but we have not fully utilised it. Data was kept for compliance purposes, but not capitalised on. Risk analysis has always been there, but the technology has advanced and many commercial finance providers are missing out on opportunities to realise other business benefits from their data. There are a number of reasons that could be cited for this slow adoption process, including:
- Failure to see what could be done, and what could be achieved via data analytics.
- Lack of communication between the retail and commercial banking divisions. Commercial finance is often seen as the ‘poor cousin’ of its retail banking counterpart. This means that rather than sharing processes, data and learnings, the two divisions generally operate in separate silos; a case of ‘never the twain shall meet.’
- Business to business (B2B) organisations have fewer clients, transactions and ultimately, less data than their business to consumer (B2C) counterparts.
A BBA and Ernst & Young report indicated that transactions worth £6.4 billion a week are made using mobile and online channels in the UK and in 2014 banking services got about seven million logins on a typical day3 – both in the B2C sectors. When it comes to domestic factoring, domestic invoice discounting, export factoring, exporting invoice discounting and import factoring, total client sales for the year up to September 2016 stood at £221,038 million4. These limiting factors mean that within the commercial finance sector, big data analytics has largely been a series of theoretical and proof of concept exercises. Equally, delving into data is likely to be a more cumbersome and expensive endeavour, which may not produce the same immediate powerful results in B2B as it does in B2C. In spite of this, early adopters can already see the promise of leveraging their data to establish patterns and trends, therefore leading to better performance levels and – ultimately – increased profitability. Moreover, having the ability to retain their current customers for longer and grow their client base while reducing risk through the analysis of patterns and predictive analytics will pay huge dividends for commercial lenders. Having a better understanding of clients’ financial and operational needs will help them to compete and improve the reputation of the industry as a whole. The flipside, if they don’t mobilise their big data strategy, is the potential of being left behind. What big data should mean to the commercial finance world – and any sector – is no more than reliance on ‘gut feeling’. By analysing data to spot patterns and behavioural inconsistencies, lenders should no longer need to ‘second guess’ potential outcomes. Combining results from these proof of concepts with lessons from other industries, the commercial finance sector can question preconceived notions and move away from legacy models, starting with a fresh perspective more aligned to specific needs of the sector. For example, using analytics to promote client retention based on preferred service criteria. Data analysis of benchmark data can also be used to assess risks previously associated with certain sectors and give commercial lenders the confidence to open their doors to new potential clients who may not have previously been considered, helping promote expansion, even into new industries.
“When following the big data route, commercial finance providers should choose and focus on the strategy that is most appropriate to them and aligns with their business objectives”
The time has come for the commercial finance sector to embrace the growing levels of data being produced. The sector is maturing, and data science can support this trend by helping providers to understand their clients and use this insight to help them develop and grow. As the concept of data analytics has also grown, technologies have developed to support the entire data life cycle, providing real-time and predictive insights.
Selling the Concept of Data Science - Focus Areas
Data mining and analysis within the commercial finance sector should be centred around the themes of customer retention, fraud reduction, risk control, personalisation and supporting business growth. Rather than starting with all of these, choose the theme most relevant to your business to build your proof of concept around. From there, you can apply your learnings to build a wider strategy that includes elements of each:
RISK/REWARD This is the area that will keep most commercial finance providers awake at night. The risk/reward paradigm is evident; greater risk yields greater rewards, expense and – with it – issues. Mining data has the ability to change risk levels, to give the lender greater faith in their activity – whether that means taking on a new client who may have not qualified previously, or extending facilities to an existing client. Knowing more about your data and – ultimately – your clients and prospects enables you to make smarter, more informed decisions. It may alert a provider to a particular risk, but data knowledge could also flag opportunities in riskier markets – those which offer a higher return, and which may have previously gone unnoticed.
FRAUD REDUCTION Having a greater understanding of your customers empowers you, and puts you in control – helping to detect fraud and mitigate risks. We have already established that it can give organisations the potential – and the power – to take greater risks. Fraudsters have become more sophisticated in their approach, and smart data analysis has the potential to help combat this. Having access to data earlier on in the process will help detect potential inconsistencies and – in some instances – possible fraud before it has the power to take hold.
CUSTOMER RETENTION What would happen if a provider could extend the lifetime of a customer service agreement? How much value is there in extending the average threeyear lifecycle of a client; even if it’s only by six months or a year? Enough to make the Board stand up and take notice? We think so. Using data analytics and mapping trends can help commercial lenders gain greater insights into their customers, including highlighting triggers that they may be ready to move on. Knowing this information means you can take steps to keep them, win them back or – if they are set on leaving – at least acquire a new client to make up for any shortfall.
PERSONALISATION Business intelligence and other developments in data science technology enable providers to deliver a more personalised service, based on customer behaviours and trends. Even more importantly, it can help commercial finance providers to predict client behaviour and stay one step ahead. For example: client A could be reaching the end of their life with a provider, and the fact that they are beginning to draw less funds could be an indication of this. Is the business failing or are just reaching the end of their trading season? Either way, it’s important to understand client behaviour. Instead of letting a good client go back to the market. Commercial lenders could proactively adjust commercial terms and retain the client. There is no doubting how competitive the commercial finance market is. When business arises myriad providers chase it. Strong data analytics and support can prevent this happening – having a better understanding of customers will help to increase the continuation of business relationships.
BUSINESS GROWTH In addition to analysing data trends to help providers retain significant business and prevent the risk of lending to the wrong customer, it supports businesses in their own growth plans. For example, access to an in-depth view and understanding of your customers’ behaviour patterns allows you to offer them other facilities that align with their growth strategy – e.g. supply chain finance, trade finance or foreign exchange services. Data can also be used to support targeted marketing activity, for example defining specific parameters around an ideal customer who – perhaps – isn’t using commercial finance. Or, this may be focusing sales activity in a specific area or industry, which is particularly buoyant at a given time and generating a marketing campaign to support this. Building a compelling business case for investing in data analytics requires time and an understanding of your organisation and how it works. The next step? Moving from proof of concept into the ‘real world’ – applying your theory and learnings into creating a wider data analytics strategy aimed at embracing the core themes and growing the business.
BUILDING THE STRATEGY Commercial finance has a lot to gain from using data science. Delving into data and deriving trends will provide lenders with a much greater understanding of their clients and potential clients, allowing them to better assess risks and adopt a more customer-centric approach that will help to remove commercial finance’s tarnished reputation. There is no doubting the sector has – historically – had an image problem; often considered to be ‘the lender of last resort’.
But in order for it to be successful – and profitable – it is imperative for data science applications to have buy-in and engagement from all levels of a commercial finance organisation. The business case needs to be made; proving that investing in data will yield higher returns, reduce risk and result in greater customer retention levels. It has to be tangible in nature. Taking the leap of faith needs to be supported by a strong business case and Return on Investment (ROI). Data science is by its very nature ‘exploratory’. It’s almost akin to panning for gold; when you find something, it can be the ‘game changer’. With this in mind, here are a few over-arching tips: ◆ START SMALL – when it comes to mining data, always start with a proof of concept, and run pilot or test projects. Begin with the ‘frontier of your knowledge’. This will help ensure measurable results, and – sometimes – a very quick reward. It will also help to build overall confidence levels in the ‘big data’ concept.