In the near future, an estate agent will know who will list a property on the market in an area, and at what price, because AI-driven insights products will say so. Similarly, data will help a car dealer understand which vehicle and value-added products will provide the most joy to a specific customer on their floor while optimising their profit.
These property and auto industry scenarios are just two ways the almost limitless value of data is being unlocked to create strategic and operational benefits that help businesses sell more properties, vehicles, financial products – and almost any other consumer product or service.
Data has become an increasingly important asset over the past twenty years, and while data sources are not always effectively commoditised and their lack of “fungibility” can make it difficult to put a direct value on it, their strategic and operational value is widely recognised. Research shows that global C-suite executives believe that companies that “excel at integrating data into their strategy, operations and culture are largely outpacing their peers in revenue growth and profitability”.
Transforming data to insights is, however, the key, as idle data has little value. British mathematician Clive Humby said in 2006 that “data is the new oil” and he was referencing its potential value once it had been refined, processed and transformed into something useful.
The more complex and uncertain the world gets, the greater the demand for crystal clear insights from data because it helps provide a more predictable roadmap through economic disruption and political turmoil.
Paul-Roux de Kock, Chief Analytics Officer at Lightstone
The strategic and operational value of data grows exponentially with more history and diverse sources of data, whether externally sourced or internally produced by products and systems, and of course in how data is processed and transformed to unlock insights which have a positive impact in the real world.
The purpose of building data sets and unlocking insights, in our world, is to enable our clients with “superpowered” decision making abilities when using our products to build a competitive advantage in their business.
Building data value
Aggregating customer and market data to gain competitive advantage is not new, although the process used to be slow and limited in scope and scale.
Despite this, the value of data grew with time – the more history there was, the greater the opportunity to discern trends and improve interpretation by comparing data sets from one period to another. As an example, assessing deeds data at a certain date offers only a limited view of the property market if compared to analysing deeds data over our comprehensive historical property dataset.
Consumer technology changed the way data is collected and stored, but more importantly, it revolutionised how different data sources – from social media and GPS location to content choices and sales information – can be integrated to create deep insights into customer behaviour that were not possible before.
Now, AI and machine learning analyses a person’s “digital exhaust” which reveals so much more in terms of interests, activities and location. And when thousands or millions of “digital exhausts” are blended, the strategic and operational value of the data increases exponentially.
In a B2B context, widely available generic data is becoming less useful, but significant value can be added to the usability of the information if layered with unique data, some of which comes from using proprietary products and systems – in our case, Lightstone’s Property Toolkit (the core delivery mechanism of our analytical products to property professionals), Lightstone’s EZVal (a home valuation workflow platform used by our banking clients) and Lightstone’s Signio (an online platform that facilitates finance and insurance transactions in motor dealerships) are examples.
It’s important we use data that’s reliable, accurate, complete, relevant and timely to effectively build algorithmic “superpowers”. Managing data is critical, whether from external partners or generated internally, and it is important to build tools which track the quality of those data sources and the impact it has on the outcomes of decisions our clients make when using our analytical products.
The power of predictive and prescriptive analytics
Data improves operational efficiency and helps businesses make better decisions, and advanced AI processes and tools help our users move from merely understanding future outcomes (predictive analytics) to being recommended the most optimal course of action to take (prescriptive analytics).
Machine learning technology helps optimize decisions in “real time”. A finance and insurance representative in the automotive industry needs to offer value-added products such as an extended motor warranty or roadside assistance, to a customer on the dealership floor, at the point when they are processing the transaction and cannot wait for the algorithm to play catch-up as new information is captured.
The predictive and prescriptive technology is created from the accumulation and integration of historical with real-time data to predict with mathematical accuracy the likely outcome of a given decision. In this example, it could include the data of past value-added product sales, and real time application-specific data which is integrated with our comprehensive vehicle and property databases as well as other usable data sources. All of this data is used to train an algorithm to ultimately recommend the best course of action to our system users in real time.
We are now piloting algorithms to predict not only what’s going to be the most valuable for the dealership to sell, but also what’s going to be the most useful for the end-consumer. We want our dealer clients to not just make additional money during the sale of the vehicle, but also to delight their customers for the duration of their vehicle ownership journey.
In the property space, Lightstone uses data and algorithms to predict which homeowners are more likely to want to sell, and this helps the estate agent focus on those homeowners who are better prospects. Combined with our Artificially Intelligent Automated Valuation Model (AiVM) we can also help the Estate Agent determine an optimal listing price. For example, we can estimate how long a property would be on the market if it is listed 10% above our valuation model price or 10% below, so they can make informed decisions based on their client’s specific need.
Lightstone’s first analytical product was an automated valuation model of homes which evaluated past sales of properties in an area and the spatial aspects of comparable sales around that property. The product is constantly updated, and the latest version, AIVM, uses artificial intelligence and machine learning to provide even more accurate property valuations, and uses new data sources, including satellite imagery of the size of the property.
We combine visual and transactional data to algorithmically determine the value of the property. The AIVM saves time and money, cutting mortgage decision time from days (when a bank sends a human valuer to conduct a physical valuation) to within an hour or two for automated valuations. This unlocks immense operational efficiency, and a large part of that saving is passed on to the home buyer.
But even when a valuer still needs to carry out a physical valuation, it’s done using our data and information to make the physical valuations more accurate. So it’s not just a “superpower” for the bank, it becomes a “superpower” for the entire industry, and those using our systems get their job done quicker and with greater accuracy.
All about finding a competitive edge
Gathering and leveraging data is commonplace today, and that on its own is not enough to deliver a competitive advantage anymore.
If the full potential of “the new oil” is to be realised and competitive advantage delivered, data must be protected, used wisely, shared appropriately and utilised intelligently to solve problems and unlock opportunities in the real world.
Importantly, the competitive value of data depends on the underlying integrity and usefulness of information, the integration of diverse sources of complementary information, the technology which gathers and processes this information, and the quality of the data science deployed to deliver the insights and predictive algorithms.
This process of finding, refining, layering and interpreting data provides the insights which, if acted on correctly, can deliver transformative competitive “superpowers” to the industries we operate in.