So now we're going to look at the needs for analytics among different business functions and some examples of each. First, let's take a look at sales and marketing. For sales and marketing, what you really want to use analytics for is to better understand customer preferences, to identify trends, to find new prospects or customers, or enlarge your pipeline. You also can use it to identify new markets to streamline the sales process. Ultimately, the goal of sales and marketing analytics is to drive revenue. For example, Westpac, one of Australia's major banks, had an issue with not having a 360 degree view of its customer across all touchpoints and services it offers, such as ATMs, in-banking services, mortgages, loans, etc. This severely limited the bank's ability to target market customers. By analyzing all of this usage data by customer, they were able to increase target offerings from just 1% of customers to more that 25% of their customers, yielding over, I think, $22 million in just the first nine months. As in the previous example, we can see that customers often interact with our companies in a wide variety of ways. So it's difficult to ensure that they're happy or at least satisfied with all the points of interaction. Therefore, we can use analytics to better understand customer experience, measure customer satisfaction, identify instances of potential customer dissatisfaction. We can also use analytics to improve the information available to customer service agents and streamline the customer service process. Ultimately, it's about customer retention and company reputation. Take Burberry for example, they wanted to improve the customer in-store experience through greater customer intimacy and interaction. So they integrated and analyzed data from 800,000 followers and 15 million fans on Instagram and Facebook, including customers' Twitter posts, their purchase history and surveys, and data pulled from RFID tags on clothing. By doing all of this, they were able to identify customers the moment they walk in the store and greet them by asking about previous purchases. Sales assistants using iPads can make on the spot recommendations, including accessories for a previously purchased item. And when a customer tries on an item, RFID tags trigger interactive videos with product details. Now that's customer experience. Analytics can also be used to define new products or services. For example, it can be used to understand unanticipated product or service usage, or identify market gaps or white space. It can also be used to identify new types of raw materials that can be converted into products, or improve the information available to designers and engineers. And it can be used the streamline the entire new product introduction, or NPI, process. Ultimately, it's about identifying new sources of revenue. Let's take AM Biotech for example, they wanted to improve the antiviral efficacy of antibodies by developing synthetic antibodies. So turning to one of the most interesting and oldest data sources, DNA, they analyzed tens of billions of short DNA sequences in the creation of customized what are called x-aptamers. The resulting synthetic antibodies enabled a broader range of diagnostic and therapeutic uses than antibodies. They're faster to develop and higher quality than regular antibodies. They have greater shelf life and ease of handling than regular antibodies. And they have no inherent immune responses. Now let's look at financial analytics. Financial analytics typically is used to understand the impact of financials on the business, or identify instances of fraud, closer to or where they actually occur. It's also used to identify new leading financial indicators for improved forecasting. Improve the information available to accountants, controllers, investors, or auditors. And to streamline the financial reporting process. The ultimate goal, for the most part, for financial analytics is to maximize the return on investment, or return on resources. Ahold is a $38 billion euro international retailer. They wanted to improve the inventory accuracy and vendor relationships for its grocery stores. So they analyzed 5 million lines of transaction data per day of receiving, inventory, and sales data from 800 stores and 300,000 SKUs by store, product, color, weight, packaging, etc. By doing pattern analysis on this data, they were able to detect issues with shipment integrity and other issues. One of the things they detected was that one vendor was charging for an expensive product while delivering a cheaper product. They were able to recover from this $160,000. And the worst division in terms of shrinkage showed the highest improvement. Human resource analytics endeavors to understand the motivation and habits of people and groups. It can also be used to identify opportunities to improve people's capabilities, and identify opportunities to automate repetitive tasks. HR analytics can also identify workplace issues, and improve information available to employees to help streamline the talent acquisition, evaluation, and other HR processes. Ultimately, HR analytics is about optimizing the availability and productivity of people and teams. Take for example Lockheed Martin, the major defense contractor that makes the F-16 and the new F-35 among other aircraft. They wanted to more proactively predict the health of thousands of programs or projects that they have running. They did this to enable program managers to apply course correction more proactively and less reactively. So what they did was they correlated and analyzed hundreds of metrics for thousands of programs to identify the leading indicators of program performance. As you can imagine, they run very large complex programs to build defense equipment. They capitalized on both structured and unstructured program metrics, and uncovered specific words from a program manager's comments that are predictors of a program downgrade. As a result, they were able to identify a concise set of attributes that are predictors of program performance. In doing so, they increased program foresight by 3X, enabling earlier program assessment, which save them hundreds of millions of dollars in program delays. Supply chain analytics could help organizations understand all of the moving parts and processes of the supply chain. It can model the supply network down to multiple levels of suppliers. It can analyze and prevent systemic or ad-hoc supply chain issues. It can identify opportunities to improve, eliminate or outsource processes. And it can improve information availability to supply chain processes and personnel. Ultimately, supply chain analytics is for ensuring materials and goods are where they need to be and when they're needed. Take NCR, the maker of ATMs, cash registers, and the like, for example. They were dealing with an issue called SKU proliferation, meaning too many varieties of a product, which created an increasing product complexity that affected sale cycles. As a result, there was a lack of alignment between customer demand and solutions management, operations, and sales. So they analyzed product, sales, and inventory data across tens of thousands of configuration options and across the entire supply chain. The pattern analysis they did optimized product configuration and helped shape demand, leading to $110 million of contribution to revenues and a 5% increase in sales efficiency. Even IT folks themselves can get in on the analytics action. They can use analytics to help understand and enable flexible, dynamic capacity and performance, to enable the cooperation of disparate technologies. To analyze and limit system downtime, to protect systems from hacking, to identify opportunities to improve, eliminate or outsource technologies, also environmental concerns too. And they can correlate technology metrics with business metrics. Ultimately, organizations can use IT and operational technology analytics to ensure technology efficiency and availability. Take for example Tesco, a major UK retailer. They wanted to predict acid failure in refrigeration and improve the energy efficiency in over 4,500 supermarkets. So they created a system that collected over 200,000 refrigeration-related data points across all of their stores, including compressor pressure, liquid levels, temperatures, energy usage, currents, etc. In all, they collected 70 million data points from readings every three minutes. The system enabled them to save 20% of their energy costs, or 20 million euros annually, and enabled them to reduce maintenance by 40% by proactively addressing imminent refrigeration problems.