Organizations today face overwhelming amounts of data, rapidly changing customer behaviours, and increased competitive pressure. Simultaneously, the explosion in data and digital technologies has opened up an unprecedented array of insights into customer needs and behaviours.Effective use of Big Data analytics increase productivity rates and profitability by 5-6%. McKinsey analysis of more than 250 engagements over five years has shown that those companies that keep data at the centre of the marketing and sales decisions improve their MROI by 15-20%. Thisadds up to $150-200 billion of additional value based on global annual marketing spend of an estimated $1 trillion.
New context of customer data:
Big data refers to the ever-increasing data in terms of volume, variety, and complexity that is being generated in today’s digital eco-system. Figure 1 shows that, big data sets are generated around customers based on their online purchases, web clicks, social media activities, smart connected devices, geo-location etc. Big data analytics can help marketers to mine and analyse these data, and help them discover hidden patterns, e.g., the way different consumer groups interact and how this affects purchase decisions. Equipped with these insights, companies can then develop targeted marketing campaigns that cater to the customer’s individual preferences.
Figure 1: Next generation customer profile. (http://asystec.ie/analytics-by-industry/)
- Next best action to engage customers is a customer-centric marketing approach that considers potential offers for individual customers and then determines the optimal one. NBA offer is determined by the customer’s interests and needs as well as by the marketing organization’s business objectives, policies and regulations. This is enabled through real-time decision technology that leverages call centre data, transaction data, customer information and a set of business rules to determine the offers for which the customer is eligible at that moment, and they are prioritised and optimised to provide the best offer to customers.
Figure 2: Accenture’s next generation analytics conceptual technical architecture
- Personalization of online shopping:Today, online retailers use powerful big data systems to gather information about user preferences, user browsing and purchasing behaviour, product attributes, geographic location of purchases, inventory levels, active promotions and campaigns. These data sources are converted into information and insights by intelligent machine-learning algorithms, which identify customer interests and product affinities, trace geographic peculiarities and identify seasonal effects among others.
- Monetizing big data for targeted dynamic advertisement:Data monetization creates opportunities for organizations with significant data volumes to leverage untapped or under-tapped information and create new sources of revenue. The volume and richness of data now accessible to the mobile providers in the form of transactions, inquiries, text messages or tweets, GPS locations or live video feeds. This offers a veritable gold mine of insights and applications.
- Machine-to-machine analytics to improve product life-cycle management: There has been a tremendous advancement in sensor technology which has led to the generation of machine-to-machine data at an unprecedented rate and in real time. Companies can use the data emitted by sensors to improve the efficiency of manufacturing processes, predict device failures and identify opportunities to upsell new products to customer. The data can also provide insights for product development, customer support and sales teams who use the information to improve product features, increase revenues and lower costs.
Why big data matters to marketing:
- Customer engagement: Delivers insight into your customers
- Customer retention and loyalty: Helps you discover what influences customer loyalty and what makes them come back for purchasing.
- Marketing optimization/performance: Helps you determine the optimal marketing expenditureinvarious channels, as well as continuously optimize marketing programs through testing, measurement and analysis.
Three types of big data that are crucial for marketing purposes:
Customer: This includes behavioural, attitudinal and transactional metrics from marketing campaigns, POS, websites, customer surveys, social media, online communities and loyalty programs.
Operational: This includes objective metrics that measure the quality of marketing processes relating to asset management, marketing operations, resource allocation, budgetary controls, etc.
Financial: This includes revenue, profits, sales and other objective data-types that estimate the financial condition of the organizations.
Five ways big data can rock our world in future:
- Large enterprises will be the first to widely adopt big data and predictive analytics technologies, followed by small and medium businesses.
- Marketing spend will become substantially more accurate by leveraging the big-data insights to accurately target consumers and deploy account-based marketing strategies.
- Salespeople will slowly adopt data-driven methodsin order to target high-value prospects, retainthe existing customers, and increase existing opportunities.
- Sales forecasting accuracy will improve dramatically with the help of sophisticated algorithms
- Real-time sales data visualization technologies will emerge, thus empowering sales managers to adjust their tactics.
Strategies to Succeed:
- Use analytics to identify valuable opportunities: Analytics leaders take time to develop “destination thinking” which is the business problems they want to solve. These need looking beyond broad objectives of expanding wallet share and get down to a level of meaningful specificity.
- Start with the consumer decision journey:Understanding consumer decision journey is critical to identifying battlegrounds to either gain new customers or preventexisting customers from defecting to competitors. 35% of B2B pre-purchase activities are now digital, which indicates B2B companies should invest in websites.
- Keep it fast and simple: Companiesshould spend on “algorithmic marketing,” as it can process tremendous amount of data through a “self-learning” process to create better and relevant consumer interaction. These systems automatically track the key words and update every fifteen seconds, based on the usage of search terms, ad costs or customer behaviour. It can make real-time price changes across products based on customer preference, price comparisons, inventory and predictive analysis.
- What should our big data analytics roadmap look like to achieve marketing objectives?
- What business outcomes would we like to influence by leveraging big data around customers?
- What capabilities and services should we develop by leveraging big data that enable a strong competitive advantage?
- What technology options will enable big data analytics journey?
- Are appropriate skills and resources available in-house to embark on the big data journey?
- Knowing what data to gather:You have enormous volumes of customer data but it has to be the right data.
- Knowing which analytical tool to apply:As the data volume grows, time available for decision making is shrinking, hence choice of tools becomes critical.
- Knowing how to go from data to insight to impact:Once you have the data, how do you turn it into insight relevant for business?
1. Sampling Problem, biased data:
Accessible user-level data is not representative ofthe whole targetsegment. Hence there is inherent risk in assuming that insights from the data are applicable for all the consumers.
2. Data execution exists in selected channels:
Certain marketing channels are well suited for applying user-level data but in many channels, it is difficult to apply user data directly to execution. For premium display and offline channels, user-level data cannot be applied to execution at all.
3. User-level results can’t be presented directly:
User-level data tends to be incomprehensible except to domain experts. It needs to be converted to a property-level or daily segment-level for the results to be consumable at large.
4. Algorithms have difficulty answering “Why”:
Algorithmic analysis of data can result in predictions and recommendations but these tend to have difficulty answering questions like why should we move spending?
5. Not suited for producing learning:
Actionable learning that require user-level data like applying a look-alike model to discover previously untapped customer segments, are relatively few and require tons of effort to uncover.
6. Vulnerability to Noise:
User-level data tends to be filled with noise and potentially misleading insights, henceit needs to be filtered to get accurate results.
7. Accessibility Issue:
Due to security reasons, user data cannot be made accessible to everyone andit shouldbe carefully transferred from server to server, as everyone doesn’t have the competency to interpret big data.
Big data for better marketing:
- Use big data to gain deeper insight which can help develop specific strategies to drive growth.
- Big data insights should be provided to those who can use it, like the CMOs, front-line store managers, call-centre phone staffs, sales associates, etc.
- Big data can be overwhelming, so begin by focusing on what outcomes would you like to improve? Accordingly identify what data you would need.
Big data marketing solutions from SAS:
- Adaptive Customer Experience:Create a complete profile of the customer based on offline and online data for more targeted, personalized communications.
- Marketing Automation: Get more campaigns out the door in an automated, trackable and highly repeatable fashion.
- Marketing Optimization: Extract the highest benefit arising out of each customer contact by determining how business variables will be affecting outcomes.
This goldmine of data is a central decision point for marketing and sales leaders. Those able to drive above-market growth, though, are the ones who can effectively mine that gold.