Retail and gambling businesses rely on their data to connect with their customers and remain competitive, but as data volumes grow, how can they continue to use that data effectively?


Data is essential to the machinations of the retail and gambling sectors, from how they manage their customers through to back-office operations and their observation of industry regulations.

The merging of data-generating channels has enabled these businesses to cross-correlate data from numerous sources, providing unprecedented levels of insight. Data can be analysed in real-time to ensure online errors are corrected or friction reduced, and every customer engagement or touchpoint with the business can be tracked and demographics used to accurately predict behaviour and improve the customer experience.

But for these data-heavy industries, how you manage this wealth of data can create problems of its own. How do you ensure you’re collecting the right data and your systems aren’t getting swamped? Are you able to separate the rich data from the chatter? How do you ensure it can get from siloed source systems to where it needs to be in an intelligible format? Before we answer these and other questions, let’s look at the nuances of each sector.

Retail rewards

Retail data comes from a variety of sources, such as the website and mobile apps, point of sale systems and loyalty schemes and even social media channels and, when subjected to predictive analysis, can be used to inform marketing drives and promotions. Behind the scenes, inventory and order management system data (e.g., from SKU (stock keeping unit) barcodes) allows the retailer to apply dynamic pricing intelligently, monitor warehouse costs and stock inventory. And the data deluge is set to increase, with the Internet of Things (IoT) now adding feeds from devices in the home and instore.

Competition in retail is intense, and omnichannel retailers are now highly focused on ‘moments of truth’, which are the touchpoints that define a customer’s opinion about a specific business. To create positive moments of truth, they need to track the customer journey from start to finish and determine the regulatory of engagement or RFM (Recency Frequency Monetary) calculation to optimise and personalise the customer experience.

A great example of utilising data derived from customer touchpoints are loyalty schemes which are used to boost customer retention. The average e-commerce business spends 90% of its marketing budget on customer acquisition, so it’s important to make that investment count. The Customer Retention Rate (CCR) is a vital metric that establishes the number of customers before, during and after a marketing drive and is used to segment and measure the success of different retail strategies. Harvard Business School claims increasing CRR by 5% can boost revenue by between 25% and 95%.

Of course, with personal data comes responsibility and retailers must contend with protecting the personally identifiable data housed in their loyalty schemes and also comply with payment card data industry (PCI) regulations. These regulations can be region-specific. We recently assisted a German retailer with Schrems II compliance, for example, as well as PCI DSS, helping them obfuscate and encrypt their data.

Gambling gains

In the gambling sector, it’s all about using data to manage risk. They need to make sure their models are accurate and that they use data to adjust the acceptable parameters of risk, and the odds or bets they receive, in order to protect the business. But they also need to be aware of their customer demographics and nurture their customer base.

Online casinos track how the customer moves around the site, how they engage with other customers, their betting patterns and spend. It’s all about improving gaming experiences, perhaps through adding a new element to a game and using digital marketing to retain and excite their user base. They need to be able to marry up demographics with marketing to work out what works and what doesn’t. But they also need to track website traffic to be alert to issues as any downtime can severely compromise their income.

For betting providers, data is mainly used to improve accuracy and develop new betting opportunities that have high stability. Real-time data analysis is key so that the odds can be changed in accordance with what is happening, so they need live data that can be processed quickly.

The amount of real-time data being generated has increased dramatically. Ten years ago, a football match might have generated attack/defence statistics, yet now there’s a wealth of data on possession, attempts on a per-player basis, and this is only going to increase due to innovations such as wearable tech giving us live player information, so providers need to be able to make sure they are only analysing rich data that could influence the outcome.

Compliance regulations are numerous and complex, so gambling providers need to ensure they are adequately protecting the privacy of their customers by redacting customer data and encrypting personally identifiable information (PII) in line with data privacy laws and PCI requirements. Swisslos, for example, which runs multiple national and transnational lotteries, including the Euro Millions, uses the Splunk data analysis platform to gain real-time insights for compliance purposes and provides the authorities with direct access to their data processes and dashboards to demonstrate compliance.

Increasing stickiness with gamification 

Straddling the divide between retail and gambling is, of course, gamification. It’s now widely used to both motivate and engage customers by providing them with competitive games that allow them to gain points, progress through levels and win rewards. Some also see it as the natural successor to loyalty schemes, which are now reaching saturation point.

Gamification can generate valuable data on customers, particularly when combined with data from social media platforms if these are used as the sign-in mechanism, and the data provided can be used to compare cohorts and help with segmentation. But gamification is notoriously difficult to do well, with many projects doomed to fail simply due to poor design and execution. Users of gamification will expect instant results, so the provider must be able to process and analyse data speedily. They’ll also need to generate insights quickly to measure what’s happening and turn the results into actionable intelligence for the business. However, many expect gamification to emerge from the ‘trough of disillusionment’ forecast by Gartner back in 2012 as the concept matures and other technologies such as AI and AR increase its allure.

Data dangers

The danger for data-heavy industries is that all this data can be overwhelming. Retailers have a wealth of data, but it sits in siloed systems and needs to be cross-linked and correlated to provide an accurate picture, while gambling businesses need fast access to data to give them actionable insights. These demands can lead to the following problems:

  1. Wrongly interpreted data – for accurate customer profiling to occur, the data needs to be reliable and accurately allocated against that person. Get it wrong because the customer made a one-off purchase for their grandma and the personalised marketing drive goes to waste. Or fail to predict the odds have changed on a sports game, and your business could lose big. So rather than assuming all data is valid and churning through it, you need to identify what’s important, e.g. the rich data, before sending it to a data analysis environment.
  2. Getting to the data – these businesses have numerous disparate systems with the data recorded in different formats, so they need to be able to access and transform it so that it can be read on the destination system.ETL (extract, transform and load) technology can be used to perform this, but it typically introduces a lot of latency which then creates a delay between touch and insight. To minimise latency, you need to be able to transform the data into a compatible format between the two endpoints, e.g. the source and the data analysis platform, while the data is in transit.
  3. Meeting compliance requirements – working in these industries can see your business subject to multiple compliance obligations, requiring processes to be put in place for the handling and protection of sensitive PII. This could necessitate extensive changes to the way you process data. However, you can comply by obfuscating and encrypting the data within the data pipeline, avoiding the need to amend your legacy systems or change how you store the data.

Using an Event Stream Processor (ESP) such as Cribl LogStream can enable you to overcome these challenges because it acts on the data dynamically during transit. Data can be reduced by using techniques such as deduplication and removing events with null fields ensuring you focus on the metrics that matter and shortening the time to insight. Data can also be transformed in-flight and converted into a format that is readable, and it can be routed to multiple destinations so that the original dataset is sent to object storage, while the rich data is expedited for analysis.

If you’re in retail, gambling, marketing or gamification and you’d like to find out how Cribl LogStream can cost-effectively streamline your data analysis, contact us for a one-to-one consultation and demonstration.