Have you ever wondered how Netflix ‘knows’ which movies to recommend? Or how does the entertainment giant choose themes for some of the most popular productions in the film industry?
Data. Data. And some artificial intelligence (AI) – that would be the answer in a nutshell.
Data lies at the heart of everything Netflix does. Its customer-centricity is underpinned by the information it gathers on individual users’ viewing habits and preferences which collectively help it to plan its broader production pipeline and experiment with new user features.
Retail can draw many learnings from the way Netflix utilises data and the last couple of years full of turbulent changes have shown that retailers need to up their data game in order to keep up with the ever-changing customer needs.
Using practical examples, Jai Gandhi, VP of data and analytics at Ciklum, explores three main areas of improvement retailers should be thinking about.
Collecting data, unobtrusively
Retailers need to know their customers in order to present the most relevant offering in their physical and online stores.
What do they buy? When and where do they shop? What offers appeal to them the most? These are just some of the fundamental questions that the savvy collection and analysis of data can answer, and Netflix is a master at doing it.
The vast majority of Netflix users will not recognise how it uses algorithms and data collection techniques behind the scenes – it learns from its customers’ everyday use of the platform, tracking what is searched, watched and reviewed, and at what times.
Retailers can do the same. On online platforms such as webshops, apps and social media, information can be collected from personal accounts, while basket spend (and abandoned baskets) can also be monitored.
In bricks and mortar settings, data collection is not as straightforward but is still abundantly possible.
As a starting point, retailers can use payment data to gain a breakdown of what a customer buys in-store and online – information which could inform what products to prioritise for valuable shop floor space.
Sensors, cameras and smartphones can also serve as nodes of information for physical stores.
From scanning what customers browse and the direction they move around the store to the frequency and time of visits, there is a lot of valuable data that can be extracted by clever use of AI and Computer Vision (CV) technology. CV, in essence, enables devices to ‘see’ by detecting and classifying the content of digital images such as photographs, videos and live camera feeds from stores.
Loyalty cards are a proven way of collecting data from both online and physical retail settings. From supermarkets to fashion brands, many data-driven retail firms rely on these to capture transaction and behavioural data.
The analytical tools to analyse such data are also becoming more powerful. This is critical, as a failure to properly interpret information collected could render data collection efforts as a fruitless exercise.
Smart analytics enables retailers to provide highly personalised offers, product recommendations and multimedia content to their customers, which brings us onto the second key learning.
Personalising customer experience
Watch It Again, My List, Top Picks… these are all components of every Netflix user’s dashboard that are unique to the individual in question.
The content that populates these lists is determined by the data collected and analysed, and explains why the subscription platform commands an enviably low churn rate – indeed, Netflix continues to add to its overall subscriber numbers after the pandemic surge in uptake, suggesting those new additions have been retained.
Personalisation of the customer experience has also been high up on the agenda of retail businesses in recent years, a priority which has been exacerbated by COVID-19 and the shift from physical to online shopping as a result of the societal restrictions brought in last year.
Amazon is perhaps the most notable example of why personalisation pays dividends. Around 35% of its revenues are derived via its personalised recommendation tool, an engine powered by user data which can deliver huge boosts to revenues even with minor algorithmic improvements.
In the fashion space, ASOS makes use of machine learning and a process known as collaborative filtering. Put simply, this allows the online retailer to find customers with similar purchasing habits and recommend products that other like-minded shoppers have already bought.
This, along with other personalisation techniques such as automated sizing recommendations, has helped to propel ASOS into one of the most trusted online-only fashion retailers.
Data = the building blocks of products
ASOS is also highly successful because it stocks products that its customers want to buy.
The most simple of retail concepts, it is more important than ever to nail product offerings given the enormity of choice consumers are faced with both in-store and especially online.
Netflix, once again, has also mastered this fundamental. Thanks to its detailed collective understanding of its subscriber base, it can strategize its content calendars with a high degree of confidence, and its track record of releasing highly-acclaimed shows is no coincidence – The Queen’s Gambit, for example, was one of the most watched series in 2020.
Retailers should also develop products based on the data gathered from their customers. Larger retailers with multiple bricks and mortar stores already do this by studying data on localised demographics. This informs what products are most likely to be popular among their catchment of consumers and can therefore optimise inventory levels, signpost likely successful promotions and, ultimately, generate higher sales.
Other, more obvious data gathering can also inform retailers on what products to develop and/or sell. Social media polling and surveys tied into loyalty schemes are cost-effective and simple ways of finding out what your customers want to buy.
Netflix also leverages A/B testing to trial new features, the data generated from user trials informing what new capabilities and upgrades to deploy. This is another technique retailers can use to hone their offering – for example, by testing two versions of different webpages on a selection of customers to see which one yields the best results.
Still not convinced?
Let’s just recap here. Data collected and analysed means retailers can better understand their customers, and that’s both online and in-store. It means they can personalise experiences and have more happy customers. Retailers can not only offer relevant products, they can build data-driven products that will be highly demanded.
Translate this into a retail leader’s language and we have lower churn, higher loyalty and customer happiness, and eventually increase in sales.
By Jai Gandhi, VP of data and analytics at Ciklum