HOW SMART USE OF DATA CAN TURBOCHARGE YOUR NPD

Innovation is the lifeblood of the food and drink industry. A successful new product launch can add millions to the bottom line, excite existing customers and attract new customers to the brand. But how can your innovation succeed where others fail? And how can you identify and then capitalise on the latest food and drink trends through your New Product Development (NPD) before your competitors saturate the market?

The answer, to a large degree, lies in data. Today’s businesses are flooded with data from all kinds of different sources; but not all of this data is properly captured, and rarely is it turned into the kind of actionable insight that can unearth the latest NPD opportunity.

Cutting-edge, Artificial Intelligence (AI) technology, the ability of a machine to think and learn, is changing that. For example, AI is used in the pharmaceutical industry to crunch and analyse many sources of unstructured data (such as doctors’ notes, images, histories, formulae and genetics) to reveal new insights and unknowns which are used to develop new medicines. Using these same techniques, CPG companies can mine unstructured data sources such as social media posts and influencer blogs, and turn this raw, messy data into valuable insights to uncover the latest food and drink trends. Early access to such insights will enable them to deliver game-changing NPD before these trends reach the mainstream.

In this whitepaper we reveal:

■ The need-to-know trends driving food and drink innovation

■ The limitations with traditional approaches to trend analysis

■ How the latest AI technology is unlocking the insights that will drive NPD success

■ What the data tells us about key future food and drink trends

01 | WNS.COM

INTRODUCTION

WNS.COM | 02

Food and drink is powered by innovation. For brands and retailers, the development of new concepts, products, flavours and formats is crucial to driving growth, maintaining consumer interest and attracting new shoppers. Without innovation, the U.K. grocery sector would be a much poorer place.

The importance of innovation to the industry is growing too. In 2018, the total value of Fast-moving Consumer Goods (FMCG) that had been on shelf for a year or less grew by GBP 541 Million1 compared to the previous year. New products accounted for 13.4 percent of the total value of the FMCG market, up from 12.6 percent in 20172 and 12 percent in 20163.

Launching a new product is one thing. Making it stick is another. Of the 3,375 products launched in 2018, just 43 (13 percent) achieved sales in excess of GBP 5 Million4. The vast majority (53.4 percent) of new launches did not exceed sales of GBP 100K. A staggering 76 percent5 of new food and drink products fail within a year resulting in the loss of millions of pounds invested in research, production and marketing.

Success, of course, depends on myriad factors: securing the right channels, establishing reliable chains of supply, determining the correct price point, and offering the right promotional and marketing mix are all crucial. But this all has to begin with the fundamentals: developing a product, identifying the target audience, and establishing what will motivate shoppers to buy it.

This requires intelligent, actionable insight. As a brand with designs on launching the next big thing, you need to be able to spot an emerging trend before it becomes mainstream. You should be able to distinguish a fad from a movement that will shape the future of a category. And you need to be able to do all this before the competition.

To this end, smart analysis of data is crucial. In later chapters, we will explore in detail how AI is transforming the innovation process. But what are the trends driving the latest grocery innovations? And what attributes do they share in common?

03 | WNS.COM

CHAPTER ONE: ASSESSING THE NPD OPPORTUNITY

Plant PowerThe most influential trend of the past year has arguably been the rise of plant-based products, a sector that barely existed a few years ago. Sales hit GBP 352.1 Million in 2018, having grown 14.3 percent year on year6. This boom in plant-based eating is largely being driven by the rise of a lifestyle that didn’t exist a few years ago either: flexitarianism. That is, a largely vegetarian or vegan diet that’s occasionally supplemented with meat, fish or dairy. A recent study7 found that 21 percent of Britons now identify as flexitarian. A further 12.5 percent are either vegetarian or vegan, meaning a third of the population are either eating less meat or have cut it out altogether. There were 600,000 vegans in Britain in 2018, up from 150,000 four years earlier8.

Ethical ShoppersSmart operators have identified what’s driving this dietary revolution. Mounting evidence that meat and dairy production

importance of spotting trends early and acting quickly. It also shows why it is paramount to understand what motivates shoppers to put new products in their baskets. These motivations are forever in flux. The growing attention brands and retailers are paying to fibre content and gut health is further proof of this.

Cereal brands such as Kellogg’s Fruit & Fibre and All Bran have long marketed themselves on their fibre content, and how a fibre-rich diet promotes healthy heart function. But in April 2019, Kellogg’s changed tack, rebranding Bran Flakes, All Bran and Fruit & Fibre as its Happy Guts range, a nod to the fact fibre encourages good bacteria to grow in the gut, promoting better gut health.

Gut Health GrowsSurprisingly, some of the inspiration for this rebrand can be found in the soft drinks market. Gut health orientated drinks are booming as consumer awareness of the wider health benefits of keeping your gut happy grows. U.K. sales of fermented milk drink kefir almost tripled in 2019 to hit GBP 23 Million13, driven by a new generation of dairy drinks from brands such as The Collective and Biotiful.

Fermented, naturally sparkling tea drink kombucha, which contains

gut health boosting live bacteria, is another hot ticket, with brands such as Equinox, Lo Bros, Rude Health and others winning growing chiller space in supermarkets and food service outlets. One recent report predicted that global kombucha sales would hit USD 3.5 Billion by 2025, representing a CAGR of 13 percent.

Sugar as a VirtueThe great expectations many have for kombucha is proof there’s still space for new products with relatively high sugar levels, in spite of the imposition of the soft drinks levy in 2018 and the wider backlash against sugar that’s benefitted brands such as Coke Zero Sugar and Grenade Carb Killa. Kombucha requires the addition of some sugar to keep the all-important probiotic bacteria in it alive.

Some soft drinks brands are even making a virtue out of their sugar content. Karma Cola, for example, boasts Fairtrade organic free trade cane sugar. Other newcomers,

are far more environmentally damaging than production of plant-based alternatives is sinking in. Seventy percent of consumers say concern over the impact food production has on the planet is the key reason they have changed or are considering changing their diet9.

This explains why oat milk brand Oatly, which hit sales of GBP 29.6 Million in 201910 after just six years on the market, now communicates on pack how much C02 is produced by the contents of every carton it sells (per litre, oat milk is responsible for less than a third of the emissions of cows’ milk11).

Green BrandsAlpro, which has successfully expanded from plant-based milks into pot desserts and ice cream in recent years, also makes much of its environmental benefits in its marketing. The brand’s sales grew 11.6 percent to GBP 215.4 Million in 201912.

Then there’s new meat alternative brands such as Eden and Birds Eye’s Green Cuisine, which explicitly or implicitly allude to the environmental benefits of a plant-based diet in their packaging and marketing.

Beyond Meat, Meatless Farm and Equals, meanwhile, offer burgers and other meat substitutes that are clearly positioned at flexitarians who miss the taste and texture of the real thing. For more proof of this trend, look no further than the rise of jackfruit during the 2019 barbecue season.

Rethinking FibreThe success of these launches doesn’t just illustrate the

WNS.COM | 04

including Dalston’s Soda Co and Gunna, market their drinks not on how little sugar they contain (as has become the norm among mainstream soft drinks) but on their flavour and use of natural ingredients.

Why Insight MattersSuch tactics demonstrate great insight, capitalising on the growing number of Britons who are drinking less alcohol, but are still looking for premium alcohol-free offerings. One recent study found that 80 percent of pubgoers would like to see a greater range of craft soft drinks with natural, environmentally sound and great tasting credentials14.

What the above examples have in spades is a well-defined proposition built on great consumer insight. New launches

can encounter trouble when either of these attributes is missing. ‘Natural’ cola Coke Life was axed in 2017 after just three years on the market when it was revealed consumers did not understand the variant contained less sugar than standard Coke. April 2019 saw Nestlé pull the plug on ‘vegetarian’ range Garden Gourmet after eight months on shelf with the food giant admitting the range had not been as popular as it had hoped. Critics claimed the brand had not recognised consumers’ changing motivations for eating less meat.

All this proves why market insight is so crucial when it comes to developing new products. The next chapter will explore how structured, semi-structured and unstructured data can help provide that insight.

Plant PowerThe most influential trend of the past year has arguably been the rise of plant-based products, a sector that barely existed a few years ago. Sales hit GBP 352.1 Million in 2018, having grown 14.3 percent year on year6. This boom in plant-based eating is largely being driven by the rise of a lifestyle that didn’t exist a few years ago either: flexitarianism. That is, a largely vegetarian or vegan diet that’s occasionally supplemented with meat, fish or dairy. A recent study7 found that 21 percent of Britons now identify as flexitarian. A further 12.5 percent are either vegetarian or vegan, meaning a third of the population are either eating less meat or have cut it out altogether. There were 600,000 vegans in Britain in 2018, up from 150,000 four years earlier8.

Ethical ShoppersSmart operators have identified what’s driving this dietary revolution. Mounting evidence that meat and dairy production

importance of spotting trends early and acting quickly. It also shows why it is paramount to understand what motivates shoppers to put new products in their baskets. These motivations are forever in flux. The growing attention brands and retailers are paying to fibre content and gut health is further proof of this.

Cereal brands such as Kellogg’s Fruit & Fibre and All Bran have long marketed themselves on their fibre content, and how a fibre-rich diet promotes healthy heart function. But in April 2019, Kellogg’s changed tack, rebranding Bran Flakes, All Bran and Fruit & Fibre as its Happy Guts range, a nod to the fact fibre encourages good bacteria to grow in the gut, promoting better gut health.

Gut Health GrowsSurprisingly, some of the inspiration for this rebrand can be found in the soft drinks market. Gut health orientated drinks are booming as consumer awareness of the wider health benefits of keeping your gut happy grows. U.K. sales of fermented milk drink kefir almost tripled in 2019 to hit GBP 23 Million13, driven by a new generation of dairy drinks from brands such as The Collective and Biotiful.

Fermented, naturally sparkling tea drink kombucha, which contains

gut health boosting live bacteria, is another hot ticket, with brands such as Equinox, Lo Bros, Rude Health and others winning growing chiller space in supermarkets and food service outlets. One recent report predicted that global kombucha sales would hit USD 3.5 Billion by 2025, representing a CAGR of 13 percent.

Sugar as a VirtueThe great expectations many have for kombucha is proof there’s still space for new products with relatively high sugar levels, in spite of the imposition of the soft drinks levy in 2018 and the wider backlash against sugar that’s benefitted brands such as Coke Zero Sugar and Grenade Carb Killa. Kombucha requires the addition of some sugar to keep the all-important probiotic bacteria in it alive.

Some soft drinks brands are even making a virtue out of their sugar content. Karma Cola, for example, boasts Fairtrade organic free trade cane sugar. Other newcomers,

are far more environmentally damaging than production of plant-based alternatives is sinking in. Seventy percent of consumers say concern over the impact food production has on the planet is the key reason they have changed or are considering changing their diet9.

This explains why oat milk brand Oatly, which hit sales of GBP 29.6 Million in 201910 after just six years on the market, now communicates on pack how much C02 is produced by the contents of every carton it sells (per litre, oat milk is responsible for less than a third of the emissions of cows’ milk11).

Green BrandsAlpro, which has successfully expanded from plant-based milks into pot desserts and ice cream in recent years, also makes much of its environmental benefits in its marketing. The brand’s sales grew 11.6 percent to GBP 215.4 Million in 201912.

Then there’s new meat alternative brands such as Eden and Birds Eye’s Green Cuisine, which explicitly or implicitly allude to the environmental benefits of a plant-based diet in their packaging and marketing.

Beyond Meat, Meatless Farm and Equals, meanwhile, offer burgers and other meat substitutes that are clearly positioned at flexitarians who miss the taste and texture of the real thing. For more proof of this trend, look no further than the rise of jackfruit during the 2019 barbecue season.

Rethinking FibreThe success of these launches doesn’t just illustrate the

including Dalston’s Soda Co and Gunna, market their drinks not on how little sugar they contain (as has become the norm among mainstream soft drinks) but on their flavour and use of natural ingredients.

Why Insight MattersSuch tactics demonstrate great insight, capitalising on the growing number of Britons who are drinking less alcohol, but are still looking for premium alcohol-free offerings. One recent study found that 80 percent of pubgoers would like to see a greater range of craft soft drinks with natural, environmentally sound and great tasting credentials14.

What the above examples have in spades is a well-defined proposition built on great consumer insight. New launches

can encounter trouble when either of these attributes is missing. ‘Natural’ cola Coke Life was axed in 2017 after just three years on the market when it was revealed consumers did not understand the variant contained less sugar than standard Coke. April 2019 saw Nestlé pull the plug on ‘vegetarian’ range Garden Gourmet after eight months on shelf with the food giant admitting the range had not been as popular as it had hoped. Critics claimed the brand had not recognised consumers’ changing motivations for eating less meat.

All this proves why market insight is so crucial when it comes to developing new products. The next chapter will explore how structured, semi-structured and unstructured data can help provide that insight.

05 | WNS.COM

To understand something, you first need to measure it. Food and drink brands typically use structured, semi-structured and, increasingly, unstructured data to plan their innovation pipelines. Launching a product without having analysed a market and the size of the opportunity would be like trying to hit a bullseye blindfolded.

In this context, to innovate means to bring something new to a market. That brings risk. Doing your homework by analysing market data helps mitigate that risk, but if a product is genuinely new, there will be no comparable products already on the market. That creates a gap in your understanding. As we’re about to see, it’s now possible to fill that gap.

CHAPTER TWO: NAVIGATING THE DATA LANDSCAPE

WNS.COM | 06

Structured, Semi-structured and Unstructured DataEstablished means of assessing the size of an opportunity and how best to execute a launch still have great power. The most common forms of structured data used when assessing trends are documents such as invoices, orders, and spreadsheets expressing sales; results from consumer panels; market research; and so on. Spreadsheet columns, for instance, have headings, values are ordered, rows represent SKUs; in short, the data is structured.

An example of semi-structured data would be the results of a poll about a potential new launch. Let’s say you asked consumers to rate a new product from one to 10 and then asked them to explain their reasoning. Your results are both quantitative and qualitative. The values are structured; the reasons your respondents gave for their scores are not, although some structure can be applied to them. So, the data you have created is semi-structured.

Unstructured data can come in the form of text, images or sounds. This might take the form of technical reports, recordings from customer care lines, online reviews or menus. In the past, the gathering and analysis of such data was expensive and time-consuming – fashion industry executives, for example, would travel the world to spot the latest trends – but new technologies have created new opportunities to mine this unstructured data far more effectively.

In grocery, during the innovation

process, structured data is typically gathered in three key ways: by using Electronic Point of Sale (EPoS) data from large retail chains; by using demographically representative consumer panels to analyse factors such as purchasing habits, purchase motivations, age and geography; or through loyalty card schemes that allow retailers to spot trends in the shopping habits of their customers. All three have distinct strengths and weaknesses.

EPoS DataEPoS data allows product developers to assess the true size of a market in any given retailer that has shared their data. Pack size, price, promotional mechanic and so on can be analysed and the performance of retailers can be compared, allowing brand owners to identify the right channels, price and promotions before launching a product.

Of course, there are limitations. Not all retail chains will share their EPoS data, so this method will not always tell the whole story. There

are 18,000 unaffiliated independent grocery retailers operating in Britain15 – to source data from each individual store would be prohibitively time-consuming, so data providers extrapolate from representative panels of stores.

Crucially, if there’s not anything on the market like what you’re planning to launch, or if the market you’re entering is nascent and only catered for by a handful of independent retailers, EPoS will not give you all the information you require.

Panel and Loyalty Card DataPanel data is crucial for identifying factors such as motivations for purchase decisions, the occasions consumers are buying items for and the demographic makeup of different markets. If you’re looking to launch a plant-based milk, for example, panel data would allow you to identify the growing importance of strong environmental credentials in this market, as we saw in chapter one.

The difficulty is that panel data is not ‘real’ in the sense that results are taken from a sample of society and then extrapolated out. Even data from a panel of 40,000 consumers would have to be multiplied by approximately 1,500 before it is of comparable size to the British population. That means trends must reach a certain size before they become visible.

Retailers can use loyalty card data to profile consumers, offer tailored promotions and analyse trends in buying patterns and what motivates shoppers to switch from one product to the next. But, as with EPoS and panel data, its capability for predictive analysis is limited. It can only provide information about products that have already achieved scale.

Predictive AnalysisIn the past, structured and semi-structured sources required data to be entered and information to be extracted manually. Such processes required humans to apply context to the data and make predictions based on what it showed. The birth of Optical Character Readers (OCR) used in tandem with Robotic Process Automation (RPA) allowed businesses to reduce the manual process of computing large volumes of data. As we’ll see in the next chapter, AI will be the next frontier in this evolution in the field of unstructured data.

Conventional forms of structured and unstructured data can be said to answer the questions, what, who and why. EPoS data provides a historical record of goods sold, prices paid, promotional

mechanics used and so on. Panel data gives a further layer of detail, with crucial information on who is buying any given product and their motivations for doing so.

Data on investments in the food and drink industry can help answer the critical question: what’s next? The Coca-Cola Company’s recent USD 20 Million investment in kombucha brand Health-Ade; Molson Coors’ acquisition of Dr Kombuch; and a growing number of private equity investments in the growing adult soft drinks category all offer valuable insight into the direction the industry is heading as consumers cut back on alcohol and make health-based purchase decisions.

But the capability of structured investment data to predict the future is limited. Clearly, brands such as Health-Ade and Dr Kombucha had already established a following before

being targeted by their new owners. For manufacturers to get a real head start in product development, they need to be able to identify trends before they become mainstream and show up on the radar of conventional sources of structured and semi-structured data.

That’s where unstructured data comes in. It’s estimated there is nine times as much unstructured data than structured data available to businesses16. Not all of it is useful, of course. There’s a risk of developers being swamped by the sheer volume of data available. To exploit its value, they need to be able to navigate it quickly and efficiently and sift the useful from the trivial.

The good news is that with developments in AI, the tools businesses need to unlock the data that can turbocharge their NPD are now at their disposal.

07 | WNS.COM

Structured, Semi-structured and Unstructured DataEstablished means of assessing the size of an opportunity and how best to execute a launch still have great power. The most common forms of structured data used when assessing trends are documents such as invoices, orders, and spreadsheets expressing sales; results from consumer panels; market research; and so on. Spreadsheet columns, for instance, have headings, values are ordered, rows represent SKUs; in short, the data is structured.

An example of semi-structured data would be the results of a poll about a potential new launch. Let’s say you asked consumers to rate a new product from one to 10 and then asked them to explain their reasoning. Your results are both quantitative and qualitative. The values are structured; the reasons your respondents gave for their scores are not, although some structure can be applied to them. So, the data you have created is semi-structured.

Unstructured data can come in the form of text, images or sounds. This might take the form of technical reports, recordings from customer care lines, online reviews or menus. In the past, the gathering and analysis of such data was expensive and time-consuming – fashion industry executives, for example, would travel the world to spot the latest trends – but new technologies have created new opportunities to mine this unstructured data far more effectively.

In grocery, during the innovation

process, structured data is typically gathered in three key ways: by using Electronic Point of Sale (EPoS) data from large retail chains; by using demographically representative consumer panels to analyse factors such as purchasing habits, purchase motivations, age and geography; or through loyalty card schemes that allow retailers to spot trends in the shopping habits of their customers. All three have distinct strengths and weaknesses.

EPoS DataEPoS data allows product developers to assess the true size of a market in any given retailer that has shared their data. Pack size, price, promotional mechanic and so on can be analysed and the performance of retailers can be compared, allowing brand owners to identify the right channels, price and promotions before launching a product.

Of course, there are limitations. Not all retail chains will share their EPoS data, so this method will not always tell the whole story. There

are 18,000 unaffiliated independent grocery retailers operating in Britain15 – to source data from each individual store would be prohibitively time-consuming, so data providers extrapolate from representative panels of stores.

Crucially, if there’s not anything on the market like what you’re planning to launch, or if the market you’re entering is nascent and only catered for by a handful of independent retailers, EPoS will not give you all the information you require.

Panel and Loyalty Card DataPanel data is crucial for identifying factors such as motivations for purchase decisions, the occasions consumers are buying items for and the demographic makeup of different markets. If you’re looking to launch a plant-based milk, for example, panel data would allow you to identify the growing importance of strong environmental credentials in this market, as we saw in chapter one.

The difficulty is that panel data is not ‘real’ in the sense that results are taken from a sample of society and then extrapolated out. Even data from a panel of 40,000 consumers would have to be multiplied by approximately 1,500 before it is of comparable size to the British population. That means trends must reach a certain size before they become visible.

Retailers can use loyalty card data to profile consumers, offer tailored promotions and analyse trends in buying patterns and what motivates shoppers to switch from one product to the next. But, as with EPoS and panel data, its capability for predictive analysis is limited. It can only provide information about products that have already achieved scale.

Predictive AnalysisIn the past, structured and semi-structured sources required data to be entered and information to be extracted manually. Such processes required humans to apply context to the data and make predictions based on what it showed. The birth of Optical Character Readers (OCR) used in tandem with Robotic Process Automation (RPA) allowed businesses to reduce the manual process of computing large volumes of data. As we’ll see in the next chapter, AI will be the next frontier in this evolution in the field of unstructured data.

Conventional forms of structured and unstructured data can be said to answer the questions, what, who and why. EPoS data provides a historical record of goods sold, prices paid, promotional

mechanics used and so on. Panel data gives a further layer of detail, with crucial information on who is buying any given product and their motivations for doing so.

Data on investments in the food and drink industry can help answer the critical question: what’s next? The Coca-Cola Company’s recent USD 20 Million investment in kombucha brand Health-Ade; Molson Coors’ acquisition of Dr Kombuch; and a growing number of private equity investments in the growing adult soft drinks category all offer valuable insight into the direction the industry is heading as consumers cut back on alcohol and make health-based purchase decisions.

But the capability of structured investment data to predict the future is limited. Clearly, brands such as Health-Ade and Dr Kombucha had already established a following before

being targeted by their new owners. For manufacturers to get a real head start in product development, they need to be able to identify trends before they become mainstream and show up on the radar of conventional sources of structured and semi-structured data.

That’s where unstructured data comes in. It’s estimated there is nine times as much unstructured data than structured data available to businesses16. Not all of it is useful, of course. There’s a risk of developers being swamped by the sheer volume of data available. To exploit its value, they need to be able to navigate it quickly and efficiently and sift the useful from the trivial.

The good news is that with developments in AI, the tools businesses need to unlock the data that can turbocharge their NPD are now at their disposal.

WNS.COM | 08

The Internet is the largest receptacle of unstructured data that’s ever existed. Its scale is dizzying. There were 4.1 Billion Internet users globally by December 201817, up from 3.7 Billion a year earlier. Every day, 5 Billion searches are carried out on Google18, 500 Million tweets are sent19 and Facebook now boasts 2.4 Billion monthly users20, making it the world’s biggest social network.

The Internet is awash with reviews, restaurant menus, blog posts, newspaper and magazine articles, photographs on Instagram, technical reports, academic papers and many other sources of unstructured data. It needn’t be online, either; everything from e-mails and phone calls to office memos have the potential to be sources of valuable data.

Most experts agree that the greatest value for food and drinks companies that want to get ahead of the curve lies online. The data that’s held on on the world’s servers contains many secrets about the movements that will shape the FMCG landscape of tomorrow. The trick is in knowing where to look and to identify what’s important before anybody else does.

An emerging brand in a developing market might not show up in structured sales data, but that no longer matters. Analysis of social media or blog content from that market could reveal growing consumer interest in that brand before it is picked up in any conventional market analysis. Such insight could be crucial for a multinational looking to bolster its portfolio.

That’s where AI comes in. Trawling technical reports, polling consumers or visiting emerging markets is expensive and time-consuming. Machines can do in seconds what would take humans hours or even days. AI doesn’t just accelerate the innovation process by unearthing the latest trends far more quickly, it reduces the risk of failure saving time and money in the process.

CHAPTER THREE: UNEARTHING THE LATEST TRENDS

09 | WNS.COM

The digital revolution has transformed the world of FMCG. It’s created new ways of selling, marketing, managing supply chains, researching new products, driving efficiencies and more. The opportunities of this new digital age are wide, but challenges remain. In a world of structured and unstructured data, in order to be able to analyse the necessary insights from it, it needs to be cleansed, integrated, codified and categorised. That’s where WNS emerging brands identifier platform, Sentinel and its cognitive data capture and processing platform SKENSE comes in.

Any kind of data takes time to process. As we’ve seen, robotic process automation and optical character readers have helped cut the resources needed to effectively process structured and semi-structured data. The next frontier for businesses is effectively processing unstructured data. And that is a very different beast.

Evidence of a nascent food and

drink trend may be buried in blog posts, Instagram images, customer reviews, academic papers and more. Each presents information in a different way, making it a laborious and expensive process for product developers to make sense of. Using advanced AI, SKENSE does just that, so you don’t have to.

The research process begins with WNS, Sentinel, which crawls the internet for data on relevant consumption trends and behavioural patterns that is then verified against existing market research. Sentinel automatically scans the digital presence of emerging companies and start-ups operating in the space, looking at multiple sources such as brand pages, social data, and venture capital websites.

As well as harnessing unstructured data, WNS is able to further qualify and size this data with non-traditional structured and semi-structured data and then rank emerging brands using various algorithms for ease of consumption.

Once new trends and products have been established, SKENSE then delivers cognitive interpretation of those results using machine learning, deep learning, natural language processing, voice recognition, and other systems to process and contextualise vast sets of data quickly and efficiently over a prolonged period of time. This allows NPD teams to keep a pulse on consumers without significant ongoing operational investments. In a recent project with a leading online food delivery platform SKENSE delivered a 75% efficiency saving in the process of uploading menus to its site.

SKENSE offers much higher text extraction accuracy than comparable methods, cutting the time needed to manually check and correct text. It applies structure to the unstructured, presenting data in a format that is easy to read and quantify. In short, it revolutionises the process of managing vast amounts of unstructured data. In doing so, it can transform your NPD pipeline.

Identifying the trends in unstructured data and making sense of them with WNS’s Sentinel and SKENSE platforms

WNS.COM | 10

11 | WNS.COM

The blogosphere is a huge and valuable source of unstructured data for food and drink companies. More than 4 Million blog posts are uploaded to the internet every day21. The biggest blogs, such as Gwyneth Paltrow’s Goop or Joe Wick’s The Body Coach, have readerships greater than many national newspapers. Ignore them at your peril.

Of course, no one has the resources available to read 4 Million posts a day. But with the development of sophisticated artificial intelligence, such as SKENSE, you no longer need an infinite amount of time on your hands. By generating insights on what’s creating excitement in food and drink, SKENSE can give brands a valuable advantage over the competition. Here are three key insights from the world of blogging that you might have missed…

1. “We’re super into algae like chlorella and spirulina here in the Goop test kitchen, but it’s no secret that their strong and earthy flavours can be less than palatable.” Goop

Goop has almost 100k followers on Twitter, so what it says matters. This reference ignited the growth in interest in freshwater ‘blue’ algae chlorella and spirulina which, in 2019, led to Innocent launching spirulina infused smoothie Bolt from the Blue. Health conscious consumers prize these complete plant proteins as rich sources of essential fatty acids, antioxidants and chlorophyll. Expect more innovations with algae to land in the near future.

2. “You know I love a burger, so I had to include one in my new Veggie Plan. This mushroom burger with lentils and chickpeas is so juicy and it tastes amazing.” The Body Coach

Joe Wicks, AKA the Body Coach (Twitter followers: 310.1k), is a living embodiment of many key health and fitness trends. His 90-day diet veggie plan reflects the growing number of consumers who are avoiding meat but are still looking for dishes that offer the taste, texture and satisfaction of meat products.

3. “Fat Bombs are all the rage. Whether you're on the keto diet or are just looking for a dessert to make for the week, fat bombs are for everyone.” Fit Foodie Finds

Brands such as Grenade Carb Killa have developed the sort of low carb sweet treats bloggers such as Fit Foodie Finds (Instagram followers: 208k) have been raving about for years. Grenade’s sales more than doubled in 201822 and the brand now sits alongside the likes of Mars and Cadbury in retailers’ impulse fixtures.

Mining the blogosphere

WNS.COM | 12

Smart analysis of unstructured data can make the difference between success and failure in NPD. It can help you get to grips with emerging trends before the competition, enabling you to develop new products more quickly and efficiently and at less cost. Crucially, it reduces the chance that your new product will be one of the 76 percent that fail within the first year.

As we’ve seen, it’s not just trends that change; shoppers’ motivations for following them do also. By bringing structure to the millions of exchanges taking place online every day, AI allows players to not just identify trends, it allows them to devise effective strategies for capitalising on these trends with new propositions.

So what future trends does analysis of unstructured data reveal?

Social media posts featuring food and drink often reveal more than the preferences of whoever posted the content. Often, they tell us when and for what reason people are choosing to consume certain types of food or drink. For example, analysis of unstructured data online reveals a key trend in soft drinks: the use of energy drinks during exams and video game sessions to improve mental cognition.

Enter a new generation of ‘nootropic’ drinks marketed on their potential to boost alertness, memory and even creativity. New brand Wow Focus claims to offer a drink that makes you think, literally, thanks to a blend of amino acids, vitamins minerals and natural caffeine. Similarly, nootropic brand LGND claims the mix of botanicals in its energy drinks help boost productivity and focus. Expect more nootropic newcomers in the market soon.

Likewise, analysis of online activity reveals a resurgence in interest in the health-giving properties of charcoal. Online influencers including Kendall Jenner are raving about the teeth whitening properties of charcoal infused toothpaste, for example. Others are taking it in supplements, shots and face masks.

The backlash against plastic pollution is clearly evidenced online too. But so are solutions. Examples of innovative bio-packaging solutions made with everything from banana leaves to seaweed are growing, giving food for thought for FMCG manufacturers under increasing pressure to reduce the impact of their operations on the planet.

These are just three of the trends emerging from better use of unstructured data. There are countless more.

AI will be key to helping businesses navigate which trends are most important to them in the future. Those who use it smartly will gain a head start in the race to launch the next game-changing grocery innovation.

13 | WNS.COM

CHAPTER FOUR:TURBOCHARGING YOUR NPD

1 Kantar 52 w/e 30 December 20182 Kantar 52 w/e 31 December 20173 Kantar 52 w/e 1 January 20174 Kantar 52 w/e 30 December 20185 Nielsen 20146 Kantar 12 August 20187 Waitrose Annual Food & Drink Report 20188 Vegan Society 20189 Acosta Europe 201910 Nielsen 52 w/e 18 May 201911 Poore & Nemecek 2018

12 Nielsen 52 w/e 18 May 201913 IRI 201914 Dalston’s Soda Co 201915 The Grocer – GRS 201716 WNS17 HostingFacts — Dec 17, 201818 HostingFacts — Dec 17, 201819 HostingFacts — Dec 17, 201820 Facebook Q2 201921 HostingFacts — Dec 17, 201822 The Grocer’s Top Products 2018

WNS.COM | 14

WNS (Holdings) Limited (NYSE: WNS) is a leading Business Process Management (BPM) company. We combine our deep industry knowledge with technology, analytics and process expertise to co-create innovative, digitally led transformational solutions with over 400 clients across various industries. The industries include banking and financial services, consulting and professional services, healthcare, insurance, manufacturing, media and entertainment, retail and consumer packaged goods, telecommunications and diversified businesses, shipping and logistics, travel and leisure, and utilities and energy. We deliver an entire spectrum of BPM solutions including industry-specific offerings, customer interaction services, finance and accounting, human resources, procurement, and research and analytics to re-imagine the digital future of businesses. We have delivery centers worldwide including in China, Costa Rica, India, the Philippines, Poland, Romania, South Africa, Spain, Sri Lanka, Turkey, the United Kingdom and the United States.

About WNS

COPYRIGHT © 2019 WNS GLOBAL SERVICES

To know more, write to us atmarketing@wns.com or visit us at www.wns.com