Artificial intelligence (AI) is making a lot of headlines right now. Tools like ChatGPT and visual art generators are shocking audiences with their creativity, lucidity and function. Investment in everyday business-focused AI and machine learning is accelerating exponentially, with more sophisticated options appearing in the market.
AI is poised to shift the way companies operate in many industries. As a product executive for a life sciences company, I’m often asked when AI will revolutionize our industry. The answer: Realizing the wider benefits from AI for quality and manufacturing professionals may be out of reach until critical changes are implemented.
AI can be a transformative and valuable tool for healthcare companies today—most, however, aren’t ready to adopt helpful solutions. For one thing, functional AI requires large quantities of digitized data, and many life science practitioners are still storing data in paper notebooks, logbooks and binders. There’s much that would need to change before companies of all sizes could benefit from AI.
The following are key first steps before AI can be effective without putting companies and consumers at risk. Digital maturity is the foundation of life science quality and manufacturing benefitting from AI. This can be a big step for smaller companies that still utilize paper records in daily work.
AI will provide the most accurate findings, predictions and recommendations only when it has access to wider, more complete data about your manufacturing and quality processes. For example, an AI model designed to understand which portions of your manufacturing process are contributing to product defects can only consider the data within its view.
For a comprehensive view of variables, you’ll need to ensure the model has access to data about your suppliers, materials, equipment, quality processes and each step of your manufacturing process. When stripped of the ability to consider the right data factors, an AI model will draw false correlations and incorrect conclusions.
Connecting your data from across systems within your company will help ensure that you (and your AI) have an accurate and comprehensive view. Utilizing master data management will help ensure that data is uniform, consistent, understandable and appropriately modeled, providing the right information for your analytics and AI initiatives.
For example, if the materials used in your product manufacturing have different names in each software system, the AI won’t understand that each ID represents the same material. This prevents the model from having a complete understanding of the role that material plays in your overall process.
Overcoming the sizable challenge of digitizing and connecting data right now may be enough to disqualify many life science companies from achieving value from AI.
Develop a clear vision of the problems you want to solve.
When life science companies are thinking about adopting AI, one of the most important considerations in the implementation is that the margins of error in production are minuscule. Any mistakes don’t just lead to a product recall or downturn in business—they could be life-threatening.
Developing a clear understanding of your company’s challenges and where AI can help is an important early step toward implementation. In many industries, it can work to adopt a technology tool first and then identify problems and use cases appropriate for the solution.
However, the inherent risks and challenges in manufacturing for life science companies make that approach unwise. For AI to provide benefits, we must identify areas with a high volume of relevant data and a reasonable margin for error.
For example, AI can evaluate data resulting from your manufacturing process, proactively monitoring and detecting production runs and environments in which the risk of potential defects is higher than the norm. Then, it can make recommendations for further investigation and testing.
Human oversight and intervention can initially validate findings and ensure that the model is performing well and becoming more accurate over time. In this way, AI can be used to minimize risk on the manufacturing floor in a safe and controlled way.
It’s also important when utilizing AI in life sciences to ensure that models are explainable, with mechanisms for clearly demonstrating its conclusions. And in the highly regulated world of life science manufacturing, it’s also critical to ensure that models are utilizing data in compliance with industry regulations. By first identifying areas in which AI can provide real value without heightening risk, companies can get the most out of technology investments while maintaining high safety standards.
Embrace the benefits and limitations of AI.
AI is an advanced technology, but it still has limited understanding—it only “knows” the information to which you’ve given it access. This means an AI model oriented to making recommendations will do so even without sufficient data—that’s how it’s been programmed. To avoid false findings and conclusions, you must ensure that the data consumed by your AI models is complete and offers a holistic view of the problem.
When given access to the appropriate data, AI can help recognize patterns in manufacturing, supplier and quality data that aren’t detectable to the human eye. For example, a human might notice that more defects come out of a batch when an oven is set to 375 degrees Fahrenheit rather than 360 degrees. However, an AI model might be able to recommend that 363 degrees is the optimal temperature for the lowest number of defects. Or, when a substitute material from a specific supplier is used, the temperature actually ought to be 366 degrees.
AI opens a world of potential benefits for life science companies. To make the leap, most life science companies will have foundational steps to take before their efforts can bear fruit. Companies that are serious about attaining these benefits must determine what they hope to achieve with AI, ensure that the relevant data exists in their digital ecosystem and then take initial, cautious steps with an eye toward minimizing risk.
Investors with allocations to emerging market debt now need to understand the true impact on developing economies of long run factors like climate change and human capital development.
Governments everywhere are racing to lock in historically low borrowing costs by issuing ever longer dated debt – in recent years Mexico and Argentina even managed to sell century bonds. That presents several new challenges for fixed income investors. Particularly those who own emerging market bonds.
Not only do bondholders have to weigh the usual near-term factors like political, economic and commodity cycles but, in lending money to sovereigns over such extended periods, they now also have to consider the impact of longer term trends such as climate change and social development. Both can affect creditworthiness in profound ways.
This has called for new approaches to investment thinking. Economic and financial forecasts are having to be recast with climate dynamics in mind. Meanwhile, modelled pathways of climatic change are themselves subject to expectations about future technological change as well as the evolution of political thinking in these countries. The number of moving parts only grows as investors realise they also have a role to play in shaping how governments approach making their economies sustainable and low-carbon.
It’s a complex problem. But not an insurmountable one.
The greening of EM debt
In 2015, some 17 per cent of emerging market hard currency debt had a maturity of 20 years or more. By the start of 2021, that proportion had grown to 27 per cent. Even local currency denominated emerging market debt, which tends to be shorter-dated, has moved along the maturity curve. Over the same time period, the proportion of local currency debt with a maturity of five years or longer had risen 11 percentage points to 58 per cent (1).
That shift reflects growing demand for yield from investors starved of income. But at the same time, bondholders have recognised the importance of taking a long-term view on environmental issues. This is apparent in both the appetite for green bonds – capital earmarked for environmental- or climate-related projects – and, more generally, bonds that fall under the environmental, social and governance (ESG) umbrella.
Governments are happy to meet that demand. Increasingly, they recognise the need to make efforts to mitigate climate change, and given that emerging market economies make up half the world’s output, they have a significant role to play in meeting global greenhouse gas emissions goals.
In the five years to the end of 2020, annual issuance of green, social and sustainability bonds by emerging market governments grew nearly four-fold to USD16.2 billion (2). And demand is only increasing. For instance, in the first few weeks of January, Chile met 70 per cent of its expected USD6 billion debt issuance for 2021, all in green and social bonds and it plans only to issue sustainable and green bonds during the remainder of the year (3). In September 2020, Egypt became the first Middle Eastern government to issue a green bond. It raised USD750 million to finance or refinance green projects. Investors were enthusiastic – the bond was five times oversubscribed (4).
And generally, these bonds have longer maturities than conventional fixed income securities. Some 46 per cent of USD36.8 billion of outstanding emerging market ESG bonds priced in local currency terms have a maturity of more than 10 years, while for emerging markets hard currency ESG bonds, it’s 41 per cent of USD12.9 billion of outstanding bonds (5).
These bonds allow investors to track performance, while green agendas can also help governments to improve their credit ratings, which then lifts the value of their debt, thus rewarding bond holders.
Overall, green bonds generate positive feedback effects. The rising volumes of green and sustainable bond issuance highlights investors’ willingness to take more of a long-term approach to EM investing. But at the same time, governments are being made more accountable – in order to issue these bonds, governments are having to publish their sustainability frameworks in greater detail. This additional accountability helps to mitigate political risks that are a key consideration in EM investing. Investors, however, will need to analyse and monitor developments closely to ensure proceeds are used as intended.
Indeed, green bonds are the most exciting development in emerging market financing for decades and, we think, will have an equivalent impact to the Brady bonds of the 1980s (6) – albeit this is dependent on improved disclosure and monitoring and industry standardisation of green labels.
Climate change matters (especially in EM)
For all the sovereign issuance of green bonds so far, a great deal more funding will need to be raised to limit climate change. Globally it will cost between USD1 trillion and USD2 trillion a year in additional spending to limit global warming, some 1 per cent to 1.5 per cent of worldwide GDP, according to the Energy Transitions Commission (7). And a significant part of those costs will need to be borne by emerging economies, not least because they are likely to suffer most.
By the end of this century, unmitigated climate change – entailing warming of 4.3° centigrade above pre-industrial levels – would cut per capita economic output in major countries like Brazil and India by more than 60 per cent compared to a world without climate change, according to a report by Oxford University’s Smith School sponsored by Pictet (8). Globally, the shortfall would be 45 per cent.
Limiting warming to 1.6° C would sharply reduce that hit to roughly 27 per cent of potential output per capita for the world as a whole, albeit with considerable variation among countries. While those in the tropics countries would be hit hard by the effects of drought and altered rainfall patterns, those in high latitudes, like Russia, would be relative winners as ports become less ice-locked and more territory is opened up to extractive industries and agriculture. And though China would suffer smaller overall losses than average, its large coastal conurbations would be subject to depredations caused by rising sea levels.
Integrating risks
As these effects are felt, investors will grow increasingly wary of lending to vulnerable countries. And climate change is already having an impact on developing countries’ credit ratings. In 2018, rating agency Standard & Poor’s cited hurricane risk when it cut its ratings outlook on the sovereign debt issued by the Turks and Caicos (9).
Investors could expect climate-related events, like droughts, severe storms and shifts in precipitation patterns, to push up output and inflation volatility in emerging economies during the next ten to 20 years, according to Professor Cameron Hepburn, lead author of the Oxford report.
That would represent a significant reversal for emerging market sovereign borrowers. Since the turn of the century the relative rate of growth and inflation volatilities between emerging and developed markets has halved (10), which, in turn, has reduced the risk faced by investors. Rising economic volatility would feed into sovereign risk assessments, eroding their credit profiles.
Other research from the Oxford team highlights the choices countries will need to take to remain on the path towards building a greener economy (11).
At Pictet Asset Management, we already use a wealth of ESG data – from both external and internal sources –as part of how we score countries. The environmental factors we monitor include air quality, climate change exposure, deforestation and water stress.
Social dimensions include education, healthcare, life expectancy and scientific research. And governance covers elements like corruption, electoral process, government stability, judicial independence and right to privacy. Together these factors are aggregated to become one of six pillars in the country risk index (CRI) ranking produced by our economics team.
Level playing fields
We believe that ESG considerations are inefficiently reflected in emerging market asset prices. This is a consequence of the market still being at an early stage in its understanding and application of ESG factors and analysis. There is also a lack of consistent and transparent ESG data for many emerging countries. We believe that using an ESG score alone is simply not enough. Having a sustainable lens through which to examine emerging market fundamentals helps us to mitigate risk and unearth investment opportunities. We use our own ESG data and analysis and engage with sovereign bond issuers to help bring about long-term change.
Emerging market economies vary hugely in their degree of development. This complicates how investors should weigh their ESG performance – after all, richer countries are more able to make the ESG-positive policy decisions that often have high front end costs for a long tail of benefits, such as shutting down coal mines in favour of solar power.
Applying the most simplistic approach to ESG – investing on the basis of countries’ ESG rankings – would squeeze fixed income investors out of the poorest developing countries, even if they are implementing the right policies to improve their ESG standing. Instead, it’s important for investors to recognise what is possible and achievable by poorer countries and allocate funding within those constraints – understanding countries’ direction of travel in terms of ESG is critical to analysing their prospects.
One solution we are implementing at Pictet AM is to weigh ESG criteria against a country’s GDP per capita. So, for example, under our new scoring system, Angola does well on this adjusted basis despite having a low overall ranking. And the reverse is true for Gulf Cooperation Council member states.
Dynamic approaches
How governments react to long-term issues like climate change or to the challenge of developing their human capital will influence their economies’ trajectories and, ultimately, play a role in their credit ratings. Those long-term decisions are only growing in importance, not least given the scale of fiscal policies implemented in the wake of the Covid-19 pandemic. Tracking these spending programmes – through, say, the likes of the Oxford Economic Stimulus Observatory (12) – then becomes an important step towards understanding the ESG pathways governments are likely to follow.
Countries with good, well-structured policies are likely to see their credit ratings improve, which attracts investors, drawing funding into their green investment programmes and ultimately driving a virtuous investment cycle.
Engaged investors
All this implies that investors have an active role to play – they can’t just passively allocate funding based on index weightings or be purely reactive to policymakers’ decisions. The most successful investors will help steer governments towards the path that boosts their credit ratings, gives them most access to the market and improves the fortunes and potential of citizens.
Like, for instance, explaining how electricity generated by wind turbines or solar can prove to be more cost-effective over the long term if financed by green bonds than ostensibly cheaper coal extracted from a mine paid for with higher yielding conventional debt. Or how fossil fuel investments could prove to be major white elephants as these sorts of polluting assets become stranded by shifts towards cleaner energy production.
Or that failing to invest enough in education is a false economy that over the long run will fail to make the most of human capital and thus depress national output – something we raised with the South African government after our meetings with our on-the-ground charitable partners in the country. To that end, The World Bank produced in 2020 a timely guide on how sovereign issuers can improve their engagement with investors on ESG issues (13).
This sort of intensive analysis – using everything from long run macro models down to meetings with leaders of youth clubs in impoverished districts – can also help to paint a rounded picture of what’s happening in a country. For instance, it helped to ensure that we weren’t caught off guard by the shift to populism in Argentina ahead of their last elections and allowed us to trim our positions in the country.
For emerging market investors, ensuring all of these cogs mesh correctly is a difficult proposition, especially given that the parts are moving all the time, many driven by forces that will develop over many decades. But by using the full breadth of analytical tools, independent research and shoe leather fact-finding, it’s possible to gain a deeper and more profitable insight into these markets than a simple reading of credit ratings or index weightings offers.
And, at the same time, influence policy makers to champion their country’s sustainable initiatives. Taking a sustainable approach to growth and issuing related bonds, emerging economies can fundamentally change their prospects for the better. It has the potential to be revolutionary for emerging markets and exhilarating for those of us who invest in them.
Mary-Therese Barton joined Pictet Asset Management in 2004 and is the Head of Emerging Market Debt. Before taking up her current position in 2018, she was a Senior Investment Manager in the team. Mary-Therese joined as an Emerging Market Debt Analyst. Prior to joining Pictet she worked at Dun & Bradstreet, where she was an economist responsible for analysing European countries.
Businesses in Australia and New Zealand that use data effectively can, on average, increase their annual revenue by 9.5%. This translates to an additional $38 million in annual revenue for large organisations in Australia with more than 200 employees.
According to a new AWS report prepared by Deloitte Access Economics, organisations with more than 100 employees improved their data capabilities in the previous year, with 34 per cent achieving Advanced or Master levels of data maturity, compared to 16 per cent in 2021.
Almost half of the organisations polled (48 per cent ) stated that effectively capturing and analysing data can lead to increased productivity, followed by improved customer experience (45 per cent ) and lower operating costs (42 per cent).
Finance and insurance companies scored the highest on the data maturity scale, with 50 per cent achieving Advanced or Master status, followed by manufacturing (45 per cent ) and information, media, and telecommunications (33 per cent ).
On the other hand, construction, healthcare and social assistance, and retail trade organisations have the lowest data maturity levels, with less than 20 per cent of surveyed organisations in these industries achieving Advanced or Master levels of data maturity.
Unusual challenges
While improving data maturity benefits businesses, large organisations in Australia and New Zealand continue to face challenges in climbing the data maturity ladder, with 42% of organisations achieving Basic and Beginner data maturity.
The main barrier cited by organisations to use data and analytics was a lack of funding (44 per cent ), which has been exacerbated by COVID-19, with 49 per cent of respondents reporting that competing priorities have resulted in fewer resources for data and analytics since the pandemic’s onset. Furthermore, 37 per cent of organisations cited poor data quality as a barrier to businesses adopting more advanced data analytics.
“We are excited to see that more organisations have advanced their data capabilities, which will help them to drive productivity, and create a positive impact on the economy while delivering significant financial returns for their business,” said John O’Mahony, partner at Deloitte Access Economics.
“Investing in cloud solutions will help businesses further their data capabilities and leverage advanced analytics tools such as artificial intelligence, machine learning, and the Internet of Things to achieve data-driven insights.
In fact, businesses that already use the cloud are 71 per cent more likely to have invested in artificial intelligence and machine learning capabilities versus organisations using on-premises data storage. To increase productivity and innovation, organisations should have a clear and practical roadmap for advancing on the data maturity ladder, invest in attracting and retaining talent, and leverage the right technology to reap the full benefits.”
According to the report, one-third of Australian and New Zealand organisations (35 per cent ) cited a lack of skilled resources as a barrier to developing their data and analytics capabilities. To improve data maturity, 33 per cent of surveyed organisations prefer to upskill their current employees, followed by outsourcing to other organisations (24 per cent ), and hiring skilled staff (24 per cent ).
“Data can be an invaluable source of growth for organisations in Australia and New Zealand. The key is recognising its inherent value, analysing it effectively, and building a data-driven culture.
“No matter what stage organisations are in their data journey, AWS is committed to helping customers leverage the scalability, cost efficiency, and security of the cloud to scale their data projects and unify their data to drive productivity and innovate on behalf of their customers,” said Rada Stanic, chief technologist at AWS in Australia and New Zealand.
“Organisations will also benefit from building data skills within their teams, which may involve upskilling current staff through on-the-job training and training courses or collaborating with organisations such as our extensive network of AWS Partners.
“As organisations increase their data maturity, it will transform how they solve problems and build customer experiences, leading to breakthroughs in all industries, including healthcare, finance, retail trade, and manufacturing operations.”
Think of your favorite movie as a kid, say in the first 10 years of your life. Now think of your favorite movie from the past decade. Do you have one? Do you have 100?
In a world with basically infinite content, choice is one of our greatest joys—and frustrations. With each passing year, consumers seem to grow more fickle and demanding, regularly moving to the platforms and publications that offer not only the best catalog but also the best customer service, content experience, user interface, and bang for the buck. And even these features may not be enough, as the recent upheaval among the major streamers has shown.
Holding on to viewers, readers, and listeners has become more important than ever. Yet most consumers can only maintain so many subscription services at once. The goal for media companies needs to be to sustain their interest, and with as much share of the consumer’s wallet as possible.
As such, churn is now the most prominent enemy of the media and entertainment industry business model. Consumers can be mercurial, sensitive to price and changes in content catalogs. Just as adding a service has rarely been easier, so is dropping one, which consumers have shown themselves more than willing to do when a channel is no longer serving their needs.
With these challenges front and center, leading media and entertainment companies are increasingly turning to data analytics and personalized content recommendations to improve customer experience and retention. In the dog-eat-dog digital world, it’s no longer the loudest bark that gets the most attention. It’s about pairing the right breed to the sensibilities of a specific person, and having the best stable of information and offerings to make that match and keep it going.
As subscriptions have risen, so has churn
A good example of the challenge of churn can be seen with streaming video. Deloitte performed a series of surveys in 2020 to gauge how consumers were changing their media consumption habits amidst the pandemic.
In January 2020, the average consumer in the United States subscribed to three paid streaming services; by October 2020, the number of subscriptions had risen to five. Overall, a positive development for media, but with the increase in subscriptions came a commensurate increase in churn.
In January 2020, Deloitte found that only 20% of people who had subscribed to a paid streaming service had cut at least one of those services in the past 12 months. By October, that number more than doubled, with 46% of consumers canceling a streaming service in the preceding six months. And at that time, 34% of consumers said that they’d both added and canceled a streaming service since the pandemic started.
Why did viewers churn? Deloitte noted that 62% of people in 2020 who had signed up for a service and then canceled it had done so because they signed up to watch a specific show, then canceled the service when they’d finished watching it. Price, as always, was also a big factor. In October 2020, 31% of people who canceled a service did so because it was too expensive. Another 28% canceled because a free trial or discount period ended. About 21% cut the service because of a lack of content they found interesting.
No matter how focused on addressing churn a company may be, what can they do when the whims of the consumer are so sensitive and fluctuate wildly?
Companies need to find ways to anticipate what their audiences want at least as well as the audience does—and certainly better than their competition. Two of the best defenses against churn are having an organized data platform, then using that data to personalize content recommendations and customer experience.
Data maturity is the first step to mitigating churn
Data maturity is the ability to have accurate and reliable data that can be utilized through cloud platforms, with advanced analytics informing every decision. It is one of the most important steps for media and entertainment companies to take in the effort to mitigate churn
In our experience working with companies as varied as Spotify, The New York Times, Major League Baseball, and Hearst, the first step to achieving data maturity is building a company culture where data is prioritized within the strategic business framework, and where funding is allocated to technology and human resources to build a mature data ecosystem.
Data maturity should not be a bolt-on to existing practices, but needs to become central to the company’s strategic business goals. Companies that have achieved data maturity tend to have specific teams or centers of excellence that manage goals, strategy, and tactics of the organization’s data framework.
In a 2020 survey by EY Global Media & Entertainment, 62% of media and entertainment executives said they saw the increasing availability of data as an opportunity. About 56% prioritized first-party data, versus only 13% who prioritized third-party data. When asked about their top three data priorities, 44% said that the consolidation of customer data was a top concern. About 40% said developing proprietary data sources was a priority, while 39% prioritized improving the relevance of data.
Consolidating data out of data silos to a unified data platform is the biggest challenge that most companies will face when building a roadmap to data maturity.
A report by Deloitte in partnership with the Google News Initiative on how news and media companies can achieve digital transformation through data outlined some of the technologies that companies can adopt to achieve data maturity. Two elements are required. First, media and entertainment companies need to be able to collect and store data that they are gathering from their planet-sized audiences and users with the tools listed below.
Data management platform (DMP) helps to manage first-party data segments and integrate third-party data and push data to other systems.
Data lake or warehouse, a central repository of data from multiple sources.
Cloud storage for reliability, security, and scalability.
Customer relationship management (CRM) the backbone of customer data that records and tracks user interactions with registered subscribers.
Customer data platform (CDP) to record and track customer data across platforms and devices.
Second, companies need to make sense of all that data and derive actionable insights from it.
Data analytics and reporting tools that can collect, organize, and analyze data from multiple sources.
Artificial intelligence and machine learning tools. Derive even more insights through AI/ML-enabled capabilities such as computer vision, speech and object recognition, and text translation.
Propensity modeling helps build a better understanding of customer preferences, fulfilling the key elements of personalization to prevent churn.
Below we describe some of the unique data sources available to media and entertainment companies and how it can be applied to artificial intelligence and machine learning.
Media and entertainment have unique data sources
Media and entertainment companies can improve personalization by tapping two unique sets of data particular to the industry: media content and audience behavior.
Media content includes easily identifiable metadata such as the title, headline, genre, topic, or format of a piece of content. But media data can also include context of the actual content itself.
For instance, AI tools like object recognition and computer vision can detect items within a movie and then add the description of the object to the searchable metadata of the content. If a television show contains a border collie, the AI can recognize the good dog and surface the show in a search for “shows with dogs.” Or with speech recognition and translation, AI can build a data set of the dialogue within a movie and make certain keywords part of the search for that show.
Behavioral data of the audience can be used in a variety of ways. Data can come from many different sources including a person’s location, device, browsing and scrolling, user profile, engagement, billing preferences, purchase and support history. Companies can help personalize experiences with this data by understanding how people interact with content and how best to engage with them, such as what times of the week are best for push notifications or when a person might be most amenable to a content recommendation.
Using artificial intelligence to personalize user experience
If you’ve ever wondered how your favorite streaming service seems to so uncannily know what you want to watch—even better than you might—the answer is probably some clever AI. Personalization is the practice of combining the new, massive datasets outlined above with machine learning and artificial intelligence to create experiences tailored to the specific needs and behaviors of an individual person.
Personalization is often associated with content recommendations. For example, about 70% of what is viewed on YouTube comes from a personalized recommendation. Certain streaming services are known to have some of the best content recommendation systems in the business. The goal with the personalization of content is to surface a new show, video, movie, podcast, song, band, album, article, or blog to the person at precisely the right moment.
Personalization is also an important element in search. Consider that with the right data inputs, two users searching for the same keywords could get vastly different results attuned to their consumption preferences. In both cases, content better suited to a person’s interest will keep them from looking around at other platforms or publications, helping to reduce churn.
The same is true for more traditional outlets, as well. Take a recent example from the (digital) pages of Newsweek. The publication’s chief technology officer, Michael Lukac, recently noted that “Google Cloud Recommendations AI has not only improved our click-through rate by 50% to 75% and subscription conversion rate by 10% but also allowed us to increase total revenue per visit by 10%.”
If you’re looking for more information about why personalization matters and how to bring it to your own services and experiences, discover more in our new ebook, Personalizing Media for Global Audiences.