AI is ubiquitous. AI is already the reality, present in many devices of everyday use. Yet, 63% of users don’t even realize they are using an AI-powered product.
What’s more, a Nimdzi Insights poll conducted across a panel of users in the United States, the United Kingdom, France, Japan, China, and India shows that AI is making an impact: Two out of three people agree AI is having a profound impact on their habits.
AI is transformative… and has the power to trigger buying decisions: 60% of users are likely to buy a personalized product recommendation they see on an online store.
People buy what they understand. This fully applies to merchandise and services recommended by AI. 9 out of 10 report they will ignore a product if it’s not in their native language.
Artificial intelligence (AI) is a term that invokes polarizing opinions — regardless of whether they may be grounded in reality or science-fiction — among business decision-makers or individual end-users. The trouble with AI today is that it isn’t understandable enough. It’s something being cooked up by a handful of English-speaking engineers at a US-based tech giant. It’s not relatable or relevant to the hundreds of millions who do not even speak English...
As you can see, in three sentences we only scratched the surface of concerns end-users may have towards AI. Companies will need to address these when thinking about deploying AI. And they don’t really have a choice if we look at how AI is becoming increasingly central to what companies do.
More than 70% of AI adopters believe that AI is critically important to their business today and that it will be key to securing market leadership in the future. Investment in AI isn’t slowing down. If anything, different forecasts put spending on AI technologies at USD 97.9 billion in 2023 — more than two and a half times the spending level of 2019.
Today’s world is one that is driven by technology, and it’s also one where the technology is put to the service of conquering new markets and reaching new users.
In this whitepaper, we explore how and why AI can be used to fashion personalized user experiences. The whitepaper explains what AI localization is, and how brands can use it to deliver unique human-centered experiences to users across the globe.
AI is ubiquitous in today’s brand-consumer interactions. Examples abound in companies that have been leveraging the technology in order to build engagement with their end-users and create stickiness (and in turn super-charge their growth). Global enterprises such as Facebook, Apple, and Google have been shaping consumer expectations for years with cutting-edge products that use AI. Others, such as Netflix and Spotify have been winning the ongoing battle for mindshare by smartly leveraging technology that transforms people’s consumption habits.
From the user’s point of view, AI is relatively invisible. For example, in 2016, 63% of global users polled by HubSpot were already using AI tools, without even realizing it. The technology is working behind the scenes, processing large quantities of data in order to learn and adapt the products to better suit users’ specific tastes. The applications of AI can result in functionality that ranges from essential and practical, to convenient and nice-to-have. However, we’ve now come to a point where users, in fact, expect quality of life improvements from the brands that they trust and use.
The research team at Facebook recently introduced a new feature in their project TextStyleBrush that allows users to emulate the style of the text detected in images. It does not only let a user change a typed email into a more personable handwritten note on the screen, but it also has the potential to introduce the photo-realistic translation of languages in augmented reality (AR). In the following example, moving an arrow from right to left allows a user to translate handwritten signs from French to English with the help of what is localized AI.
Image source: Facebook
Google’s neural machine translation and machine learning-based camera feature embedded into Google Translate allows users to instantly translate visual information into another language.
Image source: Google
A company operating in a different field, in early 2021, Spotify completed its latest step towards offering a global service by making itself available to 80 markets, and in some 36 new languages. AI is what underpins the company’s success with its users, and it is also what is behind the company transforming music consumption habits. As it happens, Spotify’s Discover Weekly and Daily Mix feature the use of AI to analyze users’ listening habits to curate and serve up musical playlists to them, thereby seamlessly fostering interactive engagement and ensuring user stickiness.
Examples of the Discover Weekly and Daily Mix features of Spotify.
For instance, the Daily Mix feature is the result of clustering techniques that identify sub-groupings in the user’s listening patterns to build tailored playlists of their six most listened-to genres. Now, the Daily Mix or Discover Weekly playlist features aren’t novel, as they have been around for several years, but they show just how curating content (music, in this case) to the user’s taste can have a positive impact on the company’s staying power and lasting popularity.
Examples of user reactions to the Discover Weekly feature. Source: From Idea to Execution - Spotify’s Discover Weekly
Netflix is another example of a business that is successfully using AI to its advantage. The company has seen millions of subscribers flocking to the platform in recent years. Through rigorous A/B testing, before new features are deployed, Netflix has been analyzing the behavioral patterns of its users. An example of how the company uses AI and machine learning is an innocuous element most users probably haven’t even noticed — personalized thumbnails of recommended content that change from user to user. As it is, Netflix found that different users will become interested in different visual cues, so by serving up customized visuals Netflix has designed a way to invite users to stay around to watch another show. After all, a user that clicks on a thumbnail they like linked to the content they love means a happy user that will remain on the platform and continue to pay for it.
Example of how Netflix uses personalized thumbnails: two different users will see two different images for the same movie. Source: becominghuman.ai
The applications of hyper-personalized AI in customer-facing products are virtually endless. Digital natives are leading the pack in adopting AI to deliver human-centric, intelligent digital experiences.
AI has been the talk of the town for years. Nimdzi has covered it in past publications too, highlighting the intersection of AI and language data — with machine translation (MT) being the prime example of how self-trained technology can be used to expand the global reach of any organization by virtue of enabling access to translated content to millions of users around the globe.
The language services industry has historically helped brands reach consumers across the globe. Language professionals are the ultimate facilitators for the sale between brand and consumer. Without their work, the sale does not take place. Translation — the bread-and-butter service offered by specialist language service providers (LSPs) — makes products and services accessible to consumers. The value LSPs provide does not end by simply offering translation from language A into language B, however. Indeed, they possess the linguistic knowledge that allows them to make translations comprehensible, but there is an additional dimension to the work they do: They understand the cultural lens that influences perception, and how users in various countries approach and consume products and services of their favorite brands. Buying behavior and user preferences DO differ from market to market, and having products available in English only – or without due consideration of local culture – may slow down the brand expansion into global markets significantly.
Adapting a product to the demands of the target market and the various audiences is called “localization”. Companies tap into this service and localize content or products to connect with audiences in meaningful ways, inclusive of culture.
Earlier we looked at the powerful results AI delivers. Now let’s take it a step further and see what would happen if AI was localized?
For AI to have an impact, it needs to be trained on an intimate knowledge of what makes users in various countries tick, what language they use, what culture or religion they identify with, and what visual or auditory stimuli most frequently trigger decision making situations and/or guide users through them. This input can be characterized through data. Data which, in turn, can (and should) be used to fuel AI models.
This is AI localization: an ensemble of services centered around collecting and curating data in order to produce clean, and thoughtfully balanced AI training datasets that will reflect how locals think and interact with the world around them.
It stands to reason, too. Today, if only judging by what the top destinations of AI investment across the globe are — with the US being home to 64% of global AI-centric transactions — there is the distinct feeling that AI is exclusively being designed by English-speaking engineers and programmers. It is a deliberately reductive assessment (and one colored by the latent influence of the Western business world, culture, and media on the collective psyche), as there are naturally other countries pushing the envelope on AI. But that’s the entire point: When crunching the data, AI should be inclusive of languages, and cultures. AI should be localized in order to bring true value and, by extension, appeal to a wide spectrum of users across the globe.
For the purpose of this whitepaper, Nimdzi Insights conducted a comparative user sentiment survey with a panel of more than 1,200 respondents aged 18 and above residing in the US, the UK, France, Japan, China, and India. The findings reveal how attitudes of people towards AI differ from country to country — and are indicative of how brands need to factor in the preferences of their potential users if they want to craft a seamless user experience.
When asked whether the participants agree or disagree with the statement that AI is already impacting the way we, as consumers, engage with brands, on average 67% of users in these six countries agree that AI is becoming commonplace in their daily lives. Users in the US, China, and India feel particularly strongly about the statement.
Source: Nimdzi Insights
One example of AI working behind the scenes is personalized product recommendations that, based on the users’ data, can offer more tailored options for shopping on a website or e-commerce platform. Here again, users in the US and Asian countries feel more receptive toward brands offering customized recommendations than users in Europe.
Source: Nimdzi Insights
Source: Nimdzi Insights
These data points convey the importance of taking a user-first approach towards how AI should be developed and used by companies. Attitudes towards brands and their products can vary significantly from country to country.
And here is where the two worlds — that of language services and AI — naturally converge. On the one hand, language as the product of culture and individual cognition and the services tailored around delivering it allow brands to reach users in new markets across the globe. On the other hand, AI technology allows brands to create novel user experiences at an unprecedented scale. Still, AI that does not consider or reflect the way users talk and think only gets half of the equation right. AI models need to be individually trained on the richness of linguistic, sociocultural, visual, and auditory stimuli that are specific to a company’s global audience.
In today’s world, brands want to stand out and remain relevant to users in their target markets. To do so global players need to incorporate a sound, scalable AI localization strategy. Here are three strategies to help ensure that the implementation of AI within a brand’s products and services will not fail at its most critical objective — ensuring a unique customer experience!
DESIGNING AI PROCESSES WITH HUMANS AT THE CENTER
Being human-centered means designing AI solutions with the needs of people at the center. To do that, a business needs to rely on a diverse set of global talent to train AI applications to be inclusive and bias-free. For example, at Pactera EDGE, we first understand the purpose of their AI applications and their intended outcomes and impacts using human-centered design frameworks. We then ensure that the data we generate, curate, and label meet those expectations in a way that is contextual, relevant, and unbiased.
INCLUSIVE AI DEMANDS DIVERSE DATA
AI relies on data. The AI Localization process relies on vast pools of hyperlocal and in-market user experiences-generated data. Organizations and brands should work with their AI enablement partners to diversify the kinds of data utilized to train their models. This can include emojis, acronyms, abbreviations, slang, and local dialects… even incomplete sentences or abnormal syntactic structure sentences. This results in a vastly superior human-centered multilingual digital customer experience.
UNLOCK THE POWER OF PARTNERSHIPS
Implementing AI localization in an organization may seem like a tall order. It will require onboarding specialist know-how. Organizations should look to augment what they can achieve by working with partners who have the global reach and local network of qualified data collection and annotation specialists they can tap into and organize to help scale their activities.
Freeletics uses AI algorithms to transform people’s training habits
Freeletics, a health and fitness app developer has been using AI to change the training habits of its 50 million users worldwide. The AI coach, via the app’s popular “Adapt session” feature can tailor users’ workouts in 14,000 different ways on any given day. The reported rating accuracy of the training sessions is 85% just after five workouts.
McDonald’s uses AI-powered drive-throughs to decrease customer wait times
In 2019 the average drive-through took six minutes and 18 seconds. In 2020 the company trimmed it to five minutes and 49 seconds. As drive-throughs account for a sizeable part of McDonald’s revenue, the company has deployed an AI-powered system voice assistant to record customers’ orders. The reported accuracy is 85%, with only 15% of orders requiring human intervention. The long-term aim is to further decrease wait times and improve their customers’ user experience.
Bank of America’s virtual chatbot Erica handles 35 million customers requests per year
In 2018 Bank of America introduced their AI chatbot Erica into their mobile banking app to help users with simple actions such as paying bills. In 2019, 6 million users have used the digital financial assistant, which processed 35 million customer requests. AI chatbots are a unique way banks can ensure a 24/7 customer service and build user engagement while driving down operational cost.
Samsung-owned Whisk Food AI helps families make smarter food decisions
Samsung leverages the power of AI via Whisk’s Food AI to transform household appliances such as an innocuous refrigerator to something that can positively impact people’s lives. Food AI, in combination with the built-in ViewInside camera, identifies food stored in the refrigerator and then recommends a curated list of recipes that map to users’ preferences and tastes. The technology allows families to plan meals more efficiently and, equally importantly, helps to combat food waste by offering recipe suggestions based on what remains in the refrigerator.
Levi’s uses a virtual stylist to cut down on customer dissatisfaction and returns
Levi’s virtual stylist program uses an AI chatbot designed to assist customers with their buying decision. It helps them find the matching pair of jeans they were looking for, while incorporating the same specialist know-how of Levi’s employees, even asking questions such as “How do you like your jeans to fit through your hips and thighs?” to offer pertinent recommendations. The service is available online and on mobile devices and has lead to increasing customer satisfaction.
With a core focus on Data, Intelligence, and Experience, Pactera EDGE helps clients achieve new levels of performance while adding brand
new digital business capabilities to drive relevance, revenue, and growth. With clarity of vision, technological expertise, operational
excellence, and a global footprint, Pactera EDGE is the partner of choice for enterprises that want to run smarter – and for those that
want to change the race.
Translation management systems (TMS) are one of the oldest language technologies out there. The first solutions appeared in the 80s with the emergence of brands such as STAR Transit and Trados, and the segment has been booming since 2010. In 2022, there are well over 160 technologies of this type on the market.
The Nimdzi Language Technology Atlas maps over 800 different technology solutions across a number of key product categories. The report highlights trends and things to watch out for. This is the only map you will ever need to navigate your way across the language technology landscape.
Imagine this: you decide to expand your very successful and popular mobile cooking app to other markets across the globe. You want to reach a wider audience and maximize your return on investment. You start by contacting translators and localization experts to ensure your app’s content is accessible to audiences from different countries. But is that enough?
As Nimdzi’s co-founder Renato Beninatto likes to say, there have only ever been three disruptive innovations in the language industry: e-mail, Translation memory software, Machine translation, and Google Translate in particular. But is there anything else? An idea so innovative that it could transform and reshape our industry?