Published 22 December 2021 in Technology • 8 min read • 
Audio available
Imagine that you need to publish a magazine for your stakeholders. In a world of information abundance, and with fierce competition for attention, the difference between a magazine that will be avidly read (and shared widely) and one that will be recycled comes down to creating content that will speak directly to the needs and motivations of the audience. It’s the same for other types of content.
I have served as Head of Digital since 2016 at Siemens, Europe’s largest industrial manufacturing company, which focuses primarily on industry, energy, health and infrastructure. In the Siemens communication department, we traditionally solved this challenge with rigorous audience research, building a detailed understanding of the needs of customers and by poring over analytics of past performance to see how we can best meet these needs.
However, this is a time-consuming process that is error prone and not scalable — detailed customer research for every magazine article would not be practical, for example. So, in 2018 we explored how artificial intelligence could help us solve such challenges, and we teamed up with Storyroom, a California-based AI startup, to develop a unique solution for communications and marketing optimization.
As the world becomes more complex and change becomes more rapid, we need to evolve our content strategy to increase our relevance to the continuously changing needs of stakeholders. Siemens provides hardware and software for everything from building maintenance to production of a COVID-19 vaccine. With AI, we’re able to connect our messages directly to external customer conversations and pain points, for example last-mile transportation, blockchain energy trading, and zero-defect manufacturing.
To pivot our messaging, we worked with Storyroom and used the startup’s unique datasets and AI to build a more fluid approach to defining positioning — the place our brand occupies in the minds of customers and how we distinguish our products and services from competitors to drive brand awareness and engagement.
Our new approach allows us to do this exercise dynamically for every customer, product or problem space. For example, when launching a new product we can take the unique attributes, compare this against the market message of our competitors and against the needs that customers are expressing, and determine where we could differentiate ourselves and create a content strategy.
We first came across Storyroom a couple of years ago after becoming convinced of the potential for “natural language processing”, which involves teaching machines to understand the nuances of human language, to transform our content strategy. The startup also uses “natural language understanding”, which deploys syntactic and semantic analysis of text and speech to work out the meaning of a sentence.
Storyroom has made leaps in the field. While computer vision and voice recognition have begun to approximate human capabilities in some use cases, human communication has been a much higher hurdle to climb. Although the complexity and ambiguity of language is to some extent a solvable problem, applying the outputs to real business situations creates another layer of complexity.
Storyroom’s unique approach (see box) has enabled us to create a “conversational map” for any given topic, on which we can layer other data sets, such as social media engagement patterns, to produce unique insights, including emergent trends as they happen. We can answer the precise questions that marketers are asking every day. For example, what opportunities are available for brand differentiation and what conversations are most relevant to our offering?
Using machine learning, we can create meaning and understand patterns in this unstructured data. This allows us to determine a number of unique insights, including where the important “white spaces” are — trending topics that our competitors are not adequately serving. It also uncovers where the conversation is saturated, and what the unique aspects are that we can bring to the discourse that will engage customers.
The AI tells us how we can connect the issues that are important to people’s everyday lives (for example, the challenges of commuting) to the solutions that we can offer in a genuine and useful way, to help us craft compelling and relevant communication. And, we can even reverse-engineer competing communication strategies which further helps us understand how to competitively position ourselves. This gives us the true “outside in” view of the world, free from internal bias.
Market research has traditionally been a time-consuming task, which in the past would require several months to establish a new brand position, increasing the costs of data collection and processing. A manual process, it’s also prone to error and human bias. The research conclusions are based on assumptions, which may quickly become obsolete in fast-changing markets.
Storyroom is a venture-backed AI startup based in California. The core of the technology is the ability to build a “digital twin” of a brand using AI. To do this, the platform ingests massive amounts of data from a multitude of media channels and uses “natural language understanding” to map the content both by and about the brand to different parts of the organization.
The company, through data crawling and data partnerships, has access to unique data sets that are used to build a large-scale polymetric inference graph that discovers latent associations between conversations, topics, entities and organizations.
Then, AI is used to construct a “digital twin” to consolidate the data around the brand and broader strategic trends. The information is used to forecast scenarios for content planning, even down to the specific messaging and AI-suggested headlines for individual content.
AI has been a game-changer for us, as it streamlines research, making analysis faster and cheaper, cutting down project timelines from months to days. The ability to assess real-time data is transformative, in that it makes market research less backwards-looking and more of a predictive indicator.
AI also unearths deeper insights, as it scours thousands of data points. And it presents the research in a structured format that is easily digestible, even to those without a data science background. This enables marketing leaders to make faster, more informed content decisions.
For example, the coronavirus pandemic rapidly changed the online conversations that many of our customers were having about the needs of their businesses. Pre-existing assumptions were turned upside down and our communication needed to adapt to that. A good example is the term “workflow” which we saw rapidly replaced by the term “air flow” as COVID-19 safety concerns became a top priority.
By applying AI, we could view how these discussions were evolving in real time, and respond with greater speed and agility to create targeted content that covers the precise talking points most relevant to our customers.
The AI even supports the creation of content. This is known as “natural language generation”, the construction of text by a machine based on a data set. For example, the AI can write headlines for articles to ensure that they are pre-optimized for search engines, suggest reference material, and the best promotion channels for the content, based on the market signals. For us, this content has been highly effective at engaging customers.
In a pilot, we used the AI to generate headlines for thought leadership pieces about infrastructure technology. The headlines were six times more effective at driving engagement in terms of views, comments and conversions than headlines written by humans in our study. For example, one AI-generated headline — ‘Connected buildings are already here, but how can we all benefit?’ — was effective because it matched exactly to a common topic of conversation among our customers: what are the broader and marketable benefits of connected buildings (i.e., when disparate building systems are connected and controlled via a single operating system)?
Despite our confidence in the process, we do often encounter reluctance. Understandably, people can be hesitant to let go of years of hard-won experience and place full confidence in a machine. Another way that we see this issue manifest is in the reluctance we sometimes see in attributing our insights to an AI. Colleagues are often wary, as they feel that the AI label could undermine confidence in the approach they are suggesting.
We also made mistakes in this regard, by expecting that our approach would be accepted as a given. It’s been necessary for us to spend a great deal of time refining the way that results are presented and the context that we provide around our data to gain greater acceptance.
For example, one key aspect in overcoming skepticism has been to put the technology into a more familiar context. So instead of presenting results in a spreadsheet or PowerPoint document, some of our AI-driven research is formatted in a way that more closely resembles a human analyst’s report. We’ve found that when people consume insights in familiar ways, they are more likely to embrace the outcomes, and in turn the technology behind them.
And once we get that acceptance, we see how engaged people become with this new paradigm, particularly due to the elimination of repetitive processes and routine tasks, thereby augmenting and streamlining many human activities. AI is unlikely to unseat workers, but it will increasingly come into its own as a valuable tool for marketing and communications.
Demand across Siemens for these insights has been high, with most of the major divisions leveraging the AI to improve marketing effectiveness and enlarge returns. While Siemens is pioneering this technique today, we believe it will become a cornerstone of marketing communications in the future.
We are also seeing multiple possibilities for this technology in other functions. The ability to understand conversations at global scale, and the ability to link data sets together, has created interest among many of our colleagues in different departments. For example, how can we understand global sentiments and trends to optimize supply chains? What can advertising trends on recruiting platforms tell us about employee needs, or even competitor strategy?
We are excited about the many possibilities that we see here. It’s little wonder that AI is being billed as one of the most significant technologies of our time. We are convinced of the huge change and benefits it will bring, and of the fact that businesses and employees alike will need to be ready for the journey ahead.
27 November 2023 • by Michael Skapinker in Audio articles
The Church of England has moved away from engagement and sold its fossil fuel shares, but to whose benefit? The shares have most likely been bought by investors who care little about...
Audio available
6 November 2023 • by Sudhanshu Sarronwala in Audio articles
Focusing on a company’s ecological and social impact as well as its economic performance has yielded mixed results. However, there is increasing evidence that even legacy companies are now making positive strides....
Audio available
30 October 2023 • by Bali Padda in Audio articles
Technology may lead to a more efficient process, but it’s not the best option if it worsens the customer experience – and without due regard to data cleansing the results can be...
Audio available
26 October 2023 • by John Elkington in Audio articles
What’s a regenerative business model? It starts with designing companies as if we mean to stay on this planet....
Audio availableExplore first person business intelligence from top minds curated for a global executive audience