Ein KI-gestütztes Customer Insight Hub für Lufthansa

PERIOD 2021 – ON GOING

Europas größte Fluggesellschaft sammelt Daten zur Customer Experience nun zentral in einem Portal, das fortschrittliche Analysen mit Hilfe verschiedener spezieller Large Language Models ermöglicht. So kann Lufthansa sein Angebot noch besser auf die Wünsche und Bedürfnisse der Kunden abstimmen.

GOAL

Kundenbezogene Daten effektiv nutzen.

Bei der Lufthansa Group, Europas größter Fluggesellschaft, darf der Begriff der „Customer Journey“ ganz wörtlich verstanden werden. Auf ihrer Reise kommen die Fluggäste an unzähligen Touchpoints mit dem Unternehmen in Berührung – von der Online-Buchung über den persönlichen Kontakt an Board bis hin zur anschließenden Kundenbefragung. All diese Interaktionen ermöglichen es der Lufthansa Group, wertvolle Informationen darüber zu sammeln, was einzelne Kunden denken oder wie sie bestimmte Dienstleistungen bewerten.

Um diese Informationen in hilfreiche Erkenntnisse verwandeln und unternehmensweit zur Verfügung stellen zu können, waren bisher aufwändige manuelle Analysen notwendig. Deshalb hat das Unternehmen TD Reply mit der Entwicklung eines zentralen Portals beauftragt, das die Verarbeitung und Auswertung der Daten automatisiert.


CHALLENGE

Die Bedürfnisse, Wünsche und Pain Points der Fluggäste mit Hilfe von Generativer KI besser verstehen.

SOLUTIONS

Fortschrittliche Analysen und externe Datenquellen:

Das neue Customer Insight Hub der Lufthansa Group ermöglicht es dank fortschrittlicher Analysen, die Bedürfnisse, Wünsche und Pain Points der Kunden besser zu verstehen. In diesem Portal werden nun alle kundenbezogenen Informationen wie zum Beispiel Bewertungen zentral gesammelt und mit Hilfe von Künstlicher Intelligenz automatisiert analysiert. Durch die zusätzliche Einbindung externer Datenquellen wie Presseberichte oder Trendstudien profitieren Produkt- und Marketingmanager unternehmensweit von umfangreichen Einblicken. Diese können Nutzer auf zwei unterschiedliche Arten bequem über einen Chatbot abrufen:

Vordefinierte Prompts:

Mit Hilfe von vorgefertigten Befehlen können Nutzer schnell Zusammenfassungen zu bestimmten Themen oder eine Übersicht über die Pains und Gains der Kunden erstellen.

Offene Fragen:

Alternativ können die Produkt- und Marketingmanager dem Chatbot individuelle Fragen stellen und so personalisierte Einblicke zu den Erfahrungen der Kunden mit ihren Services erlangen.

BMW – BRAND IMPACT ACCELERATOR

PERIOD 2021 – ON GOING

Strategically fueling and transforming a brand necessitates precise brand activation management aligned with defined objectives and a multifaceted assessment of brand impact. The Brand Impact Accelerator (BIA) empowers BMW campaign and experiential managers to make informed decisions and foster a culture of learning within the organization. Over the past few years, we’ve critically evaluated more than 50 brand activations, ranging from ArtBasel to the IAA, to assess their strategic brand impact and utilize these insights to enhance the efficacy of future activations.

GOAL

The BIA program was designed to achieve three primary goals:

Goal-Oriented Management: :

Establish a goal-driven steering process that engages all pertinent stakeholders.

Unified KPI Framework:

Create standardized Key Performance Indicators (KPIs) for strategic and tactical activation management.

Granular Brand Image Measurement: :

Develop measurement constructs that enable a detailed assessment of the BMW brand image.


CHALLENGE

The challenge lay in crafting and implementing a data-driven approach that aligns with event and campaign managers, channel owners, the BMW in-house agency, The Game, as well as the KOL agency. We aimed to cultivate a data-driven culture within the BMW Brand Impact Council.

SOLUTIONS

Shared Vision:
We embarked on defining a collective vision spanning the entire brand communication landscape.

Stakeholder Alignment:
Identifying key stakeholders and uniting them behind the common vision, approach, and KPIs. This involved defining KPIs and constructing measurement frameworks based on digital data, such as organic searches on both event and brand levels, and Digital Brand Equity, a measurement derived from social listening data assessing various facets of the BMW brand image.

Reporting Framework:
We established structured reporting formats for pre-event, during-event, and post-event evaluations. This included building a dynamic dashboard and a benchmarking database.

Learning Culture:
Our efforts extended to fostering a culture of continuous learning within the organization.

Future Positioning:
We also explored opportunities for positioning BMW within the broader competitive landscape.

By implementing the Brand Impact Accelerator, BMW has achieved a more streamlined and data-driven approach to brand activation, enhancing the brand’s strategic impact and aligning all stakeholders behind common objectives.

BAYER – DIGITAL TREND RADAR

PERIOD 2022 – ON GOING

TD Reply collaborated with Bayer to develop a Digital Trend Radar, that visualizes emerging trends, associated technologies, and their maturity levels. Leveraging cutting-edge tools such as SONAR (data-driven Trend Research) and PULSE (Dashboard and Visualization), TD Reply crafted a robust platform that empowers the Client’s INNO.X program by providing actionable insights into emerging technologies relevant to its diverse divisions and corporate functions.

GOAL

The primary objective of the Digital Trend Radar project was to democratize trend knowledge throughout the organization, fostering innovation and cross-functional collaboration at Bayer with a systematic and easily accessible tool for trend identification and monitoring. Key goals included incorporating trend insights into strategic decision-making processes, prioritizing trends relevant to specific divisions, and evaluating their potential impact on existing business models.


SOLUTIONS

The Digital Trend Radar enables Bayer to incorporate environmental scanning into the Strategy Formulation, improving the prioritization of relevant trends and assessing their impact on strategic initiatives. The development of the Digital Trend Radar included various work packages:
  1. Stakeholder Engagement:
    • Gathering stakeholder requirements and input to tailor the Digital Trend Radar to Bayer’s specific needs and objectives.
  2. Trend Discovery:
    • Leveraging TD Reply’s extensive trend universe and employing custom trend discovery methods, over 200 trends were identified, ensuring comprehensive coverage of relevant developments.
  3. Trend Prioritization:
    • A blend of quantitative and qualitative analyses was employed to quantify and prioritize trends. Using SONAR (our proprietary data-driven trend research solution), we analyzed millions of expert opinions, including blogs, news, patents, and scientific articles. This quantitative analysis was complemented by expert evaluations of each trend’s impact on Bayer, to select the most relevant trends for further in-depth analysis.
  4. Use Case Identification:
    • Building on the prioritized trends, we identified and described over 250 use cases, offering actionable insights into how these trends could be operationalized within Bayer’s context.
  5. Trend Framework Development:
    • Based on the extensive research phase, a robust and structured Trend Framework was developed, encompassing Mega Trends, Trend Areas, Trends, and Use Cases. This framework provided a comprehensive overview of the evolving landscape and served as a strategic guide for Bayer.
  6. Ongoing Automated Monitoring:
    • To ensure the Radar remains up-to-date, an automated monitoring system was established using SONAR. This continuous monitoring allows Bayer to stay ahead of emerging trends and adapt their strategies accordingly.
  7. Definition of Operating Model:
    • To ensure a seamless integration of the Digital Trend Radar into existing processes and workflows, we defined an efficient operating model.
  8. Trend Radar Set-Up:
    • The development, testing, and launch phases were executed meticulously to ensure the Digital Trend Radar meets the requirements for functionality, usability, and reliability.

Through these comprehensive efforts, we delivered a state-of-the-art Digital Trend Radar. It supports strategic decision-making, providing trend insights and empowering Bayer to navigate the ever-evolving landscape of emerging technologies and trends.

ANWB – Data-driven business transformation to enable growth

PERIOD 2022 – ON GOING

We support the ANWB in gaining a data- and consumer-driven approach to optimize and streamline marketing activities across the organisation and identify new future value spaces. Throughout the transformation process we realized short term efficiency gains of approx. 10% and mid-term value uplift of approx. 20%.

GOAL

Four major objectives:

OUTSIDE-IN PERSPECTIVE:

A neutral outside-in perspective on consumer needs & market environment

BUSINESS OPTIMIZATION:

Short-term optimization of marketing efficiency to finance growth

FUTURE VALUE SPACES:

Identify & prioritize potential propositions for new market opportunities

TRANSFORMATION STEERING:

Enabling the organization to actively steer acquisition, customer journeys and brand


CHALLENGE

Bring transparency into the value creation mechanics and growth drivers along the business lines and related products & services. Create a data-driven mindset and accountability to support the reorganisation into a platform organisation.

SOLUTIONS

Setup of a roadmap to support the 4 major goals:

1. Using digital data in particular organic search behaviour and social listening to get a

“voice of the customer” resulting in a detailed picture about:

  • Consumer interests & needs
  • Category opportunities & product interest
  • Communication perception & seasonality
  • Brand advantage
  • Competition and new players
  • Future value spaces & propositions

2. Setup and conduction of several Commercial Effect Modelling to identify and quantify direct and indirect (e.g., brand halo) revenue drivers across all anwb business lines (roadside assistance, travel, insurance, retail) as well as derivation of business line specific growth capture plans

  • Strategic budget allocation across all business lines
  • Optimization of paid and owned media, campaign and promotion in terms of timing, frequency, channel and content leading to 20% short- und 30% mid-term value increase
  • Entry point, customer journey and product assortment refinement

3. Outside-in evaluation of trends and market opportunities by a diffusion analysis based on

a) expert data (patent, scientific publications and mass media via TD Trend Sonar) and

  • b) consumer data (organic Google search)
  • Strategic trend framework along the anwb business lines and mapping on anwb value spaces
  • Prioritized value spaces/innovation funnel for annual business planning
  • Briefing input for service and product development teams

4. Derivation of a KPI set from the Commercial Effect Models and integration into a Business Steering Cockpit consolidating internal and external data sources

  • Value focused goal set with corresponding forward-looking KPIs for the entire leadership team.
  • Steering dashboard for the reorganized anwb platform organization
  • Collaborative features and engagement processes to bring the people behind the vision and drive data culture.
  • Ongoing optimization guidance for selected strategic initiatives (e.g., retail assortment, new energy proposition, owned channel architecture)

ADIDAS – Global Share of Search Program 

PERIOD 2021 – ON GOING

TD Reply has partnered with adidas to establish and manage a comprehensive global brand tracking initiative. The brand share of search program capitalizes on organic search data, integrating ‘Share of Search’ as a pivotal Key Performance Indicator (KPI) for assessing Brand Strength. By tracking organic search data, adidas gains a more precise, data-driven understanding of their brand’s position in a volatile and ever-evolving market. The comprehensive end to end tracking approach of the program facilitates informed, strategic, and operational decision-making and improved adidas’s ability to effectively adapt and thrive in a dynamic business environment.

CUSTOMER GOAL

The main objective of the project was to define a more flexible, granular, and near-time Brand KPI System that leverages digital consumer data to track brand strength, improve measurability and meets the following objectives::

Strong link to business impact :

(Strong correlations of KPI with commercial data and early indication for changes in sales and market share)

Adaptive to changes in the marketplace :

(End to End tracking from an overall Brand level, down to categories, products and partnerships)

In-time Insights:

(high frequency of data gathering to enable attribution of single activities on brand interest)

Efficient to execute :

(Significant cost advantages over existing survey-based Brand Tracking Solution and higher flexibility to adjust scope with access to historical data) 

Global Scalability:

(Global Coverage including Baidu, Good Capture of Competitors, Robust Data Volumes across countries) 


CHALLENGE

1. Developing a Global End-to-End Tracking System: One of the most daunting challenges was the creation of a tracking system capable of accurately capturing market realities on a global scale. The system had to be:

Scalable Across Diverse Geographies and Languages:

Covering 29 countries and 19 languages, the system required immense scalability and adaptability to accommodate diverse market nuances and linguistic variations.

Comprehensive for Multiple Categories, Brands,

and Products:

With the inclusion of 8 categories, over 70 brands, and hundreds of products, the complexity of the system increased exponentially. It needed to be robust enough to handle vast amounts of data while remaining precise in its analysis and reporting.

2. Balancing Stability with Market Dynamics: Maintaining the stability and comparability of the tracking system, while adapting to ever-changing market dynamics, was a critical balancing act:

Integration of New Market Elements:

The system had to be flexible to integrate new brands, products, and partners without disrupting the existing data’s consistency and historical comparability.

Responding to Rapid Market Changes:

Adapting to market dynamics in real-time, without compromising the integrity and relevance of the KPI, was paramount in a swiftly evolving market landscape.

3. Senior Management Buy-In: Securing senior management buy-in for a new KPI based on digital
behavioral data was a significant hurdle:

Shift from Traditional Methods:

Convincing the leadership to move away from traditional survey-based data to a more digital-centric approach required a strategic approach.

Proof of Concept for Trust Building:

Extensive validation and testing in the initial Proof of Concept phase were critical in demonstrating the effectiveness and reliability of the new KPI, thereby building trust in the digital approach.

Internal Education and Communication: Post gaining senior management buy-in, the next challenge was to foster a collective understanding of the new KPI across the organization:

Engaging Diverse Departments and Business Units:

Conducting internal presentations and engaging directly with stakeholders across different business units, departments, and markets was key in educating and aligning them with the new system.

Facilitating Open Dialogue:

This approach was not just about information dissemination but also about encouraging open dialogue, addressing concerns, and nurturing a collaborative environment, essential for the smooth adoption and implementation of the new system.


SOLUTIONS

The project adopted a “Pilot, Test, Scale” approach to develop and implement a global brand tracking system using Share of Search as a central KPI. The Share of Search tracking scope covers 70+ competitors, 8 categories, and hundreds of products and partnerships in 18 languages and 29 international markets (including China through Baidu Search).We embarked on defining a collective vision spanning the entire brand communication landscape.

1. Share of Search Pilot and Extensive Validation:

The initial phase focused on evaluating the feasibility, validity, and relevance of using organic search data as a Brand Strength KPI.

Building a Comprehensive Topic and Keywords Framework: A detailed framework was created encompassing over 600,000 keywords across brands, categories, partnerships, and products. This framework enabled granular insights across key layers such as Brand, Category, Product Type, and Franchises.

• Data Gathering, Analysis, and Sense Checking: The process involved collecting and analyzing data on both brand and category levels, including volume and share rankings, and search-sales correlations.

• Evaluation of Results Based on Success Criteria: The success of the pilot was measured against several criteria, including the ability to reflect market realities, tie to business performance, adapt to market changes, provide timely insights, execute efficiently, scale globally, and capture competitors effectively.

The outcome of this phase was the successful buy-in from senior management for a global rollout, underpinned by the pilot’s demonstration of the system’s efficacy and scalability.

2. Global Share of Search Program Implementation:

With the successful completion of the pilot, the project moved into its global implementation phase.

Market Rollouts: The program was expanded to additional markets, tailoring the approach to each region’s specific needs.

Tracking Scope Definition, Governance, and Management: This involved establishing the business logic, ensuring comparability, and managing stakeholder expectations across various brands, categories, and markets.

 Infrastructure Development: The implementation included setting up a robust tracking infrastructure with data pipelines, dashboarding, and reporting formats.

Target and Ambition Setting: Based on historical growth patterns and projections, target ranges were defined using different growth functions.

Ongoing Reporting and Insight Generation: The system provided continuous reporting, offering crucial insights for strategic decision-making.

 Informing Annual Strategy Process: Insights generated from the system played a key role in shaping the annual strategy process, aligning it with real-time market dynamics and brand performance metrics.

By adopting this structured approach, the project successfully implemented a comprehensive, scalable global brand tracking system. It not only provided valuable insights into brand performance but also equipped the organization with a dynamic tool to navigate and adapt to the ever-changing market landscape.



RESULTS

  • Highly granular and comprehensive End to End tracking solution 
  • Significant cost advantages over existing survey-based Brand Tracking Solution 
  • Increased flexibility to continuously adjust tracking scope with access to historical data
  • Leading indication for changes in sales and market share 
  • Better attribution of single activities on brand interest.

DIGITAL TOOL AND METHOD BOX

PERIOD: 2017

GOAL

  • Use smart data to enhance traditional market research
  • Identify new sales stories, trends, and consumer needs based on search data
  • Support content and marketing managers in their decision-making

CHALLENGE

  • Handling the analysis of over 6,500 search terms
  • Automating the smart data approach

SOLUTION

TD Reply helped DekaBank to identify trending marketing content that is relevant to the needs of its existing and prospective customers. Analyzing over 6,500 search terms and applying various algorithmic approaches, TD Reply created consumer-centric sales stories that were highly relevant to the target audience.

Thanks to an easy-access insights dashboard,” DekaBank employees can keep track of current consumer trends based on Google search data. Additional 360-degree trend reporting allows DekaBank to take deep dives into its potential content fields.

The dashboard thus allows DekaBank to monitor the financial needs and business growth of its customers and prospects much more frequently and in greater detail than traditional market research would ever allow.

DATA-DRIVEN BUSINESS TRANSFORMATION

PERIOD: 2015–2016

GOAL

In the process of exploring new and innovative ways to initiate a data-driven transformation process, Döhler has three key objectives:

  • Enhancing the decision-making process through data-driven transparency
  • Improving consumer/customer/competitor intelligence by identifying action fields and opportunities
  • Initiating a data-driven business transformation process by implementing a company-wide smart data solution across the entire value chain

CHALLENGE

  • Understanding and assessing the current smart data readiness and information needs of each department, and the development of a smart data roadmap
  • Exploring new data sources and methods to gain insights and evaluate the business value
  • Gathering and analyzing new data in order to identify growth opportunities and action fields (e.g. data as a service) that can be developed outside of Döhler’s core business for commercialization with partners, customers, etc.

SOLUTION

Market Intelligence Dashboard

The automated dashboard reflects company-specific use cases allowing the comparison of different brand and ingredient KPIs to identify market opportunities and trigger innovations and sales. Furthermore, the dashboard provides cross-cutting approaches that ensure holistic ecosystem-based management, including customer, competitor, sourcing, and processing views. The comprehensive source set includes a wide range of data, including Euromonitor data and Mintel insights, as well as buzz and search data.

FRANCHISE LIFECYCLE MANAGEMENT

PERIOD: 2018

GOAL

Due to the critical role of key franchises for overall commercial success (Top 10 franchises make between 40-70% of overall revenue across the top sportswear brands), adidas wanted to gain deeper insights on how franchises should be managed and tracked along their lifecycles in the market.

Specifically, the client wanted to:

  • Understand Market Mechanics
  • Review Competitor Best Practices
  • Improve Franchise Management

CHALLENGE

As the scope of the analysis was on understanding life cycle mechanics and patterns one key challenge was to gain “long enough” historical timeseries data, especially for key franchises that are at later stages of the life cycle.

Another challenge was to identify success factors and leading indicators for franchises success across brands and categories.

SOLUTION

TD Reply conducted an extensive Research Project analyzing 52 Franchises across 7 Brands in 5 sportswear categories.

Understand: Identification and fitting of franchise curves to statistical distributions, clustering of franchises by distribution parameters and splitting life cycles into stages

Locate: Creating a KPI set with benchmarks based on average stage performance to track franchise performance

Predict: Prediction of life cycle development in terms of the height & length of life cycles and stage entries

Steer: Development of lifecycle & portfolio frameworks (product launches, iterations, activations, …) to advance franchise management & planning

The analysis and modeling were based on around 1 Mio. data points and 17 KPIs including internal client’s data (e.g. article invest, sales and discounts), external market data (e.g. market shares) as well as digital data generated by TD Reply (e.g. google search, social media buzz and franchise image buzz).

Results of the project have been presented to the board and senior management across key business units and are used for portfolio and franchise lifecycle management decisions.

DIGITAL BRAND EQUITY TRACKING

PERIOD: 2018–ONGOING

GOAL

Drive strategy and execution from a brand perspective through a digital brand tracking system that can:

  • Measure the level of brand equity for adidas and competitors in near real-time 
  • Gives insights on how products, assets and campaigns are perceived and how they contribute to the overall brand perception
  • Measure the contribution of the brand to overall marketing success and sales

CHALLENGE

The initial challenge was to translate abstract brand personality and image items that have been tracked through traditional surveys into consumer language that enables us to track and analyze consumer conversations related to the brand drivers on a very granular level (besides overall brand perception we are able to understand how individual franchises, drops, technologies, assets etc. are perceived by consumers).

Another challenge within the project was to drive change and acceptance within the organization. As we know data projects are change projects, thus the internal communication together with the owner on the client side, was and still is a key aspect of the overall project.

SOLUTION

TD Reply developed a new approach to track the brand values of adidas and key competitors by analyzing digital data. The process involved the following steps:

Brand Meaning: Itemization of adidas brand values by translating all brand-relevant dimensions into consumer language and identifying key digital brand drivers.

Brand Performance: Data Gathering and analysis of the Digital Brand Perception for adidas and Competitors on different levels:

  • Brand Level
  • Sport Category Level
  • Franchise / Asset Level
  • Campaign Level

Brand Mechanics: Creation of an Input-Output-Outcome Model to better understand Marketing Cause-and-effect relationships and optimize the allocation of brand assets to maximize brand impact.

The approach was developed and tested in the US as pilot market and is rolled out across markets (Germany, UK, France, Japan, China).

CREATING A DIGITAL SERVICE EXTENSION

PERIOD: 2014–2017

GOAL

  • Build and maintain a community that serves as one-stop shop for professionals in the area of cataracts.
  • Continuously optimize the community with respect to content integration and customer experience optimization.
  • Integrate tools and services to deliver on key benefit: a“one-stop solution” to support users in their daily workflow overall, saving their time, improving outcome through shared expertise and enhanced knowledge, increasing efficiency and practice throughput.

CHALLENGE

  • Continuously integrating services previously launched as silos into the cataract community delivering on the promise of a “one-stop solution”
  • Creating editorial plans together with the product teams to continuously provide new and interesting content based on customer behavior and interests
  • Transforming internal thinking from a product driven company to a service driven company

SOLUTION

TD Reply implemented a modular open innovation platform based on LifeRay. The intention was to provide a wealth of content together with the possibility to ask questions, exchange with colleagues and provide access to experts. Content and page structure are continuously monitored to provide insight on how to improve customer experience and deliver content that meets customers’ interests and needs. Community management connected to specialists within Carl Zeiss Meditec guarantees fast reactions to user questions.