Innovation is essential for economic growth, technological competitiveness, and sustainable transformation. However, measuring innovation remains a complex challenge for researchers and policymakers. Although patents have long been used as an indicator of technological innovation, they only capture part of the process. Patents primarily reflect technical invention rather than market introduction or service-related innovation. By contrast, trademarks are the most widely used form of intellectual property right and provide valuable insights into market-oriented and non-technological innovation. Building on this distinction, this dissertation explores how patent and trademark data can offer a more comprehensive and reliable overview of innovation. This dissertation combines methodological, conceptual, and empirical perspectives to improve our understanding of how intellectual property data can be used to study innovation. The dissertation develops approaches for linking patent and trademark data to firm-level information. It also explores the value of this data as an indicator of different types of innovation and examines how it reveals distinctive innovation patterns. The dissertation presents these approaches across five studies and applies them empirically using large-scale datasets that integrate patent, trademark, and firm information. The first study establishes a transparent and reproducible process for linking patent data from the PATSTAT database with firm-level data. Using a sample of over ten thousand German manufacturing firms, the study introduces a seven-step matching framework that combines algorithmic and manual processes. This approach improves the reliability and traceability of the results, providing a robust basis for analyzing innovation behavior at the firm level. The second study provides a comprehensive review of trademark-based indicators, which remain underutilized despite their potential to capture product innovation, branding, and commercialization activities. It identifies nine key data sources and over forty distinct measures, classifying them into seven thematic domains. It demonstrates that trademarks can provide further insight into firms' innovation and market strategies, and it offers guidance on how to retrieve, clean, and apply such data. Furthermore, the study explores the potential applications of AI and language models in analyzing the textual and visual characteristics of trademarks. The third study examines the potential of patents and trademarks as indicators of environmental innovation. Using matched firm-level and survey data from Germany and Italy, the study assesses the reliability of these indicators in identifying firms engaged in green innovation. The results show that patents and trademarks capture different aspects of innovation: patents reflect technological invention, whereas trademarks indicate market-oriented and sustainability-related activities. The fourth study examines how the knowledge embedded in inventions influences the discontinuous potential of climate change mitigation technology patents. Using advanced text-based similarity analysis, the study distinguishes between inventions that draw more strongly on basic research and those that draw more strongly on applied technological knowledge. The results show that inventions with a stronger orientation toward basic research are more likely to generate technological discontinuities, while those based on applied knowledge tend to reinforce existing trajectories. Compared to traditional citation-based measures, the study demonstrates that content-based indicators can provide deeper insights into the knowledge foundations and transformative potential of innovation. The fifth study examines the technological innovation patterns of Hidden Champions, which are highly specialized, export-oriented niche leaders that form the backbone of the German Mittelstand. Based on patent data, the analysis shows that Hidden Champions are more active in technological innovation and exhibit greater depth and specialization than comparable firms. However, their strong reliance on internal knowledge sources limits their innovation efficiency and results in predominantly incremental rather than radical innovation. These findings empirically support long-standing theoretical assumptions about the innovation behavior of Hidden Champions and demonstrate how patent indicators can identify firms with distinctive innovation profiles. Together, the five studies offer substantial contributions to our understanding of innovation and how to measure it. They offer practical guidance for creating reproducible IPR datasets and summarize how trademark indicators can be used in innovation research. Additionally, they demonstrate how patent-based measures can be applied to study sustainability and knowledge orientation. The findings emphasize that the usefulness of IPR indicators depends on the context, as patents and trademarks capture different aspects of innovation that vary across industries and firm types. Furthermore, the dissertation shows that text-based similarity methods can enhance innovation analysis by revealing the knowledge bases and broader potential of inventions.