Identifying new innovative services using M&A data: An integrated approach of data-driven morphological analysis

Highlights

• We suggest a systematic process for generating new service ideas using M&A records.

• How to extract service actions and service contents from M&A transactions is suggested.

• The QFD structure is suggested for defining possible service types: convinced, promising, and challenging services.

• The morphological analysis is used for combining service actions and contents for each type of service.

Abstract

This study suggests a concrete framework for generating new service ideas using an M&A dataset. Addressing the limitations of previous works that neglected service-specific characteristics, we suggest methods to extract service-specific keywords and phrases from the text and restructure them to provide clear evidence for new service development. Therefore, we propose a process for building data-driven quality function deployment (QFD) and data-driven morphological analysis (MA). First, M&A transactions were collected from CrunchBase, which is an open platform that provides start-up information. Service actions and service contents are then extracted from the text using natural language processing. For each extracted keyword, a clustering analysis was performed to identify the new service patterns. For clustered service actions and contents, MA is employed to generate new service ideas. This study contributes to the technology management field by first employing M&A records for the data-driven morphological matrix and suggests how to extract service actions and service contents from the text. We also suggested a new systematic way of identifying new services using an integrated approach of QFD and MA. This work is expected to help managers in new service development by providing practical guidance and tools for utilizing textual data.

Introduction

New service development (NSD) has become critical nowadays as new and innovative services are generated and consumed quickly (Biemans et al., 2016; Kelly and Storey, 2000). While NSD is composed of several sequential tasks, the first step, i.e., new service ideation, is considered to be a vital process that determines the success and failure of the entire service (Koen et al., 2001; Lewis, 2001; Brem and Voigt, 2009; Hoegl and Parboteeah, 2007). However, the ideation process is very difficult and highly uncertain. Hence, it is sometimes called the “fuzzy front end” (Brem and Voigt, 2009; Geum and Park, 2016; Oliveira and Rozenfeld, 2010).

Many studies have been conducted to support the ideation process of NSD using concrete frameworks and systematic processes. Various methods, such as case-based reasoning (CBR) (Geum et al., 2016; Wu et al., 2006), theory of inventive problem solving (TRIZ) (Zhang et al., 2005), quality function deployment (QFD), and morphological analysis (MA) (Geum and Park, 2016; Geum et al., 2016; Kim et al., 2008; Lee et al., 2016), have been used actively. Among the many techniques and tools for service idea generation, MA has been employed extensively for new service ideation because of the creative and recombinative characteristics (Geum and Park, 2016). Yoon and Park (2005) suggested a keyword-based MA that supports the identification of technological opportunities. They developed a technology dictionary based on keywords extracted from patent documents and established a morphological matrix. Lee et al. (2017) performed CBR on keywords extracted using TF-IDF from mobile application services. Geum and Park (2016) suggested a new way of building a morphological matrix using WordNet and proposed a new ideation process. Geum et al. (2016) suggested an integrated market-pull and technology-push approach using a morphological matrix. They used this method to generate ideas for smart services.

Despite the significance of previous studies on MA, they are some common limitations. First, from the data perspective, previous studies employed databases, such as patents, application services, and dictionaries, to analyze the current phenomena in technology and markets. Particularly, customer-oriented records, such as customer reviews and crowdsourcing data, were used extensively. Even if these data are worth the effort, they simply focus on the technological perspective (patents or dictionaries) or market perspective only (application services or customer requirements (CR)). However, there is a risk of a lack of feasibility when realizing the service (Poetz and Schreier, 2012). In other words, when designing new services, how existing and new services are perceived in the market and whether the new concepts are feasible in terms of technical and commercial perspectives should be considered carefully. Mergers and acquisitions (M&A) records can help resolve this issue. As M&A represents the business potential and scalability of the acquired company, an in-depth investigation of M&A records enables us to understand what kinds of business models or services have recently become popular in marketplaces and how current businesses can be extended.

From the process perspective, a majority of previous studies focused on keyword-based analysis (Geum and Park, 2016; Yoon and Park, 2005; Yoon et al., 2014; Lee at al., 2017) without considering the service-specific characteristics of each keyword. In many cases, previous studies employed all keywords to calculate the frequency of each keyword in a document instead of extracting service-specific and relevant keywords. However, when designing new services, systematic procedures are required that consider the service-centered view (Vargo and Lusch, 2008; Vargo and Akaka, 2009). A service concept describes what needs to be resolved for customers (Kim et al., 2016; Lim et al., 2018). Therefore, the “what” and “how” of service should be considered completely to design service concepts (Lim et al., 2018). This can be characterized by the service contents and relevant actions; thus, the keywords representing service actions and contents should be filtered from the massive database and used as inputs for MA. In other studies, the extraction of service-relevant keywords should be addressed to generate new service ideas using data-driven MA.

Therefore, in this study, M&A records, which are considered as a prominent database for new service ideation, are used to develop new service concepts. Thus, a framework for extracting service actions/contents from documents is suggested in this study and the extracted keywords are used in new service ideation. Accordingly, we used QFD and MA. Three types of new services were suggested based on QFD utilization, and ideas for new services were suggested for each service type using MA.