Data collection of search data retrieved from Google Trends for brands of car manufacturers and their car brands as well as data for the brands of listed companies active in Europe. Data were retrieved using R code for a list of keywords of interest. The aim of this project was to investigate whether data from Google Trends could be used to value brands. Currently brand valuation methods are highly complex and results are generally not available for many brands and are not comparable across brands. We downloaded Google Trends data for brands in the automotive sector and for the brands of listed companies. The data were then joined to other data about the firms and brands we collected from public sources. Analysis of this data showed that Google Trends data provides a good measure of changes to brand value for well established brands with national or global reach in the automotive sector. The data capture changes in brand value that are the result of exogeous shocks, e.g. as a result of product recalls. Google Trends data also contain information about variation in stock market value of listed companies that is not contained in data released through firms' quarterly and annual reports or in other publicly available data on intellectual property rights and citations of these rights. The Google Trends data was complemented by data on car registrations and characteristics from various sources and data on various shocks affecting the car industry. These are not included in this dataset. Brands are names, phrases, symbols or designs that identify particular products, which are often protected by trade marks. Trade marks help owners protect their reputation by forbidding use of their brand name by others. Trade marks are usually registered, but not all registered marks are used. Much litigation arises over the extent of use and reputation of marks, which often results in costly surveys being run with data from trade mark registers and data from Google's search products. Data from Google's search products can be used to measure value and use of marks. In using these data, two important challenges must be overcome: heterogeneity of the data and attempts to manipulate the measures obtained. Both challenges will be addressed in the project. The aim of this project was to improve the measurement of brand value and the measurement of trade mark use by combining data obtained from trade mark registers and data from Google's search products (Google Search, Google Trends and Google Insights for Search). Data on search are heterogeneous because the reasons for searches are heterogeneous, they may reflect positive or negative brand interest. Additionally, attempts to manipulate Google's rankings of websites though the PageRank link analysis algorithm are legion. To overcome both problems we will consider ways of limiting search terms to obtain consistent measures of value from Google's search products.

Data on Google Trends derived by download from Google. The script used is available on GitHub (see Related resources). The initial input is a csv file containing two columns: “id”, which contains an identifier, and “names”, which contains the keywords to search for. The final output is a STATA table with the merged weekly series, each named from the corresponding id. Additional data collected (but not available in this dataset): (1) Automotive Companies: Data on German car manufacturers compiled from data made available by the Kraftfahrtbundesamt, ADAC and the European Commission. Data on UK car manufacturers is compiled from data made available by the DfT and the European Commission. (2) Listed Companies: Data on listed companies comes from AMADEUS, DATASTREAM and COMPUSTAT. Data on intellectual property rights was obtained from PATSTAT and from the European Trade Mark Office. Data on patent citations was created by Dietmar Harhoff.