Reliable market research and industry sizing is a science. CSA Research’s annual market sizing project benefits language service and technology providers by producing reliable and actionable data they can use in their planning. Buyers of these services, whether enterprises or governmental agencies, use this data to educate themselves on the market. Investors and the media rely on it to frame and evaluate the potential of the sector.
For the last 10 years, CSA Research has calculated the size of the market for language services and technology. We project that it will be US$49.60 billion this year, up from US$46.52 billion in 2018.
CSA Research bases our analysis and sizing on primary market research. We begin with a global supplier database of slightly more than 18,000 companies. We field a comprehensive survey that yields a representative sample. We conduct a time-consuming data collection, validation, and verification process. Each of these aspects of our process are critical to developing a reliable market analysis.
Database. We continue updating and expanding our database of more than 18,000 LSPs and technology companies, as it is a critical piece of our methodology. To qualify as an LSP, companies must have two or more full-time employees, a minimum level of revenue by region, and show visible market activity.
Survey design. CSA Research’s analyst and statistician teams develop our annual Global Market Survey, analyzing and revising it each year to optimize data as well as validate the responses received. Now in its 10th year, the most recent version contains 50 questions and provides us with 165 datapoints and hundreds of possible correlations. To produce trusted and reliable data, and to include companies with fiscal years ending on or before March 31st, we conduct the survey from January through April.
Representative sample. Statistical standards determine the number of responses that constitute a representative sample, considering population size and confidence level in the data. The methods we employ are well-established for studying population dynamics and we strive for a confidence level of 94%.A large representative sample from our database (589 LSPs this year) is a must in sizing a market in which smaller LSPs comprise 85.7% of it.
Data collection, validation, and verification. Since our launch of the current survey model in 2010, we developed and refined guidelines and tools for handling data quality issues, data security and privacy concerns, and duplicate entries. Our process ensures anonymity, checks for acceptable responses, provides for correction, and requires a signed verification for the ranking process. Critically, our algorithm excludes “double counting” of revenue (for example, sales by LSPs to other LSPs).
Algorithm. Once we identify our representative sample, validate responses with our benchmarking tools, and verify data with respondents, we move to our proprietary market-sizing algorithm. It includes 24 calculation areas and runs on custom-programmed SAS statistical software to carry out computations on multiple categories, including regions, service offerings, and technology offerings.
Adjustments. We’re not done yet. Among other sources, we incorporate data from the IMF and World Bank, two of the most reliable sources for worldwide economic data. We factor their data into our sizing model and adjust overall market and regional economic forecasts to account for macroeconomic factors that affect the language sector.
We strive to provide the language services and technology industry with rigorous market research that is at or above the level that other industries receive. Our unique methodology, such as our database and a truly representative sample, creates the complex and trusted primary research necessary for this important industry to thrive.