FACTOR ANALYSIS OF THE PHENOMENON OF MASS LAYOFFS AT STARTUPS: MIXED METHOD APPROACH WITH STRUCTURAL EQUATION MODELING

ABSTRACT


INTRODUCTION
The growth of Indonesian startups in 2019 was ranked fifth in the world after Canada, India, the UK, and the United States. This growth is also accompanied by the growth of startups based on unicorns and decacorns from Indonesia, meaning that the quality of startups in Indonesia can compete globally with a valuation of more than 1 million US dollars and more than 10 million US dollars (Kominfo, 2020). However, recently there has been a reverse phenomenon, namely the termination of employment (layoffs) and changes in the business strategy of several Indonesian startups. Some startups that do massive layoffs include Tanihub, Link aja, etc. This phenomenon is interesting to study in more depth, seeing the relatively short business cycle and changes in investor behavior to invest in new startups. Based on CNBC (2022), the phenomenon of layoffs or layoffs at startups can be seen from several aspects, including: first, a business strategy that is not right at the beginning, such as burning too much money. Second, there are not many skilled IT human resources in Indonesia, so high costs are needed, and global issues such as the outbreak of the Russia-Ukraine war and the dynamics of covid 19 are still present.
The declining performance of startups can be seen as impacting the decline of all businesses that use technology-based. Therefore this phenomenon can be categorized as a revolution in the startup world, which will filter out what startups can survive and cannot. Initial startup growth was driven by ample funding from investors and powerful strategies to encourage growth in various ways, even at high costs (Djaja, 2019). However, the condition of available human resources is still limited, so the financing offer is also high. The phenomenon of massive startup growth is not balanced with the availability of qualified human resources. Macroeconomic factors such as the spread of the COVID-19 virus and the Ukraine-Russia war outbreak have influenced the development of startups globally. The fluctuating virus has both opportunities and threats to startup survival, such as changes in consumer behavior toward health, hygiene, and lifestyle mindsets (A. Ika, Fahrika. Juliansyah, Roy, 2020). The macroeconomic factor that has been very prominent lately is the outbreak of war between two countries which has resulted in the supply of raw materials for wheat and mining as well as supply chain relationships for other raw materials between several countries. It triggers economic contraction in several countries, affecting global investors' purchasing power and behavior (Nikensari, 2018).
On the other hand, the bubble burst phenomenon is said to be the driving force behind many startups that have gone out of business lately. The bubble burst is the emergence of many new startups and high economic growth but a relatively rapid decline in market supply. So that the supply in the market no longer gets a response / high purchasing power, which causes the product/company to experience a decrease in demand (Nurhadi, 2022). These factors need to be studied in depth to discover startups' growth and defense patterns. Then conclusions can be drawn as to what kind of startups can survive amid the layoff phenomenon that is currently happening, even though startups have an essential role in the growth of MSMEs and the economy in Indonesia. Although several things threaten and disrupt MSMEs, there are also positive roles, such as helping MSME business growth and processes increase rapidly, according to the type of business suitable for similar startups (Sharma, 2020). This study will be carried out by observing the startup layoffs phenomenon and make confirmation of the factors that affect the startup layoff phenomenon. Observations for this study focused on general cases in public to better understand influencing factors. In addition, researchers try to understand how these layoff factors impact changes in startup strategy behavior. So it can be concluded that the startup strategy is to quickly adapt to increasingly uncertain global conditions.

Business Model Startup
The startup business model is generally a technology-based company that produces software to create, deliver, and capture added value (Alesander, 2010). A startup's business model is needed to provide a view of the business processes that will be carried out and planned or even explain the ongoing business. An example is a startup that is engaged in the form of a business to help analyze problems and find solutions.
New startups can use business models to design forms that sync with their startup conditions (Mardi Arya Jaya, Ridi Ferdiana, Silmi fauziah, 2017). A startup needs a perfect example of business to increase competitiveness and achieve its ultimate goal (Uzzaman, 2015). Supporting factors for startup success mean good business examples that can convey companies' views about how the market responds to their products, weaken the competitiveness of competing companies, and encourage alliances to use competing companies (mutualism) to form products that can meet market needs. These factors can be the key for the company to win the competition in the market.

Macroecnomic Factors
Macroeconomics is the study of economic discourse as a whole. Macroeconomics describes economic changes that affect many societies, companies, businesses, and markets. Macroeconomics can be used to analyze how best to influence policy goals such as economic growth, price stability, employment, and achieving a sustainable balance sheet. Macroeconomics is closely related to state finance issues. Economic changes in a country will impact a company and its market. Macroeconomics can affect economic growth, employment, price stability, and the achievement of equilibrium (Veritia, S.E., M.M. et al., 2019). Global issues currently influence global policies, such as covid-19, the Russia-Ukraine war, and other environmental issues. Since the emergence of Covid-19, various countries have taken policies to reduce the number of spreads so that it affects all forms of business and business in a country. The Russia-Ukraine war is not expected to involve other countries. However, the global impact is the primary raw materials these countries have, such as wheat, mining, and oil. Hence, some countries have to take a strategy to keep supplies running without getting involved in a war between the two countries. The next global issue is the environmental issue currently being felt globally, such as the reduction in green space and the demand for behavior change to reduce carbon so that environmental sustainability is maintained (Berry A. Harahap et al., 2018).

RESEARCH METHODS
This study uses a combination of the qualitative-quantitative sequential exploratory design method, which aims to map the problem qualitatively and is strengthened by quantitative analysis (Sugiyono, 2017). This study uses qualitative and quantitative secondary data. Qualitative data comes from journal articles, books, news, interviews, and personal opinions from startup actors in public spaces such as social media. Quantitative data was obtained from the data aggregator website and search engines. Quantitative data complement qualitative data, which aims to strengthen qualitative data. The quantitative data is panel data on startup layoffs from 2020 -2022 from various startup companies worldwide.
To identify qualitative data that is relevant to the research problem, the researcher uses a structured approach from Webster and Watson (2002), which includes 1. Searching using relevant keywords on well-known journal/news sites; 2. Selecting publications/data that match the criteria; 3. Quickly scan publications/articles through titles, abstracts, and content relevant to the research problem; and 4. Study and analyze in detail the selected publications/articles. Search using keywords including but not limited to "Startup layoffs," "Startup Bubble," and "layoff." The scope of the search is limited to reputable journal databases, search engines, and data mining from social media, Linkedin, and Twitter. The collected qualitative data is then categorized into three main categories, namely: 1. Literature review on startups: publications related to startups covering business models, nature, and mechanisms of startups. This information becomes the basis for researchers as a starting point for obtaining an overview of startups as research objects. 2. Information related to layoff phenomena: publications related to startup layoff phenomena and related matters regarding startup layoffs. This information provides insight into the variables related to the research problem. 3. Relevant macroeconomic information: publications related to external macroeconomic factors related to the startup business climate. This information is helpful in providing an overview of macroeconomic dynamics in the phenomenon being studied After the data is collected, the stages of data analysis are as follows: 1. Qualitative Analysis Qualitative analysis is carried out by studying the variables formed from the literature study and the qualitative data collected. The opinions of startup actors and those related to the mass layoff phenomenon are observed, filtered, and synthesized to form a narrative of factors that play a role in the mass layoff phenomenon.

Formulating variable framework
The variable framework is needed to determine the logical structure of the research and determine hypotheses between variables. The variable framework is determined based on previous research and empirical evidence from observation and expert opinion obtained from qualitative data collection. The formed variable framework shows a description of the relationship between variables, which will be used as the basis for quantitative analysis. Based on the observations and data analysis, a variable framework is formed as follows:

Quantitative Analysis
Quantitative analysis to be carried out is descriptive analysis and statistical analysis. Statistical analysis was conducted to test the variables using Partial Least Squares -Structural Equation Modeling (PLS-SEM). This method was chosen because the research is predictive and exploratory (Yamin, 2021). Furthermore, the advantages do not require a normal distribution (Harahap, 2020). The statistical analysis stage uses the smartPLS application with the following stages: a. Predictive Relevance Test Predictive Relevance is useful for testing the accuracy of the model. Predictive relevance is measured using the value of Q2. The value of Q2 > 0 describes the data point the model can predict well. A predictive relevance test is carried out using the blindfolding technique in smartPLS b. Hypothesis test The research hypothesis was formed after knowing the factors involved; the hypothesis was obtained after qualitative analysis as follows: Hypothesis 1: There is a significant effect of funding aspects on the number of startup layoffs. Hypothesis 2: There is a significant effect of the type of industry on the number of startup layoffs.

RESULTS AND DISCUSSION Qualitative Analysis 1. Layoff Strategy is Implications from Multifactor
The upward trend of startup layoffs began to rise this May. The increase in searches for the words "PHK" and "Startup" rose sharply in May in line with the hectic news regarding layoffs in the startup industry. Businesses that are still trying to form a good concept must make various adjustments to survive in the era of digital competition that is constantly evolving (Haryanto, 2022). Meanwhile, the CEO of MNC Group believes that the cash flow management strategy at the expense of profit has failed to gain market share, so funds from investors have decreased and forced startups to adopt a downsizing strategy (Nurrahman, 2022). CNBC believes that the cause of the layoffs is an attempt to reduce operating costs amid the declining flow of funds from investors over fears of a spike in inflation in the United States (Dewi, 2022). This opinion is in line with Wheelen et al. (2018), where the retrenchment strategy is taken when the company has a weak competitive position and is restructuring by reducing employees. This strategy is the choice for several reasons. Companies that implement employee reduction in their retrenchment strategy to narrow their business focus are proven to be able to improve their performance. However, the reduction of employees must be carried out carefully because situations can occur where the retrenchment strategy will further weaken the company (Wheelen et al., 2018).

Venture Capital Funding and Growth Priority Over Sustainability
The venture capital funding model of startups that dominates the technology sector today focuses on exit strategiesnamely, how founders and funders sell their investments in startups (Lemley et al., 2021). In line with Blank's definition of a startup, it is a temporary organization that is not planned to continue to exist (Blank, 2013). This is because most startup business cycle planning adopts the entrepreneurial innovation model from Aulet, which consists of 4 stages: startup, transitioning, scaling, and exit (Aulet, 2013). In contrast to conventional business, which is more in line with more classic models such as business life cycle theory. Although the decline stage is mentioned, the focus of business establishment is not at the decline stage. A funding model that focuses on an exit strategy gives venture capitalists a clear focus on the time limit of their funding (Gomper, 1999). This will pressure the startup to grow at a predetermined scale quickly. So startup growth indicators, especially those using venture capital funding models, are more focused on non-financial indicators such as the number of employees (Davila et al., 2003). Currently, a more contemporary startup valuation method is being developed to solve the regular harmful cash flow practice to gain market shares, such as Cost to Duplicate, acquisition rate, and Market Multiple (McLure, 2021).
Focus on increasing growth above profits for early-stage startups, forcing startups to adopt a cash-burning strategy to meet investors' growth expectations. As a result, they have minimal restrictions on spending behavior. This strategy has been very successful in the era of abundant liquidity for startups in the last five years. The high operating costs put pressure on the balance sheet so that startup liquidity must be adequate in the future until they make a profit or get new funding. The startup's resilience to negative cash flows is measured by the burn rate ratio, liquidity divided by costs over a certain period (Kemell, 2020). The availability of this liquidity is highly dependent on external parties such as venture capital when the company has not yet posted profits, so it is very vulnerable to adopt a retention strategy if funding is unstable. While venture capital invests less in good ideas and teams, it invests more in good markets (Zider, 1998), so the availability of startup funding from venture capital is highly vulnerable to macroeconomic volatility.

High Staff Cost
The IT and computer sector is an industrial sector with a salary above the average of other industries and is ranked 2nd in salary based on the industry sector. It is mainly dominated by startups (Jobstreet, 2021). For C-Level itself, the average startup sector salary is the highest compared to other sectors. Based on startup job positions sourced from the Monk's Hill Ventures Southeast Asia Tech Talent Compensation Report, the lowest salary for junior-level UI/UX designers reaches 8.7 million per month. The highest is for data scientists at 16.7 million per month. From the same source, the average salary for professionals in the startup industry is higher than in other industries due to several factors, including intense talent competition, high workload and demands, higher risk than established companies, and higher salaries for technical workers than workers non-technical (Monk's, 2021). Bank Indonesia predicts Indonesia's digital economy to grow 650% by 2030 (BI, 2021), so it requires digital talent to play a role in developing the digital economy.
The availability of digital talent in Indonesia is still not sufficient. Research from AWS and Alphabeta, only about 19% of the entire workforce has digital skills. Meanwhile, 110 million new digital talents will be needed by 2025 (Burhan, 2022). Input from universities is also not optimal, considering that only 20% of universities in Indonesia have study programs in the fields of Information, Communication, and Technology. Tech Manager Robert Walters Indonesia also mentioned that apart from the lack of talent availability, the high turnover rate of up to 30% in startups causes high employee costs. A standard industrial practice when employees apply for salary at a new company is a salary increase of 15%-30% (Avego, 2021). The high turnover in the startup sector causes salaries to rise very quickly for the same position. So there is an increase in the market salary for job positions in high demand. High liquidity in startups and pressure to improve non-financial metrics for startups that have not yet recorded finances have given startups more financial space in offering salaries. Not infrequently, this more space causes the hijacking of startup employees because one of the causes of employee hijacking is the high demand accompanied by the scarcity of supplies in specific industries, which in this context is very visible in the technology industry (Indeed, 2021).

Macroeconomic Factors
Macroeconomic factors affect all aspects of business, especially internal variables related to changes in external variables. Concerns about uncontrolled inflation caused the American central bank to adopt a tight monetary policy by raising interest rates. Along with a tighter interest rate policy, in December 2021, the Central bank announced it would indefinitely stop buying debt securities at the end of March 2022. The cessation of buying bonds, called tapering, has been proven in developing economies to cause global risk aversion, sharp and rapid exchange rate depreciation, price corrections for risky assets such as stocks, and capital returns to developed countries (Mishra, 2014). Rising interest rates make investors more risk-averse and adopt defensive strategies. Consequently, investors reconsider their investment decisions to move funds from riskier assets to safer ones. This makes investors rethink investing in venture capital, a venture capital portfolio that is a high-growth industry with high risk. The above factors cause a liquidity crisis in venture capital, making it difficult for startups to get funds.

Quantitative Analysis 1. Descriptive Analysis a. Startups companies executing layoff by year
Source: layoff.fyi, 2022

Figure 2. Number of Startups Companies Executing Layoff by Year
In 2020 and YTD July 2022, the number of companies that made layoffs was higher than in 2021. This is because in 2020, startups were affected by the pandemic crisis, and in YTD July 2022, mass layoffs were due to the shadow of the global crisis. The low number of companies making layoffs in 2021 is because, by 2020, the pandemic has already been absorbed, so that in 2021 there will not be many layoffs. b. Employee laid off by startup industries Source: layoff.fyi, 2022  The food industry ranks first in the number of laid-off employees. This is due to the accommodative barrier to entry for food startups if they are going to expand overseas. Rapid growth causes rapid employee growth, so the impact will be felt if a layoff occurs. One of the most significant contributors to the layoff of food startups is Getir, a food delivery startup from Turkey with 4,500 employees, almost 30% of the total layoff. The second largest position by industry is the financial sector. This is because the fintech sector initially was not as tightly regulated as the conventional financial sector, causing the high growth of fintech startups. The most significant layoff contributor from the financial sector is Paisabazaar digital bank from India, with 1,500 employees. The two companies implemented a layoff strategy to cut costs to maintain business continuity. c. Employee laid off by funding stage Sumber: layoff.fyi, 2022 Figure 4. Number of Employee Laid off January 2020 -July 2022 The number of startups that make layoffs based on funding shows that the most significant number of layoffs is in series E, followed by series B and series D. This data shows that startups that receive much funding do not have a layoff impact as significant as series J. This could be due to companies receiving series B funding to series E funding, which is medium-sized companies, so the financial resilience and customer base are still not firm. So startups are vulnerable to adopting a layoff strategy for cost efficiency. while companies that are more advanced in the funding stage mostly already have a mature market, so they can afford to avoid cost efficiencies to survive.

Statistical Analysis
This research employs the Partial Least Square -Structural Equation Modeling method. Based on the previous analysis, the following model was formed to illustrate the relationship between variables. Industries have a significant impact because, during specific crises, some industries are more severely affected. The global pandemic crisis that has caused restrictions on mass mobility has affected industries that depend on the real sector, for example, the food industry, retail sector, property, travel, and logistics (ECB, 2021). Meanwhile, in 2022, the global economic slowdown caused by the Russian war, supply chain bottlenecks, and the shadow of inflation in developed countries led to rising interest rates which impacted the financial sector and slowed investment (World Bank, 2022) so that the results of statistical tests show that the type of industry significantly affects the number of employees in layoffs.
The funding aspect has a significant influence on the number of layoffs. Startups that receive operational funding from venture capital have a faster growth rate than companies that do not (Inderst et al., 2009). Prominent startups can potentially have more employees and a more significant impact if the company adopts a layoff strategy to save costs.

CONCLUSION
Based on the results and data collected, it can be analyzed that high operational costs, startup dependence on venture capital funding, rapidly rising employee costs, difficulty in funding due to macroeconomic factors, and weak demand have caused startup companies to adopt a retention strategy, namely reducing the business volume to be able to grow their business. Macroeconomic factors became the initial trigger for the mass layoff. Macroeconomic factors cause difficulty in liquidity in the venture capital market and cause startups to adopt a retention strategy by cutting the number of employees. Cutting the number of employees is the most rational considering the high employee costs for startups due to several factors.
From the quantitative analysis, it can be concluded that the variable aspect of funding and the variable of the industrial sector has a significant influence. The funding aspect plays a role in the size of the startup related to the number of employees, so the impact of the layoff will be significant if the number of employees owned is also large. Meanwhile, the industrial sector has a significant impact because specific industrial sectors will be affected depending on the type of macro influence, so industrial sectors that are sensitive to the global crisis will be affected more strongly than those less sensitive to crises.