Job posting data was tagged to the SSOC 2020, and at SSOC level 5, which focuses on the job role level. This granular classification allows for a clear distinction between different job roles, providing deeper insights into job demand and specific skill requirements.

In the previous edition of the SDFE report, skills extracted from job posting data were classified into the Singapore Skills Taxonomy (SST) level 1 which has 10 clusters, with the inclusion of Apps & Tools to make a total of 11 clusters.

In this edition, a data-driven approach was taken to refine the clustering of skills to six clusters based on broad business functions and critical core skills, reflecting the skill requirements needed across job roles in the Singapore economy.

The six skill clusters are: Business and Financial Management, Critical Core Skills, Engineering and Manufacturing, Information Technology and Data Management, Operational Excellence, and Organisational and People Management.

To keep pace with evolving needs, SSG has expanded its database to encompass over 5,000 unique Apps & Tools. In this report, Apps & Tools were mapped to work functions via job roles. Since each job role was mapped to one work function and each item in the list of Apps & Tools can be mapped to more than one job role, each item in the list of Apps & Tools can be mapped to more than one work function.

To derive the proportion of Apps & Tools used by respective work functions, the total count of Apps & Tools belonging to the work function was divided by the total count of Apps & Tools in 2023.

The 24 Apps & Tools clusters based on work functions are: Accounting, Administrative, Arts and Design, Business Development, Community and Social Services, Consulting, Customer Success and Support, Education, Engineering, Finance, Healthcare Services, Human Resources, Information Technology, Legal, Marketing, Media and Communication, Military and Protective Services, Operations, Program and Project Management, Purchasing, Quality Assurance, Real Estate, Research, and Sales.

For an additional layer of analysis, Apps & Tools were also categorised as AI-related if they contributed to the creation, development, or deployment of AI, leverage AI capabilities, or are associated with AI in general.

In a given year, to derive the demand of a skill cluster, the total count of skills within each skill cluster was divided by the total count of skills across all 6 skill clusters based on job postings in a year and was presented as a percentage. This methodology can be applied at two levels: the job role level, which reveals the relative importance of different skill clusters within specific job roles, and the economy level, which illustrates the distribution of skill cluster demand across the entire Singapore economy.

In this chapter, data from five years (2019-2023) was analysed to provide a time-series view of the skills requirement for job roles. Additional analysis focused on the top 10 in-demand skills within each job role’s three most in-demand skill clusters were also included.

In a given year, to derive the demand of Apps & Tools used by a work function, the total count of Apps & Tools tagged to a work function was divided by the total count of Apps & Tools and presented as a percentage.

The same calculation for demand was done for AI-related Apps & Tools.

Analysis was also done on the top 10 demanded Apps & Tools for any particular job role. This refers to the top 10 mentioned Apps and Tools based on the number of their mentions in job postings for that particular job role.

Job demand was derived by dividing the number of job posting of a given job role in year, by the total number of job posting in the same year and presented as a percentage.

Change of job demand was derived by using Compound Annual Growth Rate (CAGR) based on job postings from 2020 to 2023. This resulted in a percentage that reflected the growth in employers’ demand for the job compounded over four calendar years.

Priority skills are a basket of skills that are deemed as currently in-demand, seeing positive growth or demanded by more than one unique job role. These skills were derived from SSG’s National Jobs-Skills Intelligence engine and validated with expert input from industry, academia and sector agencies. In the various charts within this chapter, SSG chose to spotlight skills whose demand grew in recent years, rather than just skills with a large, existing demand. The spotlighted skills were more likely to see shortages in the present and near future.

The demand of a skill represents the proportion of a given priority skill within the total number of skills listed in the economy group. The formula was derived as:

Transferability of a skill refers to the number of unique job roles that require the skill in that year. In this segment, transferability was calculated year by year based on job postings from 2022 - 2024 and for 2025 using the Skills Forecasting methodology. The formula was derived as:

Skills Forecasting refers to the predicted demand and transferability for 2025 generated by a machine-learning model.

The machine-learning model for Skills Forecasting was trained with historical demand data or transferability and employed Linear Regression to predict the value of all skills for the year 2025. The model used a sliding window approach, incorporating the average demand or transferability of each skill over three-month window, extending up to two years on a monthly basis, thus creating a total of 24 windows as predictors. Additionally, the model took in the embedding of each skill title using Word2Vec, which were also used as part of the set of predictors.

Priority skills were plotted on a scatter plot with their demand as the vertical axis and transferability (i.e. number of unique job roles requiring that skill) as the horizontal axis for each given year. Then, the priority skills were categorised using the 70th percentile mark as a benchmark for both axes, resulting in four quadrants of skills: In-demand quadrant, Up-and-coming quadrant, Niche quadrant, and Transferable quadrant. A given priority skill can be plotted on the Skills Matrix for a given year, into one of the four quadrants. Repeating this for a number of consecutive years, the position of the skill on the Skills Matrix may change, resulting in the skill movement being observed across years.

Creative professionals include job roles whose primary responsibilities involve creative work such as art, design, media, marketing, and event management. They may be working in companies whose core business is in creative fields or creative departments within businesses in other industries.

A total of 71 creative job roles, at SSOC level 5, were identified, of which job posting data of 45 creative job roles were available for analysis.

Based on the job posting data of creative job roles, the demand of a skill cluster was derived by dividing the total count of skills within each cluster from job postings each year, by the total count of skills across all six clusters based on job postings for the same year.

Data from five years (2019-2023) was analysed to provide a time-series view of the job market, comparing skills and economic trends.

The computation of the Apps & Tools was based on the methodology used in Chapter 1.

For analysis on Apps & Tools used by Creative professionals, only job posting data of Creative professionals were used.

For analysis on Apps & Tools used by non-Creative professionals, job posting data of non-Creative SSOCs were used, and those performing work relating to art, design, marketing, and media and communications were further examined.

Creative skills were identified based on skills specific to sectors in Design, Media, Built Environment (Architect) and Tourism (content and experience creation) as well as skills that were tagged to creative job roles relating to Marketing, Branding, Public Relations, Attraction and Event Management.

The transferability of creative skills was based on the methodology used in Chapter 2.

The poll was conducted to capture the general sentiments of Creative professionals on the use of GenAI at work, perceived benefits, potential applications and considerations integrating into work.

It was administered online from 24 Jul to 17 Sep 2024 by inviting Creative professionals through various channels including LinkedIn and professional communities. A total of 104 responses were collected, of which 87 indicated that they were professionals from Arts, Marketing and Media, and their responses were tabulated accordingly.

Job roles that have a) good wages, b) demanded by employers currently, and c) have good growth were identified as jobs suitable for individuals to transition into:

1. Good wages are defined as S$3,300 and above. S$3,300 was selected as a benchmark as it is 65% of Singapore’s Median wage of S$5,197. The median gross monthly income is obtained from Comprehensive Labour Force Survey, Manpower Research and Statistics Department, MOM. The median gross wage for each job role for this study is from the Occupational Wage Survey conducted by Manpower Research & Statistics Department, Ministry of Manpower.

2. Job demand was derived by dividing the number of job posting of a given job role in year, by the total number of job posting in the same year and presented as a percentage.

   

   Good demand is set at 25th percentile of market share, to bring focus on job roles that employers are actively seeking to fill.

3. Change of Job Demand (CAGR) is calculated using the Compound Annual Growth Rate (CAGR) of job postings for each job role.

   

   Using the CAGR of all job postings from 2020 to 2023, we considered job roles with CAGR of 25th percentile and above as those that are generally maintaining their importance in the job market, and not experiencing a big drop in demand by employers.

   Upon applying these filters, 146 job roles were identified as suitable for career transition. While there are 1002 job roles with unique SSOC codes, only 342 unique SSOCs have wage and job demand data relevant for this analysis. Only 5-digit SSOCs with wage information were used for this study.

Explanatory Notes

The Occupational Wage Survey covers a representative sample of private sector establishments each with at least 25 employees. The occupational wage data presented are for full-time resident employees only. This provides a more meaningful basis for comparison of wages across occupations.

Monthly gross wage refers to the sum of the basic wage, overtime payments, commissions, allowances, and other regular cash payments. It is before the deduction of employee Central Provident Fund (CPF) contributions and personal income tax and excludes employer CPF contributions, bonuses, stock options, other lump sum payments and payments-in-kind.

Career pathways comprise of the original role (i.e. the user’s current job role), followed by a first destination role, and a second destination role. The first destination role is a role identified to be potentially suitable role for career transition as mentioned in (A).

These pathways were identified using SSG developed proprietary knowledge graph¹ connecting data on jobs, skills and training courses.

Job transitions were identified based on the following criteria:

1. Job roles with many skills in common: This was measured in two ways, and in both cases greater importance was placed on skills that were considered more “unique”² to a job role:
• The ratio of skills common to both job roles, over all skills required by both job roles, taking into account the importance of a skill to an occupation (weighted Jaccard similarity). This ratio determined the ranking of the adjacent job roles that an original role can transition into. The larger the number, the higher the similarity between the two jobs, which indicated that the original role has more of the skills needed for the next job role.
• The absolute number of skills needed for both job roles
2. Count of Job postings of the destination job. To be included, jobs must have more than 10 postings in 2023 data used for this report.
3. Historical transitions between jobs: This was based on whether there were actual transitions between two SSOCs in MOM’s Occupational Employment Dataset (OED)³ from Dec 2020 to Sep 2022.