Salesforce is sold as a set of clouds, each one an application for a part of the customer relationship, and a job post usually names the ones the company runs. Sales Cloud holds leads, accounts, opportunities and forecasts for a sales team. Service Cloud holds cases, the knowledge base and the agent console for a service team, and it is named about as often as Sales Cloud, and most of all at the consulting arms and product companies. Marketing Cloud runs campaigns, journeys and email at scale, and it stands out at the pharma GCCs, where it carries the drugmakers' work with doctors. Experience Cloud builds the portals and communities that customers and partners log into, and it leads at the lenders and the document-signing company. OmniStudio and the industry editions, financial services and health, package the platform for a particular industry with ready-made data models and guided flows, and they are almost entirely an Accenture ask. A functional consultant knows how a lead becomes an opportunity, an opportunity a contract, and a complaint a case, and configures the clouds to run the company's version of that.
The platform is configured first. Objects, fields, page layouts, record types, Validation Rules, Permission Sets and sharing rules are set up through the platform's own screens, and Salesforce Flow is the tool for building automation without code, by drawing the steps a record goes through when it is created or changed. Flow is named in a fair share of job posts outside Accenture, and it has replaced the older workflow rules and process builder that experienced consultants still know.
Where configuration runs out, the developer writes Apex, the platform's own language, which looks like Java and runs on Salesforce's servers. It is the most-named skill in the role after Salesforce itself, and in the developer-led job posts, at the services firms that build deeply and at the GCCs and product companies, it is named in most of them. Apex is used for triggers that run when a record changes, for business logic too complex for Flow, for batch jobs over large data, and for the services that other systems call. SOQL is the query language for reading the platform's data, and SOSL searches across it, and both are named in a fair share of developer job posts, most at the product companies. Because the platform is shared with every other customer, Apex runs under strict limits on how much it may query and process in one go, and writing code that respects those governor limits is the mark of an experienced developer.
The screens are built with Lightning Web Components, the modern framework, and it is the single most-named skill in the whole role, because nearly every developer post asks for it. It is ordinary web development, in JavaScript, HTML and CSS, inside the platform's component model, which is why those three are named in a fair share of job posts and why frontend skills are the most common extra tag in the role. Aura is the older component framework still found in long-running systems, Visualforce the one before it, and Salesforce Lightning the name for the whole modern interface. A developer today builds in Lightning Web Components and maintains Aura and Visualforce pages until they are replaced.
A customer system has to talk to the order system, the billing system, the marketing tools and the data warehouse. The Salesforce API is how other systems reach the platform, and it is named in a fair share of job posts. MuleSoft, which Salesforce owns, is the integration platform named most beside it, with Boomi and Oracle Integration Cloud in a few job posts, and NetSuite appears where the finance system is being connected, at the product companies. Data movement between systems is tagged on a fair share of job posts, most at the product companies and the GCCs. The connector companies, whose product is the link between Salesforce and other business systems, hire for exactly this.
Salesforce sells its own AI layer, and it has become part of the role. Einstein is the older name for predictions and recommendations inside the clouds, and Agentforce is the newer tool for building AI agents that answer customers and act on records. Agentforce is named in a small share of job posts overall, but in most at the consulting arms, where consultants are selling AI agents built on Salesforce. It draws on the same Apex, Flow and data skills, with the agent's instructions and the actions it may take configured on the platform.
Salesforce development has grown a release process like ordinary software, and build and release is tagged on a large share of job posts, and on most at the services firms that build deeply and the product companies. SFDX, the Salesforce developer toolkit, lets code live in Git and be deployed from the command line. Copado is the release tool built for Salesforce, and it is named most at the services firms. Azure DevOps, Jenkins and GitHub Actions carry the pipeline where the company uses a general tool. Code moves from a developer's sandbox through test sandboxes to production, and Apex tests must cover a required share of the code before the platform will accept a deployment, which is why testing is part of every developer post whether or not it is named.
SQL is named in a fair share of job posts outside Accenture, and databases are tagged most at the product companies, where MongoDB and Oracle appear beside the platform for the systems around it. The cloud is named in a smaller share than in engineering roles, because Salesforce hosts itself; AWS appears most, mostly at the product companies and GCCs where the surrounding systems live. Java is named in a fair share of job posts outside Accenture and C# and .NET in a few, where the engineer also builds the applications around the platform, and Angular and React where a company's own portal sits in front of it. Reporting with Tableau, which Salesforce owns, is tagged on a small share.
Underneath all of it sits the way platform work is done: requirements gathered from sales and service teams, configuration and code built in a sandbox, tests written to the platform's required coverage, releases moved through sandboxes to production, and three platform upgrades a year that the company's customisations have to survive. Job posts count these as given and rarely list them as skills.
A Salesforce engineer who knows Sales Cloud and Service Cloud, can configure objects, rules and Salesforce Flow, writes Apex that respects the platform's limits, queries with SOQL, builds screens in Lightning Web Components and can maintain Aura and Visualforce, integrates through the Salesforce API or MuleSoft, and releases through SFDX with Copado or a pipeline, meets the core of nearly every job post. The variations belong to the employer. Accenture, the widest door, wants Apex and Lightning Web Components across the Service and Sales clouds, with OmniStudio and the industry editions, and often describes the role in general terms. The services firms that build deeply want senior Apex and Lightning developers with Copado, Jenkins and the most release process. The consulting arms want Agentforce with Service Cloud. The GCCs of pharma groups, banks, insurers and hotel and property firms want developers who run the parent's customer platform, with Marketing Cloud at the drugmakers and Java and AWS beside the platform. The product companies that run on or build for Salesforce want SOQL, Experience Cloud, integration and the most database work. Across all of them, the engineer who can configure the platform and then code past its edges is the one every job post describes.