Certifications

Most Salesforce problems that appear in year three were created in the data model in month one. This credential covers large data volumes, ownership skew and when data belongs outside the platform. In enterprise orgs with millions of records it decides whether reports, sharing and integrations still perform.

Salesforce awards this one for holding the Data Architect, Sharing and Visibility Architect, Platform App Builder and Platform Developer I credentials together. It stands for one judgement: configure, code or leave it out of Salesforce. Getting that call right early saves the most money in any project.

Who may see which record is a business question with legal weight. Role hierarchies, dealer access and works council requirements have to fit into one sharing model. A wrong model either leaks data or forces workarounds that users stop trusting.

Salesforce rarely stands alone: ERP, data platform, telephony and portals all exchange data with it. Whether an interface runs in real time, in batches or through events is an architecture decision with consequences for cost, error handling and operations. This credential covers those patterns and their limits.

Enterprise users log in through Entra ID or Okta, partners and customers through their own identities. Single sign-on, provisioning and external logins have to work together without locking people out or exposing data. Mistakes here surface on go-live day, in front of everyone.

Salesforce awards this one for Integration Architect, Identity and Access Management Architect, Development Lifecycle and Deployment Architect and Platform Developer I. It describes the view from outside: how Salesforce sits in the wider system landscape and how it is built, secured and shipped there.

When several partners work in one org, the release process decides more than any single feature. Branching, environments, test strategy and deployment order have to hold up with a release every two weeks. This is where I spend a large part of my governance work.

Manufacturers run processes from lead to quote, order, installed base and service, often with dealers in between. This credential tests designs that span several clouds and external systems. It matches the projects I work in, where a decision in sales shows up again in after-sales.

Every solution has to be run by someone after go-live. Knowing the admin side keeps designs operable for the internal team, with users, permissions and data maintenance they can handle themselves.

Dealer, partner and subcontractor portals put Salesforce data in front of people outside the company. Licensing, external sharing and branding have to be settled before the first page is built. A portal that partners do not use costs licences and brings nothing.

In manufacturing, after-sales often carries the margin. Case routing, entitlements and SLAs are the base that field service, telephony and portals build on. If that base is unclear, every later extension inherits the problem.

Much of what teams code in Salesforce can be configured with standard tools. Knowing those tools well keeps custom code small, and small custom code keeps upgrades and handovers cheap.

International sales teams need one shared process with country variants handled as configuration. Opportunity stages, forecasting and territory design decide whether management can trust the pipeline figures.

As an architect I review code that partners write. Reading Apex and understanding governor limits lets me spot designs that will fail under load before they reach production.

Mature orgs collect hundreds of flows, validation rules and permission sets. This credential covers the advanced tools to keep that manageable, so the org can still be changed safely after five years.

Field service connects installed assets, technician skills, scheduling rules and offline mobile work. Scheduling policies and the asset model have to be right before rollout, because technicians notice errors on their first day.

Industry clouds come with their own data model and assumptions. Working with one outside manufacturing showed me how to judge where an industry model fits and where it forces the business into the wrong shape. The same question comes up with Manufacturing Cloud.

Many failed projects built the wrong thing well. Clear requirements, user stories with acceptance criteria and process mapping are the cheapest quality measure a project has.

Asynchronous processing, complex triggers and test strategy are where Salesforce code gets expensive. This level lets me judge partner designs at production data volumes and push back with specific alternatives.

The entry credential covers the shared vocabulary of the platform. It is the one I point new team members to when they join a project, so business and IT discuss the same things.

Design thinking and discovery workshops clarify the problem before anyone picks a solution. In enterprise projects with many stakeholders this saves weeks of building features nobody asked for.

This credential looks at sales from the salesperson's side. CRM adoption depends on whether the system helps sales reps sell, and a forecast is only reliable when they enter their data willingly.

Agentforce agents act on real data and real processes. Prompt design, the Einstein Trust Layer and clear limits on what an agent may do decide whether AI in Salesforce is safe enough for production.

AI results are only as good as the data behind them. This credential covers the basics of data quality, bias and responsible use that need to be settled before any model is switched on.

The handover between marketing and sales is where leads and consent records get lost. Knowing the marketing side helps define that handover so that leads arrive qualified and consent is documented.

Software architecture methods apply beyond any single vendor: quality goals, architecture decisions and documentation. They help me write down why a design was chosen, which matters when the next partner takes over.

Data platforms and IoT services often run on AWS next to Salesforce. Understanding the AWS side helps decide where data and logic should live and what each option costs.

API-led connectivity organises interfaces in reusable layers. With many systems around Salesforce this keeps the landscape changeable, because a new system connects to existing APIs.

Copado is one of the common DevOps tools for Salesforce. Knowing its pipeline model helps set up promotion paths and an audit trail that regulated companies need.

The second level covers merge conflicts, back promotions and parallel work streams. Those are the situations where teams lose trust in their pipeline and start deploying by hand again.

Flosum runs natively on Salesforce, which matters to companies whose security rules keep source code and metadata inside the platform. Knowing it lets me recommend a DevOps tool that fits those rules.

The expert level goes into compliance features, approvals and reporting on changes. In audited environments this shortens audits considerably, because every change can be traced to a ticket and an approval.

Gearset Master is the top level of the Gearset Academy. It covers moving teams from change sets to version control, automated testing and seeding test data before UAT. Many of the orgs I review are still in the middle of that move.

Scrum only helps when sprints aim at outcomes the business can check. With several partners delivering in parallel, a shared understanding of the framework keeps planning, reviews and dependencies in step.

Large companies often steer programmes with a steering committee while teams work in Scrum. PRINCE2 gives me the vocabulary of the governance side, so both levels can be connected without friction.

Manufacturing companies think in processes, defects and DMAIC. Speaking that language makes it easier to work with operations and quality teams on service and production processes.

The Green Belt adds measurement and root cause analysis. In service projects I use it for questions like why cases are reopened or why the first-time fix rate drops, before deciding what to change in Salesforce.

Some companies or business units are better served by HubSpot than by Salesforce. Knowing both lets me give an honest make-or-buy recommendation and plan migrations in either direction.

HubSpot Marketing is a common alternative to Marketing Cloud. Knowing it allows a licence recommendation based on the requirements alone.

HubSpot Service is enough for many small support teams. It reaches its limits with installed assets, field service and complex entitlements, which is the point where Salesforce Service Cloud earns its cost.