- Migrating from Zendesk to Salesforce Service Cloud is primarily a sequencing exercise. Organizations can reduce risk by proving the architecture in one region before extending it to others.
- Existing telephony does not have to be replaced. Voice can be integrated into Service Cloud through Service Voice while the carrier infrastructure remains in place.
- Cost savings are realized when the replaced platform is decommissioned. Operating legacy and replacement platforms in parallel expands the technology estate.
- A proven deployment in one region can resolve architectural questions, accelerate subsequent regional rollouts, and reduce program risk.
- The migration should address process design, channel unification, data, automation, and decommissioning rather than treating the program as a ticket data transfer.
- A first engagement should focus on one predefined use case, with an agreed deliverable, timeline, exclusions, and discovery process.
Why are global organizations replacing Zendesk with Salesforce Service Cloud?
Global organizations replace Zendesk with Salesforce Service Cloud when service is no longer a standalone function and needs to operate on the same data as commerce, order management, and CRM.For years, Zendesk has been an effective customer service platform for organizations that needed a dedicated environment for managing tickets and support interactions. The constraint arises when an organization needs service agents to view live order data, entitlement status, and customer history during an interaction, and AI agents to act on that data under the same governance as the rest of the business. At that point, the question stops being which ticketing tool is better and becomes whether the service should run on a separate system at all.Global organizations are reevaluating the role of Zendesk within their broader technology landscape. Three situations tend to trigger the decision:- One region has already deployed Service Cloud, and the inconsistency between regions has become expensive to maintain.
- AI agents are on the roadmap, and the data they need is outside the service platform.
- The estate has grown large enough that running parallel systems costs more than consolidating them.
What is involved in a Zendesk-to-Service Cloud migration?
A Zendesk-to-Service Cloud migration involves five workstreams, and only one of them is data migration.- Current service and CRM landscape assessment: Identify which systems hold which data, where processes actually run, and which dependencies must be resolved before Zendesk can be switched off.
- Process design: Surface undocumented regional and local practices before the build begins. Teams are less likely to adopt a new environment if it removes processes they rely on without providing an effective replacement.
- Service Cloud build: Configure case management, routing, intent classification, knowledge, and the escalation paths that transfer context from automated handling to human handling.
- Channel unification: Bring voice, chat, and email into one service environment, with customer context and real-time guidance available to agents during interactions.
- Decommissioning with a defined date: Establish what will be retired, when it will be retired, and which dependencies must be resolved before shutdown. This workstream is where the financial case for consolidation is ultimately realized.
How long does a Zendesk to Service Cloud migration take?
The duration of a Zendesk-to-Service Cloud migration depends on the number of regions in scope and the adjacent systems being changed or retired. The rollout sequence often has more influence on the program than the total geographic scale.Organizations that establish a reference deployment in one region before extending the architecture tend to move faster than those that attempt to resolve every regional design requirement before the first deployment. The second region can inherit tested design decisions rather than reopening them, reducing both delivery effort and architectural debate.- A first engagement, scoped to one predefined use case, is measured in weeks.
- A full multi-region program is measured in months.
- After the first region, the rollout pace is influenced largely by regional business readiness and adoption requirements.
Do you have to replace your telephony when you migrate to Service Cloud?
You do not have to replace your telephony when you migrate to Service Cloud. Salesforce Service Voice integrates with an existing telephony provider into the service console, so voice, chat, and email are handled in a single environment while the underlying carrier infrastructure remains in place.Keeping telephony in place during the migration is usually the lower-risk approach. Service consolidation and telephony replacement are separate projects with distinct stakeholders, failure modes, and procurement paths. Running them together is one of the more common reasons these programs slip.What should you ask a partner before committing to a migration?
Six questions separate a partner who has done this from one who has implemented Service Cloud generally:- Which region would you migrate first in our estate, and why that one?
- What would you decommission, and on what date?
- How will intent classification be validated before automation volume increases?
- What happens to our existing telephony during the migration?
- What does the escalation path look like when an automated conversation becomes a human one?
- What is explicitly excluded from the first engagement?
Replacing Zendesk in a 60-country service operation: an OSF case study
The organization: A global footwear brand operating in more than 60 countries across retail, ecommerce, and wholesale channels, with customer service running on divergent platforms by region.The problem: The organization had accumulated more than 180 systems across ecommerce, planning, order management, data, ERP, CRM, and marketing, carrying high maintenance costs, fragmented data, and operational complexity. EMEA had already consolidated onto Salesforce Service Cloud with Agentforce. The US was still running Zendesk.- Service teams had no real-time access to customer and order data during interactions.
- High volumes of repetitive transactional inquiries were triaged manually, slowing response and raising the cost to serve.
- Support capacity could not grow without proportional growth in headcount.
- Inconsistent experiences across channels and geographies put post-purchase engagement at risk, which mattered particularly in a US wholesale model still reliant on discounting.
- Einstein Case Classification identifies intent in incoming requests, resolving them autonomously or escalating them with the conversation history and recommended actions attached.
- Agentforce decision logic was embedded across high-volume service processes to improve efficiency and accuracy.
- Voice was integrated into Service Voice via Amazon Connect, alongside chat and email, into a single service environment with real-time guidance for agents.
- Operational and transactional data were connected so that responses carry live customer context rather than reconstructed history.
- The program delivered 50% case containment, with routine requests resolved without a person handling the interaction.
- They achieved 90% workflow unification, lowering the total cost of ownership and enabling volume growth without a proportional increase in support resources.
- Qualification of incoming interactions now occurs without wait time, accelerating both response and resolution.
- Service operations are consistent across more than 60 countries, with real-time omnichannel data available for reporting and decision-making.
- Customers have 24/7 coverage with context-aware interactions that support post-purchase engagement and spend.
What does a first engagement with OSF Digital look like?
A first engagement runs as five defined steps before any build commitment.- Discovery workshop: A 60- to 90-minute session with service operations, CX, IT, and support center stakeholders to align on goals and scope.
- Readiness assessment: Evaluation of data cleanliness, routing logic, knowledge gaps, and current KPI baselines, producing a specific list of what stands between the organization and its first use case.
- Pilot scope definition: Identification of the top intents for a 90-day pilot, with the business case and expected return built against them.
- Reference demonstration: Service Voice, Amazon Connect, and Einstein showed working against the organization's own use cases rather than a generic scenario.
- Strategic alignment: Timeline, team composition, and procurement path finalized before kickoff.
Where should a Zendesk-to-Service Cloud migration start?
A Zendesk-to-Service Cloud migration should start with the region, process, or use case that can prove the architecture while keeping operational risk within a manageable scope.Every service environment is different. While the migration might seem challenging, it’s essential to identify the right region, process, or use case to prove first and define what success should look like before the migration begins.A single discovery workshop can uncover hidden dependencies, readiness gaps, and the highest-value opportunities for consolidation and automation. From there, you can build a roadmap that balances implementation risk, operational efficiency, customer experience, and AI readiness.Talk to an OSF Digital service transformation expert and scope your Zendesk-to-Service Cloud migration.Frequently asked questions
Yes, and a phased regional migration is usually the lower-risk approach. Establishing a reference architecture in one region resolves most design questions in production, thereby shortening and de-risking every subsequent region.
Historical data handling is a scoping decision made during assessment, and the practical question is how much history the service operation actually needs to retain in the new system versus retaining it in an accessible archive. Migrating everything is rarely the cheapest or the fastest option.
Containment of approximately 50% is realistic for organizations with high volumes of transactional inquiries, such as order status, returns, and delivery questions. The achievable figure depends on intent classification accuracy rather than on the AI model, and containment produced without accurate classification tends to return as repeat contacts.
Case containment means a customer's request was fully resolved without a person, while case deflection means the customer was prevented from reaching a person, regardless of whether the issue was resolved. Containment reduces contact volume durably. Deflection tends to reappear as repeat contacts and lower satisfaction.
Not necessarily for transactional service automation, which primarily requires that customer and order data be reachable in real time during an interaction. Data Cloud becomes relevant when the requirement extends to unified profiles across marketing, commerce, and service for proactive and predictive engagement.
Yes, though these generally follow rather than lead an automation program. Autonomous handling of policy execution, data subject requests, invoicing, and refunds requires higher classification confidence and tighter governance than transactional containment, so most organizations sequence them after the transactional layer is stable in production.
Service consolidation creates the operational foundation for consistent customer experiences, lower service costs, greater visibility, and AI automation grounded in reliable data and processes. Digital agents are becoming part of the customer service operating model.
The Grounding Truth: Why AI Is Desperately Seeking Data
Do You Need to Replace Your Current Telephony?
Why Voice Belongs in Your CRM


