Migrating a CRM to Salesforce can be complicated by decisions that initially seem minor: replicating current processes without reviewing them, transferring information without verifying its quality, or leaving user training until the end. Without a methodology to detect and correct these problems in time, the project can accumulate delays and cost overruns, compromise data security, or result in a system that is difficult to maintain and barely used by the teams.
At ALTIA, we see migrating to Salesforce as an opportunity to review how the organization works , leverage the platform's capabilities , and build a solution that meets the real needs of the business. To achieve this, we combine functional and technological expertise with a methodology that incorporates Salesforce best practices: understanding how things work, demonstrating standard capabilities before making adjustments, and validating the solution with those who will be using it.
Decisions that can compromise a migration
- Define the scope before understanding the work. A list of requests only tells part of the story: it might leave out an approval done by email, a report that someone corrects by hand, or a connection with another application. Understanding the entire process before defining the scope prevents these needs from surfacing too late.
- Asking for a design from people who are unfamiliar with Salesforce is a mistake. Without any reference point for what the product offers, it's understandable that users would request the same screens and rules they currently use. The problem arises when each habit becomes a custom development that will require payment, testing, and maintenance.
- Addressing data connectivity and protection too late. Moving data without deciding which application maintains it, who can access it, or what it will be used for creates problems that a siloed Salesforce configuration won't solve. Data architecture, integration, quality, and protection must be part of the design from the outset.
- Mistaking speed for a lack of control. Simple improvements can quickly deliver value or create problems if they accumulate without considering their impact on the rest of the system. Furthermore, with artificial intelligence, a proof of concept must demonstrate its usefulness, cost, and fit within real-world workflows before scaling.
- Checking the data without testing the work. Just because a sales opportunity appears in Salesforce doesn't mean the team can handle it well. Validation should replicate entire processes, not just verify that the data has been uploaded correctly.
- Reaching the launch date without any testing. The schedule sets a goal, but it doesn't resolve a faulty connection or explain how to recover operations. A controlled launch requires a proven plan, clearly defined responsibilities, and a recovery scenario.
- Calling a project finished too soon is a mistake. The first few weeks reveal doubts and difficulties that don't always appear during testing. Subsequent support allows you to stabilize the solution, gather lessons learned, and prioritize its evolution.
How we approach a migration to Salesforce at ALTIA
We organized the project into seven phases , from the initial diagnosis to the stabilization and evolution of the solution. First, we understand the current situation; then we demonstrate Salesforce and agree on how we will work together. From there, we build and test incremental releases until we are ready for implementation and ongoing support. Security, data protection, and information quality are paramount throughout the entire process.
Our Salesforce team coordinates this process with our cross-functional teams. Data analytics helps organize and leverage information; cybersecurity helps prevent threats and respond to incidents; regulatory compliance helps define appropriate data use; and artificial intelligence can be incorporated when needed. In this way, we integrate the necessary capabilities based on the needs of each project, maintaining a unified view of the solution.
1. Understand how things work and set priorities
Talk to the people who do the work and those who need the results. We review real-world examples involving business, operations, and technology: how a request is handled, who approves a transaction, or where a report originates, in addition to tracking work that falls outside the CRM. Understanding the reason behind each step is crucial. A manual check might protect a critical decision or simply compensate for a shortcoming in the existing system. Distinguishing between these two situations allows us to decide what to maintain, what to simplify, and what to transform.
With that vision, we identified those responsible for the data, the quality issues, and the information requiring greater protection. We also agreed on baseline indicators, decision-makers, and key users for validation.
It is in this phase that we identify quick wins : improvements with the ability to add value with a limited effort and that can be incorporated into the roadmap if their viability is proven.
2. Demonstrate Salesforce and agree on the future way of working
Users can make better decisions when they've seen the system. That's why we organize hands-on sessions where we demonstrate Salesforce processes before they complete the setup.
The goal is not to replicate the previous CRM, but to explore how Salesforce can solve business needs in a simpler, more connected, and scalable way.
What do we work on in each workshop?
- Explore the processes offered by Salesforce. We demonstrate how to manage business-relevant processes and what platform capabilities can be leveraged.
- Compare to current work. Users identify what Salesforce brings to the table, which tasks can be simplified, and what changes will be needed in their workflow. Exceptions that have business value are also tested.
- Classify the differences. We agree on what Salesforce covers directly, what needs adjustments, what requires changing a process, and what requires a custom solution that needs to be evaluated separately.
- Closing decisions. Each session leaves an updated description of how work will be done, justified decisions, pending issues with the responsible party, and clear conditions for accepting proposed improvements.
This work allows for decisions to be made with a shared vision between business and technology, and avoids turning every particularity of the previous system into a custom development.
3. Design the solution and protect the information
Before building, you need to figure out how Salesforce will fit into your organization. These decisions shape the solution's architecture, though their effects will be felt in everyday matters: what data is accurate, how long a query takes, or who can view a document.
- Each application must have a clear purpose. We agree on which processes Salesforce handles and where the reference information is stored. If the billing data belongs to another system, it will need to be connected seamlessly, without creating two competing versions.
- Connections must be fault-tolerant. Some information needs to be updated immediately, while other information can wait. In both cases, there must be a system in place for detecting an interruption and restoring the exchange without losing or duplicating data.
- Data needs order and meaning. We leveraged the Salesforce structure and defined how customers, contacts, documents, and historical data are migrated. In addition to fixing errors, we agreed on what each metric measures to prevent two reports from giving different answers to the same question.
- The design must be tailored to the intended use. We test what happens when the number of users and simultaneous operations increases. This allows us to correct overlapping or slow-down tasks before they affect the entire team.
- Security is tested in practice. Each person should only have access to what they need. We define the uses and retention periods for information and review actual access. Problems detected in the security audit are corrected and retested.
4. Build with quality and deliver value from the first improvements
The initial releases should help the team recognize the value of the change . We prioritize, in collaboration with the business, those improvements that can generate impact from the earliest stages and that leverage standard Salesforce capabilities.
Before commissioning a custom development, we check if the function already exists and what adjustments it needs. If it needs to be extended, we clearly explain its added value and the maintenance costs. Changes are documented and tested before being implemented in the daily operating system. These tests include typical and exceptional scenarios, permissions, and simultaneous operations.
When artificial intelligence is involved, the decision to expand its use depends on the results. First, we test the case with the available data; then, users test it in real-world situations. We evaluate usefulness, cost, errors, and the need for oversight before deciding how to proceed.
Each delivery starts from a clear need and acceptance criteria and is incorporated when the result is validated and can be maintained.
5. Validate complete processes and prepare people
A useful test replicates the entire workflow: the user locates a client, reviews an opportunity, updates it, and checks how it appears in the sales forecast. This allows us to simultaneously validate the data, the process, and the user experience . We also test exceptional situations such as interrupted connections, slow responses, repeated operations, or unauthorized access.
The training is based on the already validated solution, so that teams learn about processes and tools that they will actually use.
6. Prepare a controlled start-up
The switchover day should follow a pre-tested plan . We agree on when work will stop on the old system, how the latest information will be migrated, and who will verify that everything is available in Salesforce. We also define how to handle an incident and what conditions would allow us to stop or revert the process.
The client authorizes the deployment based on the results of the checks. Users know when they should start working in Salesforce and where to get help. The goal is for the move to production to be a controlled change, not the end of the preparation process.
7. Stabilize the service and organize its evolution
During the first few weeks, we address any questions and closely monitor the system's operation. Some issues require immediate attention, while others indicate areas for improvement that should be planned.
Real-world usage provides insights that testing doesn't always reveal. Beyond checking speed and connectivity, we observe whether teams keep opportunities up-to-date and trust the reports. This monitoring allows us to identify new opportunities for improvement and prioritize them based on their business impact.
Permissions and uses of information change over time. Therefore, reviewing access, controls, and policies is also part of the solution's evolution.
From a migration to a platform that evolves with the business
A well-executed implementation is evident in daily work: teams understand the processes, can trust the information, and know who to turn to when a question arises.
To achieve this, at ALTIA we combine business knowledge, technological capabilities, and support throughout the entire project. We compare needs before finalizing the scope, demonstrate Salesforce before deciding on its customizations, and test the entire solution before launch. Design and data protection are integral to these decisions.
We then continued working with the client until the operation was stabilized and the necessary knowledge was transferred. The result is a solution ready to meet current needs and evolve with the organization.