Before You Go Live on Salesforce, Do These Three Things
Most of the decisions that determine whether a Salesforce implementation succeeds get made before a single field is configured. That’s not an exaggeration. The quality of the process definition work, the state of the data going in, and the readiness of the team to actually use the system are set before the build starts. By the time go-live arrives, those things are largely fixed. If they were handled well, the implementation has a strong foundation. If they weren’t, the build inherited the problems and made them harder to fix.
This is why pre-launch preparation matters as much as the configuration work itself. Not as a formality before the real work begins, but as the work that makes everything after it possible.
There are three things every company should do before going live on Salesforce. Not because they appear on a checklist, but because skipping any one of them reliably creates problems that show up later and cost more to fix.
One: Define the Process Before You Build It
The most important thing that happens before a Salesforce implementation is the process definition. Not a high-level description of how the business operates, but a specific, honest account of how work should move through the organization: how a lead becomes a customer, who is responsible at each stage, what information needs to be captured and when, and where things currently break down.
That definition becomes the blueprint for the build. When the build is grounded in a clear process, the configuration choices make sense. The fields reflect what the team actually needs to capture. The pipeline stages reflect how deals actually move. The workflows automate things that genuinely need to be automated rather than things that seemed like a good idea in a meeting.
When the process isn’t defined before the build, what gets built is a reflection of how things currently work, including all the workarounds and inconsistencies that developed because there wasn’t a clear process to begin with. The system inherits the confusion and makes it harder to see. That’s not a configuration problem. It’s a sequence problem, and it’s entirely avoidable. Why You Can’t Buy Your Way to a Working CRM covers why process definition is the most important investment a company makes before the build starts.
The process definition conversation also needs leadership in the room. Not because leaders need to approve every field, but because the decisions about how the business should operate are leadership decisions. A good implementation partner facilitates that conversation. They bring the questions that surface the important decisions and the experience to recognize when a process is well-designed versus when it evolved by accident. How to Get Leadership Aligned on a Salesforce Implementation Before It Stalls covers why leadership engagement at this stage shapes the outcome more than most companies expect.
Two: Clean Your Data Before the Migration
Data cleanup is the pre-launch task that almost every company underestimates. Not because it’s technically complex, but because it takes longer than expected and requires decisions that only the people who know the data can make.
The principle is straightforward: don’t migrate bad data into a new system. Migrating bad data doesn’t clean it up. It gives it a new home, and it starts the new system’s life with a foundation that the team doesn’t trust. If the data is unreliable from day one, the team’s confidence in the system is unreliable from day one. And once that confidence is gone, it’s hard to rebuild.
The practical guidance here is to do the cleanup yourself. Your team knows your data better than any outside consultant does. You know which accounts are duplicates, which contacts are stale, which opportunity records reflect deals that closed two years ago and never got updated. Handing that work to someone who doesn’t know the business produces a result that’s technically cleaner and practically wrong in ways that are hard to predict.
The data doesn’t have to be perfect before go-live. No company we’ve worked with has achieved perfect data at migration time. But it needs to be clean enough that the system feels like a reliable source of truth when the team first uses it, because that first impression shapes how they engage with it going forward. How to Clean Your Data Before a Salesforce Implementation and Why It Matters More Than You Think walks through how to approach that work practically and what to prioritize when time is limited.
Three: Identify Your Internal Champion
The third thing that needs to happen before go-live is identifying the person inside the organization who will own the platform after the implementation partner leaves. This person is the internal champion, and their presence or absence shapes what happens to adoption in the months after launch more than almost any other single factor.
The internal champion doesn’t have to be a Salesforce administrator in a technical sense. They do have to understand how the system was built and why. They have to be credible with their peers, so that when they explain why a process works a certain way, people listen. They have to be responsive to the team’s questions and problems. And they have to be supported by leadership, because without that support, their ability to drive adoption and escalate issues is limited.
Identifying this person before go-live means the system has a caretaker from day one. Questions that come up in the first weeks, which they always do, get answered. Problems that surface get addressed rather than accumulating. The team has someone to go to who knows the system and is invested in making it work.
Leaving this decision until after launch means the system goes live without a clear owner. Responsibility diffuses. Problems get worked around rather than fixed. And the momentum that exists in the weeks immediately after go-live, which is real and valuable, dissipates before it produces the habits that make adoption stick. Why Salesforce Adoption Breaks Down in Growing Companies and What Leaders Miss covers how the internal champion connects to the broader adoption picture and what happens when that role is left unfilled.
Why These Three Things and Not Others
There are a lot of things that go into a successful Salesforce implementation. Training, integration planning, user acceptance testing, change management communication. Those things matter. But the three outlined here are the ones that, when skipped, most reliably produce problems that are expensive and slow to fix.
A poor process definition produces a system that doesn’t reflect how the business works. That’s a structural problem that requires rework to address. Bad data going in produces a system the team doesn’t trust. That’s a confidence problem that compounds over time. No internal champion produces an implementation that loses momentum after launch. That’s a governance problem that shows up gradually and is hard to reverse.
The other preparation work is important. These three are the ones that determine whether the foundation is solid before anything gets built on it. Why Salesforce Succeeds in Some Companies and Stalls in Others covers how these foundational decisions connect to the broader conditions that determine whether an implementation delivers over the long term.
The Work That Happens Before the Build Is the Build
One of the reframes that helps companies approach pre-launch preparation with the right level of seriousness is this: the work that happens before the configuration starts is part of the implementation, not a preliminary to it. The process definition, the data cleanup, and the identification of the internal champion are not administrative tasks that precede the real work. They are the real work. The configuration that follows is the translation of those decisions into a system.
When companies treat the pre-launch work as a formality and rush through it to get to the build, the build reflects that. When they treat it as the foundation it actually is, the build reflects that too. The difference in outcomes between those two approaches is significant, consistent, and entirely predictable.
Listen to the full podcast episode here.