Sauvetage Salesforce ou réimplémentation : comment réparer une org Salesforce cassée
Anes, Founder of Digital Stratify
November 5, 2026
9 min read

Sauvetage Salesforce ou réimplémentation : comment réparer une org Salesforce cassée

Les commerciaux ne saisissent plus rien, personne ne croit au forecast et le dernier partenaire a laissé un chantier. Comment choisir entre sauvetage et redémarrage, avec un plan en 3 semaines.

Your reps stopped logging months ago, the forecast gets rebuilt in a spreadsheet every Monday, and the partner who set up Salesforce no longer answers emails. You are paying for a system your team routes around.

I rebuild CRMs for a living, and a large share of my work is Salesforce orgs that someone else built. The question I get asked first is almost always the same: do we fix this, or do we start over? This article is the answer I give, including when starting over is the right call.

The warning signs of a broken Salesforce org

A broken org rarely announces itself. It degrades.

Reps are not logging

Activity sits in inboxes and Slack, not on records. When you ask why, the answer is some version of "it takes too long" or "nobody looks at it anyway." Both are true, and both are symptoms, not causes.

The data is stale

Close dates in the past on open opportunities. Accounts with no activity in a year still marked as active. Duplicate contacts with three spellings of the same name.

Nobody trusts the forecast

If your head of sales keeps a private spreadsheet, your Salesforce forecast is already dead. The pipeline report exists, but decisions are made elsewhere.

Too many custom fields

I regularly open Opportunity objects with 150 or more custom fields, of which maybe 20 are populated on more than half of records. Each one was a reasonable request at the time. Together they make every page feel like a tax form.

A previous partner left a mess

Undocumented flows that fire on every edit. Validation rules nobody dares to delete. Apex classes with no test coverage. A managed package that was never fully configured. You cannot change anything because you do not know what will break.

If you recognize three or more of these, you have a broken org. The next question is what kind of broken.

Rescue vs reimplementation: what each actually means

A rescue keeps your existing org, your data, your logins and your integrations. It strips out what is not used, fixes the core objects and processes, rebuilds the automation that matters, and gets the team using it again. The org that comes out is recognizably the same org, just clean.

A reimplementation stands up a new org (or wipes the current one back to a blank configuration), redesigns the data model from scratch, migrates only the data you decide to keep, and relaunches. It is a new project with the discipline and timeline of a new project.

Most people assume a reimplementation is the "proper" fix and a rescue is a patch. In my experience it is closer to the opposite. A reimplementation is often an expensive way to repeat the same mistakes with a fresh coat of paint, because the real problem was never the configuration.

The decision checklist

Each "yes" on the left pushes you toward a rescue, each on the right toward a reimplementation.

QuestionRescue ifReimplement if
Is the core data model (Accounts, Contacts, Opportunities) roughly right?Yes, the objects make sense but are clutteredNo, the model is built around a business you no longer run
Is your data worth keeping?Yes, history matters and most records are real customersNo, the data is mostly junk or belongs to a dead product line
Do integrations work?Yes, billing, marketing and support sync correctlyNo, every integration is broken or maps to the wrong fields
Is there Apex or a managed package you cannot remove?Little or none, mostly Flow and configHeavy custom code with no tests and no documentation
Did the business model change?No, same motion, same customersYes, new pricing model, new segment or a merger
Are you on the right platform?Yes, Salesforce fits your size and complexityUnsure, you might be better on HubSpot or Attio
How many users?5 to 50 reps who need to be live fastHundreds of users across divisions with different processes

If you landed mostly on the left, rescue. If you landed mostly on the right, reimplement, and before you do, read my Salesforce vs HubSpot vs Dynamics comparison, because a reimplementation is also your one chance to change platforms cheaply.

If you are genuinely split, a short diagnostic is cheaper than guessing. I offer a free 30-minute CRM diagnostic call where we walk through this checklist against your actual org. Book it here.

What a 3-week Salesforce rescue looks like

This is how I run the Revenue System Rescue for companies with 5 to 50 reps. The scope is deliberately narrow: get the org clean, live and trusted. Everything else waits.

Week 1: Diagnose and decide

Days 1 to 2 are interviews. Reps, managers, the person who runs the forecast, whoever owns billing or marketing integrations. I want to know what people actually do and what they avoid.

Days 3 to 5 are the audit. Field usage analysis on every core object, a map of every active Flow and validation rule, installed packages and whether they are used, integration health, and a data quality scorecard. By Friday you get a written diagnosis and a cut list: what stays, what goes, what gets rebuilt. You sign off before anything changes.

Week 2: Rebuild

All work happens in a sandbox. Unused fields are removed or hidden, page layouts are rebuilt around the 15 to 25 fields reps actually need, stage definitions are rewritten with exit criteria, and automation is consolidated into a small number of documented Flows. Duplicates are merged. Stale opportunities are closed or archived with a reason.

Reports and dashboards are rebuilt from the leadership questions down, not from the existing report folder up.

Week 3: Validate, train, go live

Monday and Tuesday are user acceptance testing with two or three reps and the sales leader. Wednesday is deployment to production, timed around your sales calendar. Thursday is training, which is short because the org is now simple. Friday is the first pipeline review on the new setup, with me in the room. Then two weeks of hypercare: I watch adoption, fix what surfaces, and hand over documentation.

The Revenue System Rescue is fixed price, paid in milestones, with one guarantee: clean, live and trusted by your team in 3 weeks, or I keep working free until it is.

Typical costs and timelines in the market

I do not publish my own package pricing here, but you should know the general ranges so you can judge a quote.

  • Salesforce health check or audit from a consultancy: typically one to three weeks and a low five-figure fee, sometimes credited against follow-on work.
  • Focused rescue of a small or mid-size org (5 to 50 users): typically three to ten weeks. Boutique firms and independents usually land in the low to mid five figures. Large system integrators quote the same scope at several times that, with a longer timeline and a more junior delivery team.
  • Full reimplementation: typically three to nine months. Expect a cost comparable to your original implementation, often higher once data migration and integration rework are included. My Salesforce implementation timeline breaks down where the months go.

If a quote for a rescue looks like a quote for a reimplementation, ask what they plan to rebuild and why.

Mistakes that make a broken org worse

  • Adding fields to fix a process problem. Reps are not logging because logging does not help them, not because a field is missing. Every new required field lowers adoption further.
  • Buying an add-on before cleaning the base. Forecasting tools, AI features and enrichment products amplify whatever data you feed them.
  • Letting the original partner run the rescue. Sometimes they are the right choice. Often they built the mess and will defend it. Get an independent audit first.
  • Rebuilding every report. Start from the five questions leadership asks every week. Delete the rest.
  • Skipping the sandbox. Deleting fields and Flows directly in production is how a one-week fix becomes a three-month recovery.
  • Treating go-live as the finish. Adoption is won or lost in the first two weeks after launch.

One more: if you have fewer than five users and a simple pipeline, you may not need a rescue at all. Spend a weekend removing fields, archive dead opportunities, rebuild one dashboard, and see if adoption recovers.

Frequently asked questions

How do I know if my Salesforce org needs a rescue?

Three or more of these: reps not logging activity, stale opportunity data, forecasts built outside Salesforce, far more custom fields than anyone uses, or undocumented automation inherited from a previous partner. One is friction. Three means the system is being routed around.

How long does a Salesforce rescue take?

For a company with 5 to 50 reps and a roughly correct data model, three to ten weeks depending on the firm. My Revenue System Rescue runs about three weeks: one week to diagnose, one to rebuild in a sandbox, one to validate, train and go live, followed by two weeks of hypercare.

Is it cheaper to reimplement Salesforce or fix the existing org?

Fixing is almost always cheaper when the core data model is roughly right and the data is worth keeping. Reimplementation makes sense when the business model has changed, the data is mostly junk, or heavy undocumented code makes the current org unsafe to change. Reimplementation typically costs as much as the original project, sometimes more.

What is a Salesforce health check and do I need one first?

A health check is a short structured audit of field usage, automation, data quality, integrations and adoption, ending in a written diagnosis. It is the right first step if you are unsure whether to rescue or reimplement. If the warning signs are obvious, it can be folded into week one of a rescue instead of bought separately.

Should I switch from Salesforce to HubSpot or Attio instead of fixing it?

Only if the platform is the problem, not the configuration. A company with 5 to 50 reps, a simple sales motion and no complex integrations may be better served by a lighter CRM. A company with multiple product lines, quoting, service and billing integrations usually should stay on Salesforce and fix it.

Not sure which path you are on?

I offer a free 30-minute CRM diagnostic call. Share your screen, show me the org, and I will tell you whether it is a rescue or a restart. Book a call.

Si ça ressemble à votre CRM, regardons-le ensemble.

Trente minutes, sans slides, sans pitch commercial. Vous repartez avec un diagnostic dans tous les cas.