Sales Cloud and Service Cloud in 90 Days: The Setup Plan for Growing Businesses
A practical 90-day plan to set up Salesforce Sales Cloud and Service Cloud: discovery, data migration, integrations, training and support after go-live.
Sales Cloud and Service Cloud 90-day setup plan
Most Salesforce projects don't fail because of the software. They stall because nobody agreed how the business actually works, the data was a mess on day one, or the launch happened and then nobody owned what came next.
If you're a growing business planning to start Salesforce in the new year, this is the plan I'd want you to have in your hands first. It covers Sales Cloud (for your sellers) and Service Cloud (for the people who look after customers after the sale), in four phases across roughly 90 days.
One honest caveat before we start: 90 days is a realistic target for a focused scope, not a promise. A single team and a handful of processes can move faster; a company-wide rollout with heavy custom development takes longer. Your own timeline comes out of the discovery phase, not out of a blog post.
Phase 1 (Days 1 to 15): Decide how the business works
I know, you want to see screens. Resist. The most valuable two weeks of the project are spent on paper.
We start by walking through how a lead becomes a customer and how a customer gets help. Where does a new enquiry come from? Who responds? What are the stages a deal passes through, and what has to be true for a deal to move to the next one? When a customer raises a problem, who owns it, how fast must it be answered, and when does it escalate?
Write these down in plain language. Then decide the numbers you want Salesforce to show you in six months: lead response time, win rate, pipeline by stage, first-response time on cases, cases resolved within target. If you can't name the report, you can't design the system.
This phase is what we'd call strategic consulting: a process assessment, a solution design and an architecture review. Skipping it is the single most common reason I see a Salesforce setup get reworked a year later.
Phase 2 (Days 16 to 45): Build the core
Now you build, in the order that gets value into people's hands soonest.
Sales Cloud first. Lead capture from your website and other sources, with assignment rules so every lead has an owner. Accounts, contacts and opportunities with your stages and required fields. Duplicate rules, so the same company doesn't appear four times. Quotes and products if you sell from a price list. A forecast view for the sales head, and a small set of dashboards that answer the questions you wrote down in phase 1.
Service Cloud next. Cases created from email and your web form, with queues and assignment rules, and a status flow people can follow. Service targets (how fast you promise to respond), with escalation when you're about to miss one. A knowledge base of short, accurate articles so agents answer consistently, and so customers can answer themselves. If you want a customer-facing portal, that's where Experience Cloud comes in, giving customers a login to raise and track requests.
Two rules I hold to here. Keep it simple: every extra field you add is one more thing someone has to fill in, and people stop filling in forms they don't believe in. And configure before you code: most of what a growing business needs is built with Salesforce's standard tools (flows, validation rules, page layouts), with custom development kept for what genuinely can't be done any other way.
Phase 3 (Days 30 to 70): Move your data and connect your tools
This phase overlaps with the build, and it's where timelines slip if you leave it late.
Clean before you load. Moving messy data into a new system just gives you messy data in a new place. Before anything is imported, agree what you're keeping (do you need seven years of closed-lost deals?), merge duplicates, standardise the fields that matter (country, industry, stage names) and decide who owns each record.
Pick the right loading tool. Salesforce's own guidance is that the Data Import Wizard suits smaller, simpler imports: fewer than 50,000 records, into supported objects with fewer than 50 fields, and it can prevent duplicates by matching on account name and site, contact email or lead email. For larger volumes, complex field mappings, scheduled loads or objects the wizard can't handle, Data Loader is the tool, and it supports files of up to 150 million records. Whichever you use, load a small sample first, check the results against the source, then load the rest.
Connect what you already use. Most growing businesses have an accounting tool or ERP, email, and perhaps a marketing platform. The question for each is simple: what information needs to move, in which direction, and how often? If your finance and operations run in an ERP such as Odoo, the connection between the two is where "one version of the truth" is won or lost. We cover that choice in Salesforce vs. Odoo: which CRM-ERP stack fits your business. On our side, integration is part of how we implement and develop: ERP, marketing and third-party connections using APIs and middleware, with our Starter package including up to three integrations.
Phase 4 (Days 60 to 90): Train, launch, and don't disappear
Training is not a one-hour demo. People need to practise on their own real records, in a test environment, with their own scenarios. Train sellers on the three or four things they'll do every day, not on every button. Train service agents on the case flow and the knowledge base. Name a champion in each team: someone colleagues will ask first.
Launch in a way you can recover from. A common pattern that works: pilot with one team for a couple of weeks, fix what you find, then roll out to everyone. Keep the old process available for a short, defined overlap, then retire it clearly so nobody keeps one foot in each system.
Plan for the weeks after go-live. The first weeks always surface things nobody predicted: a missing field, a report that doesn't match the old spreadsheet, a user who's stuck. Have a named person and a response time for these. Beyond the first weeks, Salesforce keeps changing under your feet with regular releases, so someone needs to test upgrades, keep an eye on org health, answer user questions and tune performance. That's the work of managed support: a help desk, incident handling, release upgrades and testing, ongoing training, and periodic health checks.
Choosing the size of the project
We offer three packages on our Salesforce implementation services page, and the differences are about scope more than anything else. Starter suits small teams setting up Sales or Service Cloud with up to three integrations and 30 days of post-launch support. Growth is for multi-cloud implementations, with Einstein AI, unlimited integrations and 90 days of support, led by a dedicated project manager. Enterprise covers a company-wide rollout with custom development, complex integrations, change management and 12 months of managed support, with a dedicated solution architect. Your discovery phase will tell you which one you're really buying.
If you want to see how this has gone for other teams, our Salesforce case studies are the best place to look. And for a closer look at what a partner should include, and what it costs, read what Salesforce implementation services include and how to choose a partner.
Five mistakes I'd help you avoid
- Starting with features, not processes. If you can't describe how a deal moves, no tool will do it for you.
- Importing everything. Bring only the data you'll actually use, and clean it first.
- Customising too early. Start with standard tools and add custom code only when you must.
- Treating training as an afterthought. Adoption is the project. A perfect system nobody uses is a failed project.
- Having no plan for after go-live. Decide who owns support, upgrades and improvements before you launch.
What comes after day 90
Once Sales Cloud and Service Cloud are running and your data is trustworthy, the next steps get interesting: deeper integrations with your ERP, customer portals, richer dashboards and, when you have clean data and a clear use case, AI. Chatbots over your knowledge base, lead scoring and forecasting all depend on the foundations you've just built. If that's where you want to go, our AI and machine learning services start with exactly the same discipline: discovery, data assessment and a proof of concept before anything goes live.
And if you're still deciding whether to bother, the true cost of not using a CRM or ERP is worth five minutes.
Conclusion
A good Salesforce setup isn't a technology project with a business side effect. It's a business project that happens to use Salesforce. Decide how you work, build the core in the order that helps people soonest, move clean data, connect the tools you already depend on, train properly, and plan for life after launch. Do that and 90 days is a sensible target for a focused scope.
If you're planning a Salesforce start in the new year, talk to our team. We'll tell you what a realistic scope and timeline look like for your business, which package fits, and what to leave out of phase one.
People Also Ask (FAQs)
How long does a Salesforce Sales Cloud and Service Cloud implementation take?
For a focused scope, about 90 days is a realistic target: roughly two weeks of discovery, a build phase, data migration and integrations overlapping the build, then training and launch. Larger or more customised rollouts take longer. Your own timeline should come from a discovery phase.
What should we do before starting a Salesforce implementation?
Document how leads become customers and how customers get support, define the reports you want in six months, and clean your data. Decide what to keep, merge duplicates and standardise key fields. Clear process and clean data prevent most rework.
Should we use the Data Import Wizard or Data Loader?
Salesforce says the Data Import Wizard suits imports of fewer than 50,000 records into supported objects with fewer than 50 fields, and can prevent duplicates by matching on account name and site, contact email or lead email. Data Loader suits larger volumes, complex mappings, scheduled loads and unsupported objects, and handles files up to 150 million records.
Do we need both Sales Cloud and Service Cloud?
Not necessarily. Sales Cloud supports your sellers and Service Cloud supports customer service teams. Many businesses start with one and add the other once the first is working. If the same customers move between sales and support, running both on one platform gives you a single customer record.
Can Salesforce integrate with our ERP or accounting system?
Yes, through APIs and middleware. Decide for each tool what data should move, in which direction and how often. Symake's Starter package includes up to three integrations, and the Growth and Enterprise packages include unlimited integrations.
What support do we need after go-live?
Plan for a named owner and response time during the first weeks, and ongoing managed support after that: help desk and incident management, testing of Salesforce release upgrades, user training and periodic org health checks.
Which implementation package should we choose?
Symake offers Starter (small teams, up to three integrations, 30 days of post-launch support), Growth (multi-cloud, Einstein AI, unlimited integrations, 90 days of support) and Enterprise (company-wide rollout, custom development, change management, 12 months of managed support). A discovery phase will show which fits your scope.