Bespoke CRM, SaaS or a layer on top of your ERP: how Swiss SMEs should choose

Almost every service SME we meet already has "a CRM". Sometimes it is a SaaS subscription paid for the last four years and opened by two people out of ten. Sometimes it is an ERP module nobody ever really configured. Often, it is a shared Excel file, a team inbox and the sales manager's memory.

So the question is not "do we need a CRM?", but which type of CRM fits the way your business actually works. There are three routes. None is better in absolute terms. Each one answers a specific situation.

Route 1: off-the-shelf SaaS CRM

HubSpot, Pipedrive, Salesforce, Zoho: solid tools, used by thousands of companies, ready in a few days.

When it is the right choice:

• Your sales process is standard (prospect, meeting, offer, signature).
• You do not have an ERP yet, or your ERP does not handle customer relationships.
• You want to get started quickly and are happy to adapt to the tool.

Where it falls short in practice:

• The cost per user climbs as the team grows. The useful features (automations, advanced reporting, lead assignment) often sit in the higher tiers.
• Double data entry with the ERP. Sales updates the CRM, the back office updates the ERP, and the two versions drift apart within weeks.
• Your business process gets bent to fit the tool. An installation company does not have the same cycle as a SaaS agency: technical visit, priced offer, signature, scheduling, on-site work, invoicing, follow-up. Generic CRMs handle the start well and the rest poorly.
• Adoption. A tool that takes five clicks to log a call will not get used. And an empty CRM is worthless.

Route 2: the custom-built CRM

Software built around your process, your stages, your vocabulary and your assignment rules.

When it is the right choice:

• Your process has specifics that give you an edge (geographical areas, types of service, languages, scheduling constraints).
• You receive enquiries from many channels (website, advertising, phone, chatbot, AI search engines) and you lose time sorting them.
• You want automations that understand your trade: follow-ups at the right moment, call prioritisation, meeting preparation.

What bespoke makes possible and SaaS does badly:

• Fine-grained automatic assignment: by area, by product, by language, with absence handling and reassignment when a file goes quiet.
• A "Top 3 to call today" list based on urgency, value and likelihood of signing, rather than a list of 80 leads sorted by date.
• A meeting preparation sheet generated automatically: history of exchanges, comparable projects nearby, points to watch.
• A price estimator calibrated on your own quotes. In one project we delivered, the estimator was calibrated on 241 of the company's real quotes, not on a theoretical price grid.
• One-click actions from an email (signed links), so people in the field can update the CRM without opening it.

What to watch out for:

• You need a provider who understands the business, not just the code.
• You need to plan for maintenance and evolution, as with any living tool.
• If you already have an ERP that is the reference, rebuilding everything in a separate CRM is a mistake. Hence the third route.

Route 3: a CRM layer on top of your existing ERP

This is the most underrated route. Many Swiss SMEs already have an ERP (Odoo, bexio, or industry-specific software) that manages contacts, offers, invoices and projects. The problem is not a lack of data: it is that the ERP is painful for day-to-day sales work.

A CRM layer is a lightweight application designed for sales staff and management, connected to the ERP through two-way synchronisation. The ERP remains the source of truth for accounting and projects. The layer adds what is missing: prioritisation, follow-ups, dashboards, AI automations.

When it is the right choice:

• Your ERP is in place and accepted by the back office.
• Sales staff work around it (Excel, notes, WhatsApp).
• You want reliable dashboards without re-keying data.

What to demand from a serious CRM layer:

• Regular synchronisation in both directions (activity, signatures, amounts, stages, owner, invoices, lost deals).
• Safeguards: an amount validated in the ERP is never overwritten automatically; bulk changes go through a simulation (dry run) and then human approval.
• Duplicate prevention when contacts are created on the ERP side.
• A "truth scan" that regularly compares CRM and ERP and flags discrepancies, instead of letting them grow.

The decision framework

Ask yourself these five questions, in order:

1. Do you have an ERP that is already the reference for offers and invoices? If so, start by assessing a CRM layer. Building a second reference system costs more than it seems.
2. Is your sales cycle standard? If so, and without an ERP, a well-configured SaaS is often enough.
3. Where do your enquiries come from, and how long do you take to handle them? Measure the time between a lead arriving and the first action. For a technical SME in French-speaking Switzerland that we work with, this time dropped from 216 hours to 95 hours after automatic sorting and assignment were put in place.
4. How many hours a month go into repetitive tasks? Follow-ups, record updates, reports, meeting preparation. This is where bespoke software and AI pay off most.
5. Who will maintain the tool? A SaaS maintains itself but does not adapt. A custom-built tool adapts but requires a long-term partner.

The real cost to compare

Do not compare a monthly subscription with a development quote. Compare the total cost over three years:

• licences and the higher tiers needed for automations;
• hours lost to double entry and manual reconciliation;
• leads handled too late or never called back;
• advertising budget spent on channels whose real return nobody measures.

That last point often comes as a surprise. When the CRM finally links each lead's source to the signature, the gaps become visible. In the case mentioned above, leads from the website signed 2.3× more often than those from paid social media, and around CHF 20,000 a year of unprofitable advertising budget was identified and cut. No SaaS in its default configuration makes that connection without real integration work.

And where does AI fit in?

AI does not replace the architecture decision; it makes it more important. AI agents that read emails, prepare follow-ups or write reports need clean data and a clear reference system. Plugging AI into an empty CRM, or one that conflicts with the ERP, simply automates the mess.

We explain this approach (AI prepares, a human approves) in our article on AI agents for SMEs.

In short

• SaaS: fast and proven, ideal for a standard process without an ERP.
• Bespoke: when your process is your advantage and generic tools hold you back.
• CRM layer on the ERP: when your data already exists but nobody enjoys using it.

The right choice is the one your team will use every day. To see what a CRM built around your business looks like in practice, visit our bespoke CRM page or our full range of business software and AI automations. And if you would like to talk it through, get in touch.