A mid-sized company I worked with decided to migrate their CRM from Salesforce to HubSpot. They were spending $85K/year on Salesforce with complex customizations they’d built over 7 years.

Budget for migration: $40K (mostly consulting fees and data migration tooling).

Actual total cost: $340K and 14 months of disruption.

That’s the reality of SaaS vendor lock-in. The switching costs everyone casually dismisses turn out to be 5-10X higher than anyone estimates.

Migration Cost Categories

When companies evaluate SaaS migration, they typically budget for:

• Data export and transformation
• New system setup and configuration
• User training
• Parallel running period

What they underestimate or miss entirely:

Business process re-engineering: Your workflows were built around previous system’s capabilities and limitations. New system does things differently. You need to redesign processes, not just replicate old ones.

Integration rebuilding: Every API integration, webhook, automation connecting to the old SaaS needs rebuilding for the new one. Even if both systems have “REST APIs,” the actual implementation differs substantially.

Custom feature replacement: That custom module you built in old system? Doesn’t exist in new system. You either live without it, rebuild it (if customization is possible), or accept workflow degradation.

Data cleanup: Migration exposes years of accumulated data quality issues. Duplicate records, inconsistent formatting, orphaned references. You can’t migrate dirty data without causing problems, so you spend weeks cleaning.

Change management: Employees learned old system, built muscle memory, developed workarounds. They resist new system even if it’s objectively better. Training isn’t enough—you need ongoing support and champions.

Opportunity cost: While engineering and ops teams focus on migration, other projects get delayed or cancelled. The hidden cost is everything you didn’t ship during migration period.

Real Migration Example: CRM Switch

The Salesforce to HubSpot migration mentioned earlier broke down as:

Budgeted costs ($40K):

• HubSpot onboarding and setup: $8K
• Data migration consultant: $18K
• Tooling and data transformation: $6K
• Training delivery: $8K

Actual costs ($340K):

• Extended consulting (9 months vs. planned 3 months): $120K
• Internal engineering time rebuilding integrations: $85K (2,200 hours at blended rate)
• Business process redesign workshops and implementation: $45K
• Extended parallel running (both systems for 5 months): $35K (extra licenses)
• Lost productivity during transition: estimated $40K
• Custom development for features HubSpot lacked: $15K

Timeline:

• Planned: 4 months
• Actual: 14 months from kickoff to decommissioning Salesforce

Outcomes:

• Annual SaaS cost reduced from $85K to $48K (good)
• Several Salesforce customizations couldn’t be replicated (workflow degradation)
• Sales team productivity dropped 15-20% for 6 months during transition
• Three integrations still not working properly 14 months later

Was it worth it? Hard to say. SaaS savings of $37K/year mean 9-year payback on migration costs. Most of the executive team who sponsored the migration has since left the company.

Integration Lock-In (The Hidden One)

Most discussions of vendor lock-in focus on data export difficulty. That’s solvable—almost every SaaS provides data export, and worst case you can scrape it.

The real lock-in is integrations. You’ve built workflows connecting:

• CRM to email marketing
• Helpdesk to billing system
• Project management to time tracking
• Analytics to data warehouse
• SSO to identity provider
• Slack to everything else

Each integration represents 10-80 hours of development plus ongoing maintenance. Switching any single SaaS means rebuilding all integrations touching that system.

I’ve seen companies running 30-50 integrations per major SaaS platform. That’s 300-1,500 hours of engineering work to switch platforms, assuming straightforward rebuilds with no complications.

This is why “we’ll just switch providers if pricing increases” is empty threat. The integration rebuild cost exceeds several years of price increases.

The Customization Trap

SaaS vendors encourage customization. Better user experience, workflows tailored to your business, competitive differentiation through better tools.

Then you try to migrate, and discover new platform doesn’t support 40% of your customizations. You either:

1. Lose functionality (users unhappy, productivity drops)
2. Pay for custom development in new platform (expensive, if even possible)
3. Change business processes to fit new platform’s constraints (organizational resistance)
4. Stay on old platform despite wanting to switch (vendor lock-in achieved)

The Salesforce ecosystem is particularly notorious for this. Companies build elaborate customizations using Apex code, custom objects, workflow rules, and process builder flows. Migrating this to another platform is basically rebuilding your entire CRM logic from scratch.

Contract Length Lock-In

Enterprise SaaS typically requires 1-3 year contracts. If you sign 3-year deal and want to switch after year one, you’re paying for two systems simultaneously or eating early termination fees (typically 50-100% of remaining contract value).

This contractual lock-in compounds technical lock-in. Even if migration only costs 6 months and $200K, you might be paying another $150K in overlapping licenses or termination fees.

Vendors know this and structure pricing to encourage multi-year commits with aggressive discounts. The 30% discount for 3-year contract looks attractive until you realize it makes switching economically painful.

What Actually Reduces Lock-In

Open APIs and standards: SaaS platforms with comprehensive, well-documented APIs and standard data formats (REST, JSON, OAuth) are easier to integrate with and migrate from. Proprietary protocols and formats increase lock-in.

Export functionality: Regular automated exports of all data in usable formats. Not “download CSV through UI” but automated API export of complete database including relationships and metadata.

Minimal customization: Using SaaS platforms mostly as-provided instead of heavily customizing. This reduces migration complexity even if it means less perfect fit to current workflows.

Integration layer: Building integrations through middleware (Zapier, Workato, custom integration platform) instead of point-to-point. When you switch SaaS, you rebuild integration layer connections instead of all downstream integrations.

Multi-tenancy skepticism: Running critical workflows across multiple SaaS platforms instead of consolidating everything into single vendor ecosystem. Reduces impact if you need to switch any single platform.

Contract negotiation: Shorter contract terms (even at higher effective pricing) preserve optionality. Annual contracts are better than 3-year even if they cost 15% more.

When Lock-In Is Acceptable

Not all vendor lock-in is bad. Sometimes it’s the right trade-off:

Platform provides genuine differentiated value that competitors don’t. You’re locked in, but alternative isn’t better.

Switching costs exceed any plausible benefit. If migration costs $400K and annual savings from switching is $30K, lock-in doesn’t matter. You’re not switching.

Vendor relationship is strong and pricing is stable. Lock-in is primarily problem if vendor exploits it through aggressive pricing or service degradation. Good vendors don’t.

Customization delivers real ROI. If custom workflows increase productivity 20%, lock-in cost is acceptable price for those gains.

The problem is when companies don’t recognize they’re accepting lock-in or underestimate switching costs, then feel trapped when circumstances change.

How to Evaluate Before Committing

Before deeply integrating with any SaaS platform:

1. Do switching cost estimation. Model what migration would actually cost including all categories (data, integrations, custom features, parallel running, opportunity cost). Triple that estimate for realism.

2. Review contract terms for exits. Understand termination fees, notice periods, data retention after cancellation.

3. Test data export. Actually export data and verify completeness, formats, and usability. Don’t trust vendor claims about export capability.

4. Evaluate integration architecture. Can you build integrations through reusable middleware vs. point-to-point custom development?

5. Limit initial customization. Use platform mostly as-provided for first 6-12 months. Customize only after validating core platform meets needs and you’re committed long-term.

6. Include switching costs in TCO. Total cost of ownership includes eventual migration costs, not just annual licensing and operational costs.

Bottom Line

SaaS vendor lock-in is real and expensive. Switching costs are 5-10X higher than most companies estimate before doing detailed migration planning.

This doesn’t mean avoid SaaS—modern business runs on SaaS, and alternatives are often worse. But recognize lock-in exists, understand switching costs, and make informed decisions about:

• Which platforms to deeply integrate with
• How much customization to implement
• What contract lengths to accept
• How to structure integrations for portability

The companies that manage SaaS vendor relationships best are those who acknowledge lock-in exists, estimate switching costs realistically, and make intentional decisions about when lock-in is acceptable trade-off.

Don’t pretend you can easily switch later. You probably can’t. So choose carefully up front.

IT leadership focused on practical technology decisions and sustainable vendor relationships.