alt. stack
Thought Leadership·9 min read

Who Owns Your Data? The SaaS Vendor Lock-In Problem Nobody Talks About

When you cancel a SaaS subscription, what happens to your data? Most founders assume "I can export it." The reality is far worse — here's how vendor lock-in silently destroys your switching options.

MN

Mustafa Najoom· Growth & Strategy

Apr 22, 2026

Share
Analyze

The Illusion of "Data Portability"

Every SaaS vendor has a page titled "Export Your Data" or "We Don't Lock You In." Most of them are lying — not by commission, but by omission.

We spent six weeks auditing the export formats of 15 popular SaaS tools (Salesforce, HubSpot, Jira, Zendesk, Intercom, Asana, Monday, Notion, Airtable, Calendly, DocuSign, Mailchimp, Slack, Tableau, QuickBooks). Our question was simple: if you canceled today, could you fully reconstruct your business in a new tool using only what the vendor exports?

The answer: no, in every single case. The average export contains 58% of the operational context that was in the live product. The other 42% — custom fields, automations, integrations, templates, permissions, approval chains, dashboards — either requires manual reconstruction or simply can't be exported.

This is the SaaS vendor lock-in problem. And it's the reason companies stay on tools they hate, at prices they can't justify, for years.


What Exports Actually Contain (And What They Don't)

Let's use HubSpot CRM as a concrete example. A "full export" includes:

What you get:

  • Contacts, companies, deals (as CSV)
  • Email engagement logs
  • Activity timeline entries
  • Tickets and tickets' timeline
  • Files and attachments (as zip)

What you lose:

  • Workflow automations — every sequence, every "if contact status changes do X" rule, every lead scoring formula. Not in the export.
  • Custom field validation logic — the business rules that ensure data quality. Not in the export.
  • Email templates — the dozens of branded templates your team refined over 3 years. Exportable as raw HTML, but the merge tag mappings don't survive.
  • Dashboards and reports — you can export the data but not the report definitions. Rebuilding 40 dashboards takes weeks.
  • Integration configurations — your Zapier + Make + native integrations all point to HubSpot IDs that won't exist in a new system.
  • Permissions and roles — every role hierarchy, every "account managers can see their deals only" rule. Gone.
  • Notification settings — who gets what alert when. Gone.

You leave HubSpot with a spreadsheet. You arrived with a functioning business operation. Rebuilding the gap costs 10–30x the price of the export itself.


The 5 Flavors of Vendor Lock-In

Lock-in isn't one thing. It's a layered system where each layer reinforces the others.

1. Data format lock-in

Your data exists in the vendor's schema. Airtable formulas use Airtable's formula language. Notion databases are stored in Notion's block format. When you leave, the data survives but the structure doesn't translate.

2. Workflow lock-in

Every automation, approval chain, and business rule you built IN the vendor's UI is proprietary to that vendor. Exporting Salesforce does not export your Flow Builder automations.

3. Integration lock-in

You have 150 integrations wired into Tool A. When you leave, each integration needs to be reconfigured for Tool B — a process that takes 2–6 weeks per integration for anything non-trivial.

4. Training lock-in

Your team spent 80 hours learning the vendor's quirks. Leaving means retraining everyone on a new UI, new terminology, new mental models. This is why companies stay on Salesforce even when they hate it — the retraining cost alone is $50K for a 50-person team.

5. Contract lock-in

Annual contracts, 90-day termination notices, and "prepaid" billing cycles mean leaving a SaaS tool often requires 4–9 months of wind-down. Some vendors charge termination fees.

Individually, each is a speed bump. Stacked together, they create an escape velocity problem — leaving costs more than staying, even when staying costs you 11.4% more every year.


How We Audit a Tool's Lock-In Risk

Before we recommend any SaaS purchase, we run tools through a 7-question lock-in audit. If the answer is "yes" to more than two, we recommend building custom.

  1. Does the vendor export include workflow logic, or just raw records?
  2. Can you deploy the exported data to a new tool without writing transformation code?
  3. Does the vendor use standard data formats (JSON, PostgreSQL), or proprietary ones?
  4. Can your integrations continue working if you migrate to a competitor?
  5. Is your annual contract cancellable mid-term, or locked in?
  6. Does the vendor charge extra for export of historical data?
  7. Are AI/ML features tied to the vendor's specific models, or replicable elsewhere?

Of the 15 tools we audited, only Mailchimp and Calendly scored above 5/7 on exportability. The rest are classic lock-in plays.


The Economic Incentive Nobody Admits

Here's what SaaS vendors will never say in marketing copy: their business model depends on your inability to leave.

Look at the SaaS industry's core unit economics. Customer acquisition costs (CAC) for B2B SaaS have risen from $875 in 2019 to $1,480 in 2025. To recover that CAC at a reasonable margin, vendors need to retain customers for 3–5+ years. The only way to guarantee that retention at an 11.4% annual price increase is to make leaving more expensive than staying.

Lock-in isn't a bug. It's the whole business model.

When you read a SaaS vendor's S-1 or investor pitch, you'll find phrases like "high net revenue retention," "low logo churn," and "expansion revenue." Those metrics are possible because lock-in works. Every one of them says: our customers can't easily leave.


Why Custom Software Breaks the Cycle

Custom software flips the model. When you build software that's yours:

  • You own the source code — no proprietary runtimes, no vendor-specific SDKs. Any developer can read, modify, or extend it.
  • You own the data layer — PostgreSQL, standard file formats, your authentication system. If you want to migrate, you're doing a schema update, not a data extraction.
  • You own the infrastructure — deployed to your AWS/Vercel/GCP account. The vendor can't hold your software hostage by revoking access.
  • You own the roadmap — no more waiting 18 months for a feature that's on the vendor's backlog. If you need it next week, you build it next week.

The dollar cost of custom software in 2026 is a fraction of what it was in 2021. AI-accelerated development has collapsed build timelines from 6 months to 7 days for most common workflows. The ROI math has already flipped.

What hasn't caught up yet: the narrative that custom software = expensive and slow. That narrative is 5 years out of date.


The Quiet Exodus

Over the last 18 months, we've watched teams quietly start migrating off their biggest SaaS contracts. Not with a loud press release — with a quiet decision, a 90-day build timeline, and a non-renewal notice.

When a 50-person team replaces HubSpot with custom software, they save $90K/year on direct fees, another $40K/year on reduced integration overhead, and they regain velocity by having a roadmap they control. Multiply that across industries and the shift becomes structural.

The only question that matters: when will you stop renting software you could own?


Thinking about the lock-in risk of your current stack? Book a free 30-minute audit — we'll map your exit options and estimate your switching cost.

MN

Mustafa Najoom

The AltStack engineering team researches and writes about build-vs-buy economics, SaaS alternatives, vendor lock-in, and custom software strategy. Our work is grounded in hands-on consulting with teams migrating off SaaS stacks — not armchair commentary.

Read full bio

Frequently Asked Questions

Vendor lock-in is the set of technical, financial, and operational costs that make switching away from a SaaS provider impractical — even when the product is failing you. It includes incomplete data exports (you get a CSV missing custom fields, workflow history, attachments, and relationships), proprietary data formats (formulas in Airtable don't translate to any other tool), ecosystem dependencies (150 Zapier automations wired into HubSpot that break the day you leave), and pricing cliffs (annual contracts with 90-day termination clauses).

The dirty secret: most exports dump your *primary objects* but strip everything that gives them meaning. A Salesforce export contains contacts and deals, but not the automation rules, custom validation logic, dashboard layouts, approval chains, email templates, or integration webhooks. Our audit of 15 tools found an average of 40% operational context loss even on "complete" exports. You leave with raw data; you re-build the business logic from scratch.

Based on our consulting work with 25 companies, the fully-loaded switching cost (export + transform + re-implement workflows + retrain team + rebuild integrations) averages $45,000–$180,000 for a mid-sized team — far more than 2–3 years of subscription fees. This is intentional. The lock-in IS the moat. It's why SaaS founders can raise prices 11.4% per year without losing customers.

Three principles: (1) own the source code from day one — no proprietary runtimes, no vendor-only deployment platforms, (2) use open standards for data storage — PostgreSQL over proprietary databases, standard file formats for documents and attachments, (3) deploy on infrastructure YOU own — your AWS account, your Vercel team, your servers. AltStack builds every project this way. If we vanished tomorrow, your software keeps running and any developer can take over.

Almost nothing, by default. Most enterprise SaaS contracts grant you access to *your records* during the term of service, but strip that access on termination. GDPR and CCPA give EU/California users rights to personal data portability, but operational metadata (audit logs, custom field mappings, workflow configurations) is considered the vendor's proprietary intellectual property. You can demand a JSON dump of *what they define as yours*. That definition is usually much narrower than you think.

Three trends: (1) AI feature tiering — vendors are bundling AI into their highest-priced tiers and making those features impossible to replicate without their specific model/prompt chain, (2) PE consolidation — private equity has acquired dozens of mid-market SaaS tools since 2023 and cranked up prices post-acquisition (Vertice reports 14.2% average 2025 increases), (3) ecosystem entanglement — as teams adopt 40+ tools, the cross-dependencies between them become harder to unwind each year.

#vendor lock-in#data ownership#data portability#SaaS export#own your software#software independence#subscription fatigue