1,200 tools: Klarna reportedly decommissioned Salesforce during its SaaS breakup
Klarna reportedly cut Salesforce during a 1,200 tool SaaS breakup. For a 10 person Pro Suite team, the rent is $36,000 over 36 months.
Tim Neunzig
1,200 tools were reportedly on Klarna's chopping block. Salesforce was one of them, according to the cited reporting. Workday was named too.
That is the part that matters to you. Not the AI noise. Not the founder theater. A large company looked at rented software and decided it wanted fewer landlords.
The exact Salesforce shutdown date is unknown from the cited sources. The internal replacement name is unknown. The architecture is unknown. The bill math is not.
What happened
Klarna's CEO-led SaaS breakup was reported as a plan to decommission 1,200 external software tools. Salesforce and Workday were included in the reported list. The replacement was described as internal tools built through vibe coding.
Business Insider covered the market reaction around AI replacement and the future of software. Sequoia published an interview with Sebastian Siemiatkowski. SalesforceBen wrote about why vibe coded CRMs are not the end of Salesforce. A GeoMusings source in the set is dated 2026/06/22. The exact date when Klarna decommissioned Salesforce is not established by the cited sources.
That is important. Do not turn a report into a fake case study. The sources do not give you Klarna's Salesforce contract value. They do not give you seat count. They do not give you the internal CRM name. They do not give you its database, language, hosting, data model, or migration plan.
They do give you the shape of the event. Klarna reportedly attacked external software sprawl. The reported figure was "1,200 external software tools." Salesforce was named. Workday was named. The replacement direction was internal software.
Salesforce's own pricing gives you the smaller version of the same problem. Sales Cloud pricing lists Starter Suite at "$25 per user per month" and Pro Suite at "$100 per user per month." Enterprise is listed at "$165 per user per month." Unlimited is listed at "$330 per user per month." Einstein 1 Sales is listed at "$500 per user per month."
The add-on page adds more rent paths. Marketing Cloud Account Engagement starts at "$1,250 per month" for up to 10,000 contacts, according to the cited pricing summary. Agentforce for Sales is listed at "$125 per user per month." Revenue Intelligence is listed at "$250 per user per month." Sales Dialer starts at "$5 per user per month." Sales Engagement is listed at "$50 per user per month."
That is the machine. Start with CRM. Add seats. Add tiers. Add modules. Add a new name for a feature bundle. Keep paying.
For the model here, the rent is simple. Pro Suite is $100 per user per month. A 10 person team pays for 36 months. That is $36,000.
Klarna's case is not proof that every company should rip out Salesforce tomorrow. It is proof that a serious buyer can ask a brutal question: why are we renting a giant platform when the daily job is smaller?
What it actually did for you
Salesforce Sales Cloud is a CRM. Strip the cathedral paint off it and the job is plain.
You store leads, contacts, accounts, and opportunities. You assign owners. You update stages. You log calls and emails. You chase the next step. You open cases when customers have problems. You run reports so managers can see pipeline, activity, conversion, and risk.
That is the daily surface for many teams. Customer record. Deal record. Activity history. Case queue. Reminder. Report. Export.
The useful parts are not exotic:
- Lead, contact, account, and opportunity management. This is the record system. Who you sell to, what is open, what stage it is in, and who owns it.
- Email integration. Sales activity has to sit next to the customer record. Otherwise it dies in private inboxes.
- Case management. Customer issues need an owner, status, history, and follow-up path.
- Workflow and automation. Owner assignment, reminders, stage changes, approvals, and case status changes remove grunt work.
- Forecasting and pipeline inspection. Managers need open pipeline by owner, stage, expected close period, and amount.
- Reporting. You need basic views for pipeline, cases, activity, and conversion. The cited sources do not establish daily-use frequency for every report.
Now the padding.
The source set does not provide usage-frequency data good enough to prove which Salesforce features are ignored. So the honest answer is narrower. You pay for a platform surface far larger than the replacement scope. You may not use all of it. The cited sources do not prove what Klarna used or ignored.
Salesforce's own product and pricing pages show the breadth. Advanced AI, Agentforce for Sales, Einstein 1 Sales, Revenue Intelligence, Sales Engagement, Sales Dialer, partner tools, marketing automation, and many add-ons sit around the core CRM. They may matter to a large sales organization. They may be dead weight to you.
That distinction is the kill. Do not rebuild Salesforce. Rebuild the work.
The math
The model is a 10 person team on Salesforce Pro Suite for 36 months. The listed Pro Suite price is $100 per user per month. The owned build range is $2,500 to $8,000.
| Option | Price per user per month | Team size | 36 months | Total |
|---|---|---|---|---|
| Salesforce Pro Suite | $100 | 10 | 36 | $36,000 |
| Owned CRM build | $0 license rent | 10 | 36 | $2,500 to $8,000 build, hosting unknown |
You keep the code, the database, the exports, the workflows, and $28,000 against the $36,000 rent if the $8,000 build replaces the Pro Suite subscription. Payback is 8 months.
After that, the per-seat tax is gone for this CRM scope. Hosting, maintenance, and future changes still exist. The cited sources do not price those.
What you would build instead
Build a narrow internal CRM. Not a Salesforce clone. Not a marketplace. Not a second platform religion.
- CRM database. Store accounts, contacts, leads, opportunities, cases, activities, owners, stages, notes, and audit timestamps in PostgreSQL.
- Sales workflow service. Run owner assignment, stage reminders, due-date notifications, case status changes, and simple approval steps.
- Pipeline and forecast views. Show open opportunities by stage, owner, expected close period, and forecast amount.
- Reporting screens. Cover pipeline, cases, activity, and conversion. Export to Excel-compatible files.
- Admin and roles. Let admins manage users, roles, fields, stages, and simple permissions.
- Import and export tool. Import Salesforce-style records for leads, contacts, accounts, opportunities, cases, activities, owners, stages, and notes. Keep the raw export as an archive.
- Email and calendar links. Connect Outlook or Gmail for email context. Use Google Calendar if the team actually uses it.
- Notifications. Send Slack alerts for ownership changes, due dates, stuck deals, and case movement.
Use PostgreSQL for the relational data. Use a Python FastAPI service for the application and business logic. Use server-rendered HTML with HTMX so the UI stays boring. Put Nginx in front for reverse proxy and TLS termination. Run it on one self-hosted Linux server controlled by the client, unless load or compliance forces a split.
For open-source building blocks: PostgreSQL uses the PostgreSQL License. FastAPI uses MIT. HTMX uses BSD-2-Clause. Nginx uses BSD-2-Clause. SuiteCRM is an AGPL-3.0 CRM reference or alternative, but the project URL was not established by the cited source set. EspoCRM is an AGPL-3.0 CRM reference or alternative, with the same URL gap in the cited source set.
Data migration is mechanical, not glamorous. Export Salesforce leads, contacts, accounts, opportunities, cases, activities, owners, stages, and notes. Map only fields the team actually uses. Import into PostgreSQL. Keep the raw export. The cited sources do not establish Klarna's actual Salesforce data model, so do not pretend yours matches theirs.
You lose real things. You lose Salesforce's broad packaged CRM surface outside the replacement scope. You lose vendor support and managed cloud operations. You lose packaged Mobile CRM unless you build a mobile app. You lose advanced Salesforce AI and add-ons such as Einstein 1 Sales, Agentforce for Sales, Revenue Intelligence, Sales Engagement, and Sales Dialer. You lose the larger integration ecosystem and ready-made connections to tools you may use later. You lose complex admin configuration, marketplace extensions, and enterprise governance features outside the owned build.
Good. If you need those, rent them. If you do not, stop funding them.
Who should do this
Do this if you are a 10 person sales or account team on Pro Suite and your real day is records, pipeline, email, cases, workflows, and reports. Your 36 month rent is $36,000. A $2,500 to $8,000 owned CRM can pay back in 8 months.
Do this if you have stable sales stages and a small operations team that knows the process. You do not need a giant admin console when the workflow is known. You need clean tables, clear ownership, imports, exports, and boring screens.
Do this if you hate per-seat pricing more than you love packaged breadth. Growth should not automatically mean a higher CRM bill for the same narrow workflow.
Keep renting if Salesforce is your operating system. If you depend on advanced AI, packaged mobile, deep partner tools, heavy marketplace extensions, complex enterprise governance, many third-party integrations, and vendor support, do not build a toy and call it strategy.
Keep renting if your process changes every week and no one owns requirements. Custom software punishes chaos. SaaS hides it on a monthly invoice.
The SaaSKiller Take: Salesforce Pro for 10 people is $36,000 over 36 months. Klarna reportedly put Salesforce inside a 1,200 tool SaaS breakup. Your lesson is not to copy Klarna. Your lesson is to stop renting a platform when your team only needs records, pipeline, email, cases, workflows, and reports.
Sources
- https://www.salesforce.com/sales/cloud/guide/
- https://www.salesforce.com/products/
- https://www.businessinsider.com/salesforce-microsoft-stocks-down-ai-replacement-future-of-software-2026-4
- https://www.capterra.com/p/61368/Salesforce/
- https://www.linkedin.com/posts/arnaud-contival-aa947_is-saas-dead-no-but-its-traditional-per-user-activity-7500981276309397504-nhwX
- https://www.linkedin.com/posts/jacquesbughin_saas-ai-reinvention-activity-7492283080926343168-9srQ
- https://www.salesforce.com/sales/pricing/add-ons/?bc=OTH
- https://sequoiacap.com/podcast/training-data-sebastian-siemiatkowski
- https://www.salesforceben.com/why-vibe-coded-crms-are-not-the-end-of-salesforce/
- https://www.larksuite.com/en_us/blog/salesforce-pricing
- https://frontdeskreview.com/software/crm-software/salesforce/
- https://blog.geomusings.com/2026/06/22/vibe-coding-ai-disruption-and-the-restructuring-of-the-saas-market/
- https://www.xpay.sh/saas-pricing/salesforce-sales-cloud/
- https://vendorbenchmark.com/vendors/salesforce-sales-cloud-pricing
- https://topdynamicspartners.com/learn/erp-comparison/d365-vs-salesforce
AI-generated