Go from Salesforce to Bigin
The job: Small-team sales pipeline (Contacts, Companies, Deals, Tasks, Simple pipeline report). Pick where you are today and where you want to land; the move below follows your pick.
Every way to do this job · pick one
Switch app
Open source
Build it with an agent
Agent skill
The money
$1,000/mo today , then $70/mo
Salesforce Pro, 10 seats
Modeled from public list prices, with your time at $50 an hour. Show the math
Does it do the job? 5 of 5 covered
- Covered: Contacts
- Covered: Companies
- Covered: Deals
- Covered: Tasks
- Covered: Simple pipeline report
The work
- Setup
- About 10 h + $100
- Upkeep
- 0.5 h a month
- Watch out
- No Salesforce CPQ, territory logic, required advanced permissions or complex integration estate.
Make the move in four steps
Export from Salesforce
A backup of your org's data, by the data types you choose. Format: Zip archive of CSV files
Export stepsSet up Bigin
10 users; contacts, one deal pipeline, tasks and simple reports
DetailsSwitch links and integrations
Point links, embeds and integrations at Bigin, then send a test through every one.
Cancel Salesforce
Turn off renewal before your renewal date, once the new setup works. Access runs until: The end of the current subscription term.
How to cancel Salesforce
The whole recipe: 16 ways out, best first
What you use
- Contacts
- Companies
- Deals
- Tasks
- Simple pipeline report
Not covered
- CPQ
- Territories
- Complex forecast rules
- Large integration estates
- Enterprise security parity
Check before switching
- All must-have features covered
- Export relationships verified
- No required CPQ or territory model
- Compare downgrade and managed CRM first
Agents: get_recipe("crm-pipeline") on the Stackcut MCP server.
16 ways out, best first
How we got this
| You pay now: Salesforce Pro | $100/user/mo × 10 users | $1,000/mo |
|---|---|---|
| New cost: Bigin Express | $7/user/mo × 10 users | $70/mo |
| Your upkeep time | 0.5 h/mo × $50/h | $25/mo |
| Difference each month | $1,000 − $70 − $25 | +$905/mo |
| Over 12 months | +$905 × 12 | +$10,860 |
| Setup, once | 10 h × $50/h + $100 one-time | −$600 |
| Year one | +$10,860 − $600 | +$10,260 |
- Not counting your time (16 h in year one): saves $11,060 in year one.
Year one = 12 × (what you pay now − the new cost − upkeep) − setup, with your time at $50/hour. Prices are public list prices; hours are Stackcut’s planning estimates. Price it for your team
How we got this
| You pay now: Salesforce Pro | $100/user/mo × 10 users | $1,000/mo |
|---|---|---|
| New cost: Replace with AI: bespoke pipeline | hosting and API costs (modeled) | $40/mo |
| Your upkeep time | 2 h/mo × $50/h | $100/mo |
| Difference each month | $1,000 − $40 − $100 | +$860/mo |
| Over 12 months | +$860 × 12 | +$10,320 |
| Setup, once | 50 h × $50/h + $500 one-time | −$3,000 |
| Year one | +$10,320 − $3,000 | +$7,320 |
- Not counting your time (74 h in year one): saves $11,020 in year one.
Year one = 12 × (what you pay now − the new cost − upkeep) − setup, with your time at $50/hour. Prices are public list prices; hours are Stackcut’s planning estimates. Price it for your team
How we got this
| You pay now: Salesforce Starter Suite | $25/user/mo × 1 user | $25/mo |
|---|---|---|
| New cost: Zoho CRM (free tier) | free plan | $0/mo |
| Difference each month | $25 − $0 | +$25/mo |
| Over 12 months | +$25 × 12 | +$300 |
| Setup, once | 1 h × $50/h | −$50 |
| Year one | +$300 − $50 | +$250 |
- Not counting your time (1 h in year one): saves $300 in year one.
Year one = 12 × (what you pay now − the new cost − upkeep) − setup, with your time at $50/hour. Prices are public list prices; hours are Stackcut’s planning estimates. Price it for your team
How we got this
| You pay now: Pipedrive Lite | $24/user/mo × 1 user | $24/mo |
|---|---|---|
| New cost: Zoho CRM (free tier) | free plan | $0/mo |
| Difference each month | $24 − $0 | +$24/mo |
| Over 12 months | +$24 × 12 | +$288 |
| Setup, once | 1 h × $50/h | −$50 |
| Year one | +$288 − $50 | +$238 |
- Not counting your time (1 h in year one): saves $288 in year one.
Year one = 12 × (what you pay now − the new cost − upkeep) − setup, with your time at $50/hour. Prices are public list prices; hours are Stackcut’s planning estimates. Price it for your team
How we got this
| You pay now: Pipedrive Lite | $24/user/mo × 1 user | $24/mo |
|---|---|---|
| New cost: Attio (free tier) | free plan | $0/mo |
| Difference each month | $24 − $0 | +$24/mo |
| Over 12 months | +$24 × 12 | +$288 |
| Setup, once | 1 h × $50/h | −$50 |
| Year one | +$288 − $50 | +$238 |
- Not counting your time (1 h in year one): saves $288 in year one.
Year one = 12 × (what you pay now − the new cost − upkeep) − setup, with your time at $50/hour. Prices are public list prices; hours are Stackcut’s planning estimates. Price it for your team
How we got this
| You pay now: Salesforce Starter Suite, 2 users | $25/user/mo × 2 users | $50/mo |
|---|---|---|
| New cost: Agent skill: keep a sales pipeline in CSV files you own | runs in the agent you already pay for | $0/mo |
| Your upkeep time | 0.5 h/mo × $50/h | $25/mo |
| Difference each month | $50 − $0 − $25 | +$25/mo |
| Over 12 months | +$25 × 12 | +$300 |
| Setup, once | 3 h × $50/h | −$150 |
| Year one | +$300 − $150 | +$150 |
- Not counting your time (9 h in year one): saves $600 in year one.
Year one = 12 × (what you pay now − the new cost − upkeep) − setup, with your time at $50/hour. Prices are public list prices; hours are Stackcut’s planning estimates. Price it for your team
- Tested by Stackcut: 14 of 14 synthetic cases passed a file checker and a strict rubric after 3 rounds (11, 13, 14); an independent round (a fresh agent ran every case, a separate agent graded it) passed 14 of 14. our test
- No new subscription if you already pay for an agent that runs skills: a skill is a folder of files the agent loads. agentskills.io
- Your pipeline is five CSV files you own. Every change is shown as a diff first, and the script refuses changes that would leave the files invalid. Deletions need a separate yes. our test
- No email sync, reminders, pipeline board or mobile app: you paste messages in and ask for the due list. our test
- One editor at a time: no concurrent editing and no user permissions. our test
- HubSpot's free CRM covers up to 2 users and 1,000 contacts at $0, with a UI. hubspot.com
Step-by-step: set up, move your data, check, roll back · about 3 h (estimate)
Keep your pipeline in five CSV files you own. Paste in emails, call notes and LinkedIn messages, approve the diff your agent proposes, and a script writes it.
Adapting the reference files (what you sell, owners, currencies, stages) and moving your records into the five CSVs. In our test, writing the skill, references, script, 14 cases and checker took about 11 minutes of agent time, and each test round (runner plus grader) 2.6 to 4.6 minutes.
Before you start
- Export before you cancel. HubSpot: CRM → the object → Export (CSV, XLSX or XLS; the emailed link expires after 30 days). knowledge.hubspot.com
- Pipedrive: Tools and apps → Export data (Excel or CSV). Notes and activities linked to deals are separate exports, file attachments aren't included, and export files stay for 28 days. support.pipedrive.com
- Salesforce: Setup → Data Export gives a ZIP of CSV files, deleted 48 hours after the email. help.salesforce.com
Set up
- Use the Agent Skills layout: SKILL.md plus references/ (pipeline rules, matching, deletion requests, drafts, business facts) and scripts/crm.py (Python 3, no network). agentskills.io
- Run crm.py init on a folder kept in git or a synced drive with version history. Then write what you sell into business.md: in our test, drafts went wrong until the skill could tell a customer's choice from your list apart from a request only you can decide. our test
- Write 10 to 15 realistic test cases, and the change-set you expect for each, before the first run. The included checker compares them.
Move your data
- Map the export to the five CSVs: one row per company, person, deal and open task, with past emails and calls in activities.csv. Run crm.py validate until it passes, and keep the export as your archive.
Check it worked
- Have one agent run the test cases and a separate agent grade them against the rubric and tests/check.py. Fix the references, not the outputs. our test
- For two weeks, keep the old CRM read-only and compare its pipeline with the weekly report.
Roll back
- When a HubSpot cancellation takes effect, the account moves to HubSpot's free tools, so there is still somewhere to import back into. knowledge.hubspot.com
- Pipedrive imports XLS, XLSX and CSV (people, organizations, deals, activities, notes), up to 50,000 rows per file. support.pipedrive.com
Watch out for
- HubSpot: a cancellation takes effect on your renewal date, and mid-contract downgrades aren't permitted. knowledge.hubspot.com
- Salesforce's help page offers no self-serve cancel for a paid subscription: you contact its billing department. help.salesforce.com
- Deleting a contact removes them from the five files, but not from git history, synced-drive versions, exports or your inbox. Clean those up yourself for a GDPR request. our test
- Fixing one rule can break another case. In our test, the rule that stopped drafts agreeing to a customer's terms failed a case where the customer only picked a listed price tier. Re-run every case after each change. our test
Reviewed 2026-09-29.
How we got this
| You pay now: Salesforce Starter Suite | $25/user/mo × 1 user | $25/mo |
|---|---|---|
| New cost: Twenty | $12/user/mo × 1 user | $12/mo |
| Difference each month | $25 − $12 | +$13/mo |
| Over 12 months | +$13 × 12 | +$156 |
| Setup, once | 1 h × $50/h | −$50 |
| Year one | +$156 − $50 | +$106 |
- Not counting your time (1 h in year one): saves $156 in year one.
Year one = 12 × (what you pay now − the new cost − upkeep) − setup, with your time at $50/hour. Prices are public list prices; hours are Stackcut’s planning estimates. Price it for your team
How we got this
| You pay now: Pipedrive Lite | $24/user/mo × 1 user | $24/mo |
|---|---|---|
| New cost: Twenty | $12/user/mo × 1 user | $12/mo |
| Difference each month | $24 − $12 | +$12/mo |
| Over 12 months | +$12 × 12 | +$144 |
| Setup, once | 1 h × $50/h | −$50 |
| Year one | +$144 − $50 | +$94 |
- Not counting your time (1 h in year one): saves $144 in year one.
Year one = 12 × (what you pay now − the new cost − upkeep) − setup, with your time at $50/hour. Prices are public list prices; hours are Stackcut’s planning estimates. Price it for your team
How we got this
| You pay now: HubSpot Sales Hub Starter | $10/user/mo × 1 user | $10/mo |
|---|---|---|
| New cost: Zoho CRM (free tier) | free plan | $0/mo |
| Difference each month | $10 − $0 | +$10/mo |
| Over 12 months | +$10 × 12 | +$120 |
| Setup, once | 1 h × $50/h | −$50 |
| Year one | +$120 − $50 | +$70 |
- Not counting your time (1 h in year one): saves $120 in year one.
Year one = 12 × (what you pay now − the new cost − upkeep) − setup, with your time at $50/hour. Prices are public list prices; hours are Stackcut’s planning estimates. Price it for your team
How we got this
| You pay now: HubSpot Sales Hub Starter | $10/user/mo × 1 user | $10/mo |
|---|---|---|
| New cost: Attio (free tier) | free plan | $0/mo |
| Difference each month | $10 − $0 | +$10/mo |
| Over 12 months | +$10 × 12 | +$120 |
| Setup, once | 1 h × $50/h | −$50 |
| Year one | +$120 − $50 | +$70 |
- Not counting your time (1 h in year one): saves $120 in year one.
Year one = 12 × (what you pay now − the new cost − upkeep) − setup, with your time at $50/hour. Prices are public list prices; hours are Stackcut’s planning estimates. Price it for your team
How we got this
| You pay now: HubSpot Sales Hub Starter | $10/user/mo × 1 user | $10/mo |
|---|---|---|
| New cost: Small Business plugin (Anthropic knowledge-work-plugins) | runs in the agent you already pay for | $0/mo |
| Your upkeep time | 0.5 h/mo × $50/h | $25/mo |
| Difference each month | $10 − $0 − $25 | −$15/mo |
| Over 12 months | −$15 × 12 | −$180 |
| Setup, once | 1 h × $50/h | −$50 |
| Year one | −$180 − $50 | −$230 |
- Not counting your time (7 h in year one): saves $120 in year one.
Year one = 12 × (what you pay now − the new cost − upkeep) − setup, with your time at $50/hour. Prices are public list prices; hours are Stackcut’s planning estimates. Price it for your team
How we got this
| You pay now: Salesforce Starter Suite | $25/user/mo × 1 user | $25/mo |
|---|---|---|
| New cost: Twenty | a small server (modeled) | $5/mo |
| Your upkeep time | 2 h/mo × $50/h | $100/mo |
| Difference each month | $25 − $5 − $100 | −$80/mo |
| Over 12 months | −$80 × 12 | −$960 |
| Setup, once | 4 h × $50/h | −$200 |
| Year one | −$960 − $200 | −$1,160 |
- Not counting your time (28 h in year one): saves $240 in year one.
Year one = 12 × (what you pay now − the new cost − upkeep) − setup, with your time at $50/hour. Prices are public list prices; hours are Stackcut’s planning estimates. Price it for your team
How we got this
| You pay now: Salesforce Starter Suite | $25/user/mo × 1 user | $25/mo |
|---|---|---|
| New cost: SuiteCRM | a small server (modeled) | $5/mo |
| Your upkeep time | 2 h/mo × $50/h | $100/mo |
| Difference each month | $25 − $5 − $100 | −$80/mo |
| Over 12 months | −$80 × 12 | −$960 |
| Setup, once | 4 h × $50/h | −$200 |
| Year one | −$960 − $200 | −$1,160 |
- Not counting your time (28 h in year one): saves $240 in year one.
Year one = 12 × (what you pay now − the new cost − upkeep) − setup, with your time at $50/hour. Prices are public list prices; hours are Stackcut’s planning estimates. Price it for your team
How we got this
| You pay now: Pipedrive Lite | $24/user/mo × 1 user | $24/mo |
|---|---|---|
| New cost: Twenty | a small server (modeled) | $5/mo |
| Your upkeep time | 2 h/mo × $50/h | $100/mo |
| Difference each month | $24 − $5 − $100 | −$81/mo |
| Over 12 months | −$81 × 12 | −$972 |
| Setup, once | 4 h × $50/h | −$200 |
| Year one | −$972 − $200 | −$1,172 |
- Not counting your time (28 h in year one): saves $228 in year one.
Year one = 12 × (what you pay now − the new cost − upkeep) − setup, with your time at $50/hour. Prices are public list prices; hours are Stackcut’s planning estimates. Price it for your team
How we got this
| You pay now: HubSpot Sales Hub Starter | $10/user/mo × 1 user | $10/mo |
|---|---|---|
| New cost: Twenty | a small server (modeled) | $5/mo |
| Your upkeep time | 2 h/mo × $50/h | $100/mo |
| Difference each month | $10 − $5 − $100 | −$95/mo |
| Over 12 months | −$95 × 12 | −$1,140 |
| Setup, once | 4 h × $50/h | −$200 |
| Year one | −$1,140 − $200 | −$1,340 |
- Not counting your time (28 h in year one): saves $60 in year one.
Year one = 12 × (what you pay now − the new cost − upkeep) − setup, with your time at $50/hour. Prices are public list prices; hours are Stackcut’s planning estimates. Price it for your team
How we got this
| You pay now: HubSpot Sales Hub Starter | $10/user/mo × 1 user | $10/mo |
|---|---|---|
| New cost: EspoCRM | a small server (modeled) | $5/mo |
| Your upkeep time | 2 h/mo × $50/h | $100/mo |
| Difference each month | $10 − $5 − $100 | −$95/mo |
| Over 12 months | −$95 × 12 | −$1,140 |
| Setup, once | 4 h × $50/h | −$200 |
| Year one | −$1,140 − $200 | −$1,340 |
- Not counting your time (28 h in year one): saves $60 in year one.
Year one = 12 × (what you pay now − the new cost − upkeep) − setup, with your time at $50/hour. Prices are public list prices; hours are Stackcut’s planning estimates. Price it for your team
How to leave Salesforce
Export first
- All your data: Setup → Data Export → Export Now or Schedule Export (ZIP of CSV files). Weekly exports need Enterprise, Performance or Unlimited; other editions export monthly. The ZIP files are deleted 48 hours after the email, not counting weekends. help.salesforce.com
Cancel
- Export your data first (see export).
- For a paid subscription, contact Salesforce's Customer Service Billing Department: its help page offers no self-serve cancel.
How to leave HubSpot
Export first
- Contacts, companies, deals and other records: CRM → the object → Export at the top right of the list view (CSV, XLSX, XLS). The download link arrives by email and expires after 30 days; very large exports arrive as several files in a ZIP. knowledge.hubspot.com
Cancel
- Export your records first (see export).
- Account name → Account & Billing → Subscriptions → Cancel auto-renewal.
- Takes effect
- On your renewal date, not right away. Mid-contract downgrades aren't permitted.
- Notice
- For a downgrade, contact HubSpot at least 5 business days before the renewal date. A notice of non-renewal goes to the billing admin.
- After you cancel
- The account moves to HubSpot's free tools. You can delete it once no subscription has an active commitment.
Tested by Stackcut
Small-team CRM seat → an agent skill that keeps the pipeline in CSV files you own
Question. Can a skill folder plus a small script keep a 1-3 person sales pipeline in plain CSV files, updated from pasted emails, call notes and LinkedIn messages with every change approved as a diff, well enough to replace a CRM seat? Path: Agent skill: keep a sales pipeline in CSV files you own.
Tested on. Stackcut CRM-in-files skill: A fictional two-person web studio (Juniper & Slate Studio) with a synthetic pipeline: 9 companies, 11 people, 10 deals in USD, EUR and GBP, and a week of emails, call notes and LinkedIn messages to process.
Numbers
- 14 synthetic cases, a file checker and a strict rubric
- Pass rate by round: 11, 13, 14 of 14; independent round (fresh runner, separate grader): 14 of 14
- The file changes passed the checker in every round; every failure was a draft or a summary line
- Author rounds were self-graded; the independent round used a fresh runner and a separate grader: 14 of 14
- About 2.6 to 4.6 minutes of agent time per round
Verdict
For one to three people who already pay for an agent, it keeps a working pipeline in files they own: contacts, companies, deals, tasks and a weekly report, with every change shown and approved first. It passed all 14 cases by round 3. An independent round (a fresh agent ran every case, a separate agent graded it) also passed 14 of 14. At HubSpot or Pipedrive seat prices it doesn't pay back in money once upkeep is counted; the gain is owning your data.
Good fit when
- One to three people, one editing at a time
- You already pay for an agent that runs skills and can run Python
- You want the pipeline in files you own (CSV, kept in git or a synced folder)
- A few dozen open deals, updated from messages you paste in
Not a fit when
- You need email or calendar sync, reminders, or a mobile app
- Several people update the pipeline at the same time, or you need permissions
- You want a pipeline board or dashboards
- HubSpot's free CRM (up to 2 users and 1,000 contacts) already covers you
What we found
- The file changes were never the problem: the checker passed all 14 cases in every round, including the dedupe, rename, revival, two-deal, two-currency, conflict and deletion cases. Every failure was in the words around them (reply drafts and report summaries).
- Drafts decide things for the person unless told not to. Round 1 drafts said 'Splitting the invoice into two payments works' and 'Yes, we can pick this up again' before the person had decided. A rule to leave '[you decide: ...]' fixed it.
- Fixing one rule broke another case. The new rule failed a round-2 draft that accepted a customer's move to a cheaper tier, because nothing listed what the studio sells. Writing the price list into business.md fixed it. Re-run every case after each change.
- Summaries drift from the numbers. Round 1's 'what needs attention' lines added a task date from the due list and called one deal the largest across three currencies. Limiting those lines to restating report lines fixed it.
- The script is what keeps the files valid. In our script tests it refused a second open task on a deal, an email address already on another contact, a company name already used as someone's former name, an unknown currency, a won deal without a close date, a stale update, and any delete without --confirm-delete.
- A GDPR-style deletion works inside the files: the contact, their tasks and activities, and their name in other rows are all removed, after a separate yes. It doesn't reach copies outside the files (git history, synced-drive versions, exports, the inbox). The skill says so; it can't erase them.
- Process slips hide when the runner knows the data: in round 3 the author's runner wrote three change-sets before running the lookup the dedupe case depends on. The independent round, run by a fresh agent that didn't know the data, passed all 14 cases.
How we tested
- Wrote the skill in the agentskills.io format: SKILL.md (steps, the exact output block, a Never list), six reference files for a fictional studio (pipeline rules, matching, deletion requests, drafts, business facts, file format) and scripts/crm.py (Python 3, standard library, no network).
- crm.py is the only thing that writes the five CSV files (contacts, companies, deals, tasks, activities). It previews every change as a diff. It refuses any change that would leave the files invalid, any update whose old value no longer matches, and any delete without --confirm-delete <id>. It also prints the due list and the weekly report.
- Before the first run we wrote 14 cases and a pass/fail rubric. The cases are a week of messages plus requests for the due list, the report, follow-up drafts and two out-of-scope asks. We also wrote the starting files and an answer key: the change-set we expected for each case, and the expected final files.
- The hard cases: a person writing from a second address, a renamed company, a lost deal that comes back, one email about two deals, a call with no next step, a GDPR-style deletion request, a quote shown in EUR and USD, a budget conflict, and spam.
- The checker (tests/check.py) applies the runner's change-set for each case to the expected state before that case, passing only the deletion the person approved, and compares the structured fields with the expected state after it. It also applies all change-sets in order from the start, checks that the deleted person's name and address appear nowhere, and checks that the due list and report are the script's own output. We tested the checker first: the answer key passes 14/14, and five planted mistakes fail exactly their cases (check.py --selftest).
- The author's rounds: the same agent (the skill's author) did a runner pass, reading only SKILL.md, references/ and cases.md, and then a separate grader pass (rubric, references, run file, checker output), because no Agent tool was available in that session. The runner knew the answer key, which is why an independent round followed.
- Between rounds we changed only the skill (SKILL.md and references/), never the outputs, and re-ran all 14 cases.
- After the skill was final, Stackcut ran one independent round: a fresh runner agent (skill folder and case inputs only) and a separate grader agent (rubric, references, checker). Result: 14 of 14.
Compare before you switch
Read the checked comparisons and workload guides for the tools this recipe replaces.
- Salesforce vs Bigin Checked migration details, public switching accounts.
- Salesforce vs Twenty Checked migration details, public switching accounts.
- Salesforce and HubSpot CRM alternatives in 2026: what to use instead, and what it really costs Compare the wider set of alternatives and their tradeoffs.
More CRM recipes
- Close: Sales CRM $430
- Attio: Sales CRM $370
- folk: Sales CRM $310
- Zoho CRM: Sales CRM $190
- Bigin: Sales CRM $58
All 139 recipes · Savings leaderboard · Every vendor, with prices and cancel steps