Marketing agencies and consultancies
Workamajig, HubSpot, Harvest and Asana data export: what agency software terms say
On this page
The short answer
Agency software terms mostly say the agency owns its data. HubSpot, Harvest, Float, Asana, monday.com, Function Point, Productive, Teamwork, Salesforce and Figma each say the customer keeps ownership of what it puts in. The catches are elsewhere: Float deletes at cancellation, Harvest may delete after termination, HubSpot gives most paid hubs no access after the term ends, and HubSpot, Asana and Slack limit what third-party apps may do with your data.
So export your Workamajig, HubSpot, Harvest and Asana data yourself, while subscriptions are active, and never through a buyer's connector app. Workamajig's license agreement is not public.
Agency software data terms at a glance
| System | Who owns the data | After you cancel | AI and access notes |
|---|---|---|---|
| Workamajig | License agreement not public; customer controls what it processes (privacy policy) | Not public | Ask for your signed agreement |
| Function Point | "exclusive property of Customer" (6) | May permit 30 days of access, then deletion (1.6) | No access by methods it has not provided or approved (1.7(d)) |
| Productive | "Customer will own all Customer Data" | No obligation to keep data after termination | Export allowed during the subscription |
| Teamwork.com | Customer owns all Customer Data (8.2) | Written request within 10 days for the latest backup (16.5(d)) | Separate AI policy not reviewed |
| Harvest | "you own and retain all rights" (7.1) | May delete at its sole discretion (6.5) | AUP bars training on Harvest AI prompts or output |
| Float | You retain all rights to your content | Scheduled for deletion upon cancellation | No training on paid or enterprise content |
| Asana | Customer retains all right, title and interest (7.2) | Export on written notice within 30 days (5.4) | API data may not be shared with third parties (API Terms 4.10) |
| monday.com | Customer retains all right, title, interest and control (3.1) | Exporting before termination is the customer's sole liability (11.4) | No access other than its interface, e.g. scraping (AUP 1.1) |
| HubSpot | "You own and retain all rights" (5.1) | Most paid hubs: no access after the term (PST 5.1.1) | May train its AI unless you opt out (5.3.1); apps may not train AI on customer data (Dev 4.B.8) |
| Salesforce | Acquires no right, title or interest in Customer Data (6.2) | Export on request within 30 days (2.2) | AI features process data only per your instructions (2.5) |
| Slack | Workspace export by owners and admins | Plan decides what an export covers | Third-party apps may not train an LLM on API data |
| Figma | Customer retains all right, title and interest (2.8) | 30 days to retrieve content (8.2(b)) | Content training on by default for Starter and Professional |
Owning the data in the tool is not the same as owning everything inside it. A client's brief stored in Asana is still the client's confidential information. Our guide to who owns client work covers that second layer.
Workamajig, Function Point, Productive and Teamwork data export
Agency management systems hold the single most valuable record an agency has: a job number that links the estimate, the approved scope, every time entry, the purchase orders, the invoices and the final margin. Five or more years of that history in one system is, in our view, the best single signal that an agency's records are worth licensing.
- Workamajig. Its license agreement is not published. The privacy policy, dated February 5, 2026, refers to customers “under the Workamajig License Agreement” and says each customer controls the information it processes. It also says Google Workspace APIs and Workamajig APIs are not used to train generalized AI or ML models (source: Workamajig Privacy Policy (Feb 5, 2026)). Ask Workamajig for your signed agreement and a full export of jobs, estimates, timesheets and billing before any change.
- Function Point. “Customer Content is the exclusive property of Customer” (6). After termination Function Point “may permit” 30 days of access to remove content, after which data is permanently deleted (1.6). Customers may not access the service by any method Function Point has not provided or approved (1.7(d)) (source: Function Point Terms of Service (Oct 27, 2020)). The terms are dated October 27, 2020.
- Productive. The June 15, 2026 customer terms say the customer “will own all Customer Data”, allow export during the subscription, and say Productive has no obligation to keep or provide data after termination (source: Productive Customer Terms of Service (Jun 15, 2026)).
- Teamwork.com. The customer owns all Customer Data (8.2). After termination, Teamwork may dispose of it unless you ask in writing, within ten days, for the most recent backup (16.5(d)) (source: Teamwork.com Terms of Service). Ten days is the shortest window in this group.
Harvest and Float: export data before you cancel
Time and resourcing history is the record buyers describe as task execution history: years of hours by role, task and job, planned against actual (source: program terms, Regulr buyer directory). It is also the record agencies lose first.
- Harvest. “As between you and us, you own and retain all rights to the Customer Data” (7.1). After termination, Harvest may delete your data “at our sole discretion”, and you are responsible for backups (6.5) (source: Harvest Terms of Service (Apr 15, 2026)). Its acceptable use policy bars using prompts or output generated by Harvest's AI features to train an AI model (source: Harvest Acceptable Use Policy), so leave AI-feature output out of any export.
- Float. You retain all rights to your content, but “All of your content will be scheduled for deletion from the Service upon cancellation” and cannot be recovered. Float says it will not use content from paid or enterprise accounts to train its models (source: Float Terms of Service v2.5 (Oct 7, 2026)).
Export Harvest and Float data first
Neither gives a guaranteed window after you cancel. Export time entries, projects, clients, people, rates and invoices to CSV while the account is active, and check that each entry still carries its job or project code.
Asana export all data, and monday.com data retention
- Asana. “Customer and its licensors will retain all right, title, and interest” in Customer Data, and admins can control product-improvement use in the Admin Console (7.2). After expiration or termination, Asana makes data available for export only if it gets written notice within 30 days (5.4) (source: Asana Subscriber Terms (Sept 1, 2026)). Its API terms bar sharing data obtained through the APIs with any third party (4.10) (source: Asana API Terms (Mar 14, 2022)).
- monday.com. “Customer retains all right, title, interest and control” (3.1). “It is Customer's sole liability to export the Customer Data prior to such termination or expiration”, after which data is deleted (11.4) (source: monday.com Terms of Service (May 5, 2026)). Its acceptable use policy bars access by any means other than its interface, with scraping, spidering and crawling as examples (1.1) (source: monday.com Acceptable Use Policy (Dec 20, 2023)).
Project histories are valuable because they show multi-step work with dependencies, approvals and delays. Task text often restates client strategy, though, so it is reviewed before it goes into a license.
HubSpot export all data, and Salesforce
- HubSpot. “You own and retain all rights to the Customer Materials and Customer Data” (5.1). HubSpot may use Customer Data to train its AI models, and you can opt out in account settings (5.3.1) (source: HubSpot Customer Terms of Service (Sept 16, 2026)). For most paid hubs HubSpot says it will not provide access to Customer Data after the subscription term (5.1.1); Marketing Hub Professional and Enterprise customers who ask in writing within 30 days get temporary access to retrieve data (3.3.3) (source: HubSpot Product Specific Terms (Sept 16, 2026)). Its developer terms bar apps from using customer data to train, fine-tune or develop AI models or to generate labeled data, except a model built solely on one customer's data for that customer's benefit with its permission (4.B.8) (source: HubSpot Developer Terms (May 4, 2026)).
- Salesforce. Salesforce “acquires no right, title or interest” in Customer Data (6.2), makes it available for export on a request made within 30 days after termination (2.2), and says AI features process Customer Data only according to the customer's documented instructions (2.5) (source: Salesforce Main Services Agreement (Oct 15, 2026)).
CRM history matters because new-business pipelines carry outcomes: a proposal sent, a pitch won or lost. Prospect contact details come out; deal stages, dates, values and win or loss reasons usually stay.
Slack export all data, and Figma files
- Slack. Slack's API terms, effective October 10, 2025, bar third-party apps from using API data to train a large language model and from bulk exporting message and file data except where another agreement allows it (source: Slack API Terms of Service (Oct 10, 2025)). The owner's own export depends on plan: Free and Pro exports cover public channels, Free exports include file links only from the last 90 days, and Business+ owners can apply to export private channels and DMs (source: Slack Help Center, export your workspace data).
- Figma. The customer “retains all right, title, and interest in Customer Content” (2.8), and content is available for retrieval for 30 days after termination (8.2(b)) (source: Figma Terms of Service (Mar 11, 2026)). Content training is on by default for Starter and Professional teams, and admins can switch it off (source: Figma Help Center, content training settings by plan). Design files are mostly client work, so they are the files most often scoped out.
What each agency tool's export holds for a buyer
Buyers value records that link across systems, so the export from each tool matters most for the fields that tie it to a job or a client token. Here is what usually comes out of each kind of system and why buyers care.
| Tool type | Fields to keep | Why buyers value it | Usually removed |
|---|---|---|---|
| Agency management (Workamajig, Function Point, Productive, Teamwork) | Job number, estimate lines, approved scope, change orders, POs, invoices, status dates | Estimate against actual, with scope changes explained | Client names (tokenized), client pricing marked confidential |
| Time tracking (Harvest, Float) | Date, hours, role, task, job code, planned against actual | Years of how expert work is actually spent | Staff names, individual pay and rates |
| Project tools (Asana, monday.com) | Tasks, dependencies, assignees by role, due and done dates, comments | Multi-step workflows with approvals and delays | Client strategy in task text; staff names |
| CRM (HubSpot, Salesforce) | Deal stages, dates, values, source, won or lost and the reason | B2B selling with outcomes | Prospect contacts and emails |
| Chat and email (Slack, Teams, Gmail, Outlook) | Internal threads by channel and date | The reasoning behind decisions | Names, phones, HR matters, client channels |
| Docs and wikis (Notion, Confluence, Drive) | Playbooks, SOPs, templates, proposals, post-mortems | The firm's know-how written down | Client deliverables and client research |
Two practical tips. Keep the original IDs (job numbers, project IDs, deal IDs) in every export, because they are what lets a buyer join the timesheet to the estimate to the invoice after names are replaced. And export more history than you think you need: buyers value years, and older jobs often have the cleanest outcomes.
Owner exports, not connector apps
Read together, these terms point one way. Vendors let you, the customer, export and use your own data. They restrict third-party apps that pull data through their APIs, especially for AI training. HubSpot's developer terms, Asana's API terms and Slack's API terms each say a version of this. So the records for a license should come from your own in-app exports or your own API use, delivered to the buyer you chose, and not from a buyer's connector logged into your accounts.
Some buyers offer to process data inside your environment or to have a specialist strip identifying fields before anyone sees it (source: program terms, Regulr buyer directory). Our buyer directory lists the programs and links to their published terms. Ask any buyer how it collects records from each tool, and decide the method in writing.
What to export, in what order
- Time and resourcing (Harvest, Float). No guaranteed window after cancellation.
- Agency management (Workamajig, Function Point, Productive, Teamwork). Jobs, estimates, time, POs and invoices, with job codes intact. Teamwork's backup request window is ten days.
- CRM (HubSpot, Salesforce). Deals, companies, activities and notes. HubSpot gives most paid hubs no access after the term.
- Project tools (Asana, monday.com). Projects, tasks, comments and status updates.
- Chat and email (Slack, Teams, Google Workspace, Microsoft 365). Owner or admin exports only.
- Files (Drive, SharePoint, Notion, Figma). Playbooks, templates and proposals first; client deliverables stay where they are.
Law firms face a similar sorting job in their practice-management systems; our practice management data export guide shows how that category handles it. For what buyers want and who qualifies, see the marketing agency and consultancy data licensing guide. If you are winding down or selling, our guide to selling a marketing agency covers the timing, and you can see what your agency's exported records could qualify for.
How Regulr is paid, and how we write these pages
Regulr is a sell-side advisor. When a business we introduce signs a license, we may be paid a fee by the buyer or through a program's referral terms. We list every program we can verify whether or not it pays us, order is never set by fees, and every figure is the company's own advertised term with a link to where it says so. Regulr has not closed a data deal yet, so nothing here describes our results.
Frequently asked questions
Who owns the data in HubSpot, Harvest or Asana?
Under the terms we read on October 18, 2026, the customer does. HubSpot says you own and retain all rights to your Customer Data (5.1), Harvest says the same (7.1), and Asana says the customer retains all right, title and interest (7.2). Float, monday.com, Function Point, Productive, Teamwork, Salesforce and Figma use similar language. Owning the data in the tool does not settle whether a client owns the content inside it, which is a separate question.
How do I export all data from Asana or monday.com?
Use the export features in each product's admin settings while your subscription is active. Asana makes data available for export after termination only if it receives written notice within 30 days (5.4). monday.com says exporting before termination is the customer's sole responsibility (11.4). Do not rely on a third-party app to pull it for a buyer, since both vendors limit that route.
Can I export my Workamajig data?
Workamajig's license agreement is not public, so we could not read its export or termination terms. Its privacy policy refers to the Workamajig License Agreement and says each customer controls the information it processes. Ask Workamajig for your signed agreement and a full export of jobs, estimates, timesheets and invoices before you change anything.
What happens to Harvest or Float data when I cancel?
Float's terms say all content is scheduled for deletion upon cancellation and cannot be recovered. Harvest's terms say it may delete your data after termination at its sole discretion and that you are responsible for backups. Export time entries, projects, clients and invoices before you cancel either one.
Does HubSpot train AI on my agency's CRM data?
HubSpot's terms say it may use Customer Data to train its AI models and that you can opt out in your account settings (5.3.1). That is HubSpot's own use and does not limit what you do with your data. Separately, HubSpot's developer terms bar apps from training AI models on customer data, except a model built for one customer's own benefit with permission (4.B.8).
Can an AI buyer connect to our Slack to collect the data?
That is the route to avoid. Slack's API terms bar third-party apps from using API data to train a large language model and from bulk export except where another agreement allows it. The workspace owner's own export is the usual route, and what it covers depends on your plan: Free and Pro exports cover public channels, while Business+ can apply to export private channels and DMs.
Do I need to keep paying for software to keep the data?
Until you have exported it, often yes. Several agency tools give no guaranteed window after cancellation, and others give 30 days only on written request. Export first, check the files open and link by job, then downgrade or cancel.
Sources
Vendor terms read on each vendor's own page on October 18, 2026; Teamwork's terms show no date and Harvest's acceptable use policy shows none. Workamajig's license agreement is not public. Program terms checked against their own pages on October 15, 2026; our buyer directory lists the programs and links to their published terms.
- program terms, Regulr buyer directory
- Workamajig Privacy Policy (Feb 5, 2026)
- Function Point Terms of Service (Oct 27, 2020)
- Productive Customer Terms of Service (Jun 15, 2026)
- Teamwork.com Terms of Service
- Harvest Terms of Service (Apr 15, 2026)
- Harvest Acceptable Use Policy
- Float Terms of Service v2.5 (Oct 7, 2026)
- Asana Subscriber Terms (Sept 1, 2026)
- Asana API Terms (Mar 14, 2022)
- monday.com Terms of Service (May 5, 2026)
- monday.com Acceptable Use Policy (Dec 20, 2023)
- HubSpot Customer Terms of Service (Sept 16, 2026)
- HubSpot Product Specific Terms (Sept 16, 2026)
- HubSpot Developer Terms (May 4, 2026)
- Salesforce Main Services Agreement (Oct 15, 2026)
- Slack API Terms of Service (Oct 10, 2025)
- Slack Help Center, export your workspace data
- Figma Terms of Service (Mar 11, 2026)
- Figma Help Center, content training settings by plan
Brian Boesen · Founder, Regulr AI
Brian runs Regulr AI, which works for established businesses that want to license de-identified records to AI companies. He checks each program's published terms against its own pages and updates these guides when they change. Last updated October 18, 2026.
This page is general information, not legal advice, and Regulr is not a law firm. Rules vary by state, by profession and by contract, and they apply to your facts, not to a category. You approve exactly what is in scope before anything is shared.
Keep reading