Snapshot 310420
Normalized text
Scripts and page chrome removed; this is what change detection compares.
How Fibbler keeps your data safe Before you connect your ads and CRM to anything, you should know exactly what happens to your data. Here's how we handle it. The quick answers are right below, and every detail your IT team will ask about is further down. Last updated: November 2026 SOC 2 Type 1 Sensiba audited us in September 2026, and Atoro pen tested us in June 2026. Our Type 2 audit is underway. Your data stays in the EU We host it on Google Cloud in Belgium, with a cache in Frankfurt. Company-level only We work with companies and deals. We don't read your contacts or anyone's personal ad activity. We only write when you ask We read your tools, and only write when you turn a feature on. We never touch budgets, bids or creatives. CRM data deleted in 30 days If you disconnect your CRM or cancel, the CRM data we stored is deleted within 30 days. 2FA on every plan Owners can require it for the whole workspace. You can sign in with Google or Microsoft, and full single sign-on is on our roadmap. What Fibbler can see and do in your tools We only read by default. Anything we write is something you turn on, and you can turn it off again any time. ToolReadsWrites, only if you turn it onNever LinkedIn Ads Reads: Campaigns, ad sets, performance, and which companies your ads reached and engaged. Writes, only if you turn it on: Company and job title exclusions, impression caps, and turning campaigns and ad sets on and off on a schedule. Never: Budgets, bids or creatives. Google Ads and Meta Ads Reads: Campaigns, keywords, search terms and performance (Google), campaigns and performance (Meta), plus which companies visit your site from your ads. Writes, only if you turn it on: Nothing Never: Anything in your ad accounts. We only read them. Your CRM Reads: Companies and deals: names, domains, amounts, stages and dates. Writes, only if you turn it on: Fibbler's own fields with your ad data, lists of warm accounts for sales, and new companies if you choose to. Never: Your contacts, or deleting anything. Claude and ChatGPT (MCP) Reads: Your Fibbler numbers, when you ask a question. Writes, only if you turn it on: Nothing Never: Anything. The MCP is read only. Need documents for your security review? The SOC 2 report, our policies and our sub-processors are in the Trust Center. For a DPA, an NDA or a filled-in security questionnaire, email support@fibbler.co. Open the Trust Center ↗ The details, for your security team Frequently asked questions Where is data stored? All customer and application data is hosted in the EU, on Google Cloud in the Belgium region, with a cache from Redis Cloud on AWS in Frankfurt (also in the EU). None of your data in the product leaves the EU. Four things do run outside the EEA: support conversations, marketing email, billing details and the advertising tags on our marketing website. Each of them is covered by Standard Contractual Clauses or the EU-US Data Privacy Framework, and we go through them in Where your data lives. Do you process personal data? Only a little. We have your account name and email, plus the IP addresses and browser signals we use to keep accounts secure (rate limits, request logs and a bot check at sign-up). We don't process individual contact records from your CRM, personal identifiers, LinkedIn messages or sensitive information of any kind. The company-level and deal data we take from your CRM (company names, domains, deal amounts and deal stages) is business data, not personal data, and we use it for attribution and analytics. What do you read in LinkedIn, Google Ads, Meta Ads and my CRM? We only read the company-level and deal data that the features you turn on actually need. We list it object by object and field by field, for every platform, in What we read and write in your tools. Do you store any of that data? Yes. We store your LinkedIn Ads data, your Google Ads and Meta Ads campaign data, and your CRM company and deal records (company names, domains, deal amounts, deal stages and related metadata) in our database, because that's what attribution, analytics and the MCP run on. All of it is encrypted at rest with AES-256. If you disconnect or cancel, the CRM data we stored is deleted within 30 days. And if you'd rather we didn't store CRM data at all, you can opt out by emailing support@fibbler.co. There's more on this in What we store, and for how long. Do you push anything back into my CRM? Only if you turn it on. CRM sync writes your ad engagement data into fields Fibbler creates on your Company or Account records. Ad signals keeps a list of accounts for each signal (a tag in Zoho CRM), writes its own fields saying why each account was sent, and leaves a note on the record in every CRM. On a record that already exists, we write to our own fields. We explain both features in What we write back into your CRM. Do you ever delete anything in my CRM? No. We never delete a record, a field or a property in your CRM. You'll see the word "delete" in some HubSpot permission names, because HubSpot bundles create, change and delete into one permission and doesn't offer anything narrower. We only use the create and change part. Do you sell customer data? No. We don't sell, resell or profile your data, not to anyone, and we never will. Do you support two-factor authentication (2FA)? Yes, on every plan at no extra cost. It uses time-based one-time passwords (TOTP), so any standard authenticator app works, like 1Password, Google Authenticator, Microsoft Authenticator or Authy. Workspace owners can require 2FA for every member of a workspace, and we check that on our server on every request in the app. There are no self-serve recovery codes today, so if you get locked out, email support@fibbler.co and we'll help you get back in. You can read the rest in Authentication below. Do you support single sign-on (SSO)? Partly. You can sign in with a Microsoft work account or a Google Workspace account, so your own identity provider, its MFA and its Conditional Access policies apply when your team signs in. We do not offer SAML single sign-on, enforcing SSO for a domain, or SCIM provisioning today. Do you have a security certification? Yes. We're SOC 2 Type 1 compliant. Our auditor, Sensiba, went through our controls for security, availability and confidentiality, and gave us a clean report in September 2026. The same audit checked us against the GDPR's data protection principles, and we passed that too. You can request the full report in our Trust Center, and read every control and policy behind it there. Our SOC 2 Type 2 audit is underway, also with Sensiba. We've also booked ISO 27001 with the same auditor, but that audit hasn't started, so we're not ISO 27001 certified yet. Google Cloud, where we host Fibbler, is SOC 2 Type II and ISO 27001 certified. Do you run penetration tests? Yes, once a year. The latest one was done in June 2026 by Atoro, an independent security firm. On top of that we do our own security reviews, and we get automatic alerts for vulnerable dependencies and cloud misconfigurations. The June test found seven issues: five Medium, two Low, and nothing High or Critical. We've fixed all seven, and the fixes are live in production. They were about how long sessions stayed valid after a password or 2FA change, input validation on free-text fields, our Content Security Policy, our SPF, DMARC and CAA records, and a few older TLS ciphers. app.fibbler.co now gets an A+ on SSL Labs. Do you run background checks on new hires? No. We don't do criminal record, credit or employment history checks. Instead, at least two people at Fibbler interview every candidate and assess them against the role before we make an offer, and we write that assessment down and keep it. Everyone signs confidentiality and information security terms as part of their employment agreement, and gets access based on their role. Our SOC 2 controls describe it the same way. Is our data used in benchmarks, and can we opt out? Yes, your numbers are part of the benchmarks every customer compares against, but only aggregated and anonymized. Nobody can see another company's numbers. If you don't want to be included, you can opt out at any time. Benchmarks work both ways, so a workspace that opts out stops seeing them too. Can Fibbler connect to Salesforce as a dedicated integration user? Yes. Fibbler works with whichever user connects it, so you can give it a dedicated user with narrower permissions than an admin. For attribution and reporting, that user can be fully read-only. It only needs more if you turn on features that write back to Salesforce. There's more in our Salesforce setup guide. Do you have an incident response or recovery plan? Yes. Our business continuity, incident management and risk management policies are all in our Trust Center. In practice, we back up stateful systems every day and can restore customer-critical infrastructure within 24 hours. If there's ever a personal data breach, we tell the customers affected without undue delay, and within 48 hours. What we read and write in your tools No certification answers this part for you, so we keep all of it on this page. We only access the company-level and deal data that the analytics, attribution and reporting features you turn on actually need. LinkedIn Ads We access your LinkedIn Ads account through the LinkedIn Marketing API, and this is what we use: Data Access Fields Why it's used Campaigns and ad sets Read name, ID, status, objective To show campaign performance and attribute it Performance metrics Read impressions, clicks, spend, engagements For attribution and reporting Company engagement Read company name, domain, engagement data To match ad engagement to CRM accounts Ad set targeting Write targeting exclusions (job titles, companies) For audience exclusions and impression caps (only if you turn them on) Campaign status Write on or paused For campaign scheduling, turning campaigns and ad sets on and off on the days and hours you pick (only if you set a schedule) We only write to LinkedIn when you set one of these features up, and we never change budgets, bids or creatives. HubSpot, Salesforce, Attio, Pipedrive, Zoho CRM & Microsoft Dynamics 365 This is what we access in your CRM: Object Access Fields Why it's used Company/Account Read name, domains, LinkedIn page, ID To match CRM records to campaigns Opportunity/Deal Read name, amount, status, created/close dates, ID For revenue attribution and funnel reporting Custom fields Read field names, and the values of the deal fields we use (in Attio, the whole deal record) So you can map ad data into your CRM Custom fields Write Fibbler-created fields on Company/Account (HubSpot, Salesforce, Attio, Pipedrive, Zoho CRM & Microsoft Dynamics 365) For CRM sync, which creates fields to hold your LinkedIn, Google Ads or Meta Ads engagement data (only if you turn it on) Custom fields Write Fibbler-created signal fields on Company/Account: signal type, date, name, channels and reason For Ad signals, so you can see why an account was sent and filter on it (only if you turn it on) Company/Account Create on new records only: name and domain, and depending on the CRM, industry, employee count, country, LinkedIn page and lifecycle stage, plus a field saying it was created by Fibbler For Ad signals, and only once you turn a signal on. If you chose this when you set up the signal, it creates an account that's missing and marks it as created by Fibbler List / View / Filter / Tag Create & update One list of companies per signal, and which companies are in it. In Zoho CRM, which has no lists we can fill, one tag per signal on the Accounts it sends, and a saved view per signal that shows them For Ad signals. It's the queue your team works from. An account leaves the list once the signal hasn't picked it for 30 days, and we don't touch the record itself Note Create A note on the Company/Account, in every CRM we support For Ad signals. It's the one-line reason a rep reads. In HubSpot it uses the same company write permission, with no access to your contacts CRM sync and Ad signals are both off until you turn them on, and both work with every CRM we support (HubSpot, Salesforce, Attio, Pipedrive, Zoho CRM, and Microsoft Dynamics 365). Everything else we do in your CRM is read-only. A word on lists, because the permission is broader than what we use it for. In HubSpot the list permission also covers contacts and deals, because HubSpot doesn't offer anything narrower. Every list we create is a list of companies, and we don't browse the other lists in your account. There's one exception in HubSpot: if you already have a static company list with the same name as your signal, we use that list instead of making a second one, and add the signal's companies to it. On a record that already exists, we write to our own fields. We never change the owner, the lifecycle stage or anything else your team looks after. We only fill in details like industry and size on a record we create. Google Ads (optional add-on) We get campaign and performance data through the Google Ads API. We only read your Google Ads data, and we never change anything in your account. Data Access Fields Why it's used Campaigns Read name, ID, status, type To display campaign performance and attribution Ad groups, ads, keywords and search terms Read name, ID, keyword and search term text, match type, bidding strategy, metrics To group performance and connect search terms to pipeline Performance metrics Read clicks, impressions, spend, conversions For attribution and reporting Website visitor data (via Leadfeeder) Read company name, domain, industry, size and location, plus the landing page, pages viewed, visit time and the campaign from your ad's tracking parameters To match website visits to Google Ads campaigns and CRM deals Meta Ads (optional add-on) We get campaign and performance data through the Meta Marketing API. We only read your Meta Ads data, and we never change anything in your account. Data Access Fields Why it's used Ad accounts Read name, ID, currency, status So you can pick which ad accounts we read Campaigns Read name, ID, status, objective To display campaign performance and attribution Performance metrics Read impressions, clicks, engagements, spend For attribution and reporting Website visitor data (via Leadfeeder) Read company name, domain, industry, size and location, plus the landing page, pages viewed, visit time and the campaign from your ad's tracking parameters To match website visits to Meta Ads campaigns and CRM deals Integrations you connect The platforms above are not sub-processors. They already hold your data as your own system of record, and we only connect after you give us access through that platform's own OAuth flow. We don't read anything until you connect, and you can take our access away at any time, from the Data sources page or in the platform itself. Ad platforms: LinkedIn Ads, Google Ads, Meta Ads CRMs: HubSpot, Salesforce, Attio, Pipedrive, Zoho CRM, Microsoft Dynamics 365 What each consent screen asks for When you connect a tool, it shows you its own screen with the permissions Fibbler asks for. This is what you'll see for the ad platforms: LinkedIn Ads: reporting, your company pages, and managing ads. The last one is only used for the features you turn on (exclusions, impression caps and scheduling) Google Ads: Google has no read-only option, so its screen says Fibbler can manage your Google Ads. But we only ever read Meta Ads: reading your ads data, and nothing else For the CRMs, we list every permission and what it's for on each CRM's setup page in our docs. What we store, and for how long We store your LinkedIn Ads data, your Google Ads and Meta Ads campaign data, and your CRM data in our database, because that's what attribution, analytics and the MCP run on. That means campaign performance metrics, audience exclusions, attribution data, CRM company records (name, domain), and deal or opportunity records (amount, stage, dates). We sync your CRM data on a regular schedule and keep it up to date for as long as the integration is connected. If you disconnect or cancel, the CRM data we stored is deleted within 30 days. And if you'd rather we didn't store CRM data at all, you can opt out by emailing support@fibbler.co. When a trial or a plan ends, we keep your data for 30 days, so nothing is lost if you come back. Logs are kept for 90 days. Backups are kept for 7 days, so anything we delete is gone from them a week later. Our database is encrypted at rest with AES-256, and the access tokens you give us when you connect a tool are encrypted again with their own key. MCP: Fibbler in Claude and ChatGPT The MCP lets you ask about your Fibbler data in Claude, ChatGPT and other AI tools. It is read only: it can't change anything in Fibbler, your CRM or your ad accounts. It answers for one workspace at a time, and only with data that workspace can already see in Fibbler Claude and ChatGPT connect with your Fibbler login. The access that gives them lasts an hour at a time, and you can remove the connection in the AI tool whenever you want Other tools use an API key. We only store a hashed version of it, it is shown once, and you can revoke it in Settings Every request is logged, and each workspace has a daily limit Once Claude or ChatGPT has an answer, it's covered by your own agreement with them. Fibbler doesn't use your data to train any model What we write back into your CRM We only write to your CRM if you turn it on yourself. Two features do that. CRM sync puts your LinkedIn, Google Ads or Meta Ads engagement data into HubSpot, Salesforce, Attio, Pipedrive, Zoho CRM or Microsoft Dynamics 365. It does that by creating custom fields on your Company or Account records. In Microsoft Dynamics 365 those fields sit in a solution called Fibbler Integration, so everything we added is in one place. Ad signals does more, so here is exactly what it does. For each signal you build, it creates a list, view or filter named after that signal, keeps the right accounts in it, and writes its own fields on each account saying why it was sent. In Zoho CRM the list is a tag on each Account. It also leaves a note on the record, in every CRM. If an account is missing from your CRM and you chose this when you set up the signal, it creates the account and marks it as created by Fibbler, so you can find them all again. Signals are optional, and nothing happens until you set one up and turn it on. You control all of it We never change a field your team owns on a record that already exists We never delete a record, a field or a property in your CRM. The only thing we ever take away is an account from a signal's own list, once the signal hasn't picked it for 30 days, and the record itself stays Authentication & 2FA You sign in with your Google or Microsoft account, or with your email and password. Two-factor authentication is there for every user on every plan at no extra cost, and workspace owners can require it for their whole team. This is how it works: Google and Microsoft sign-in use OpenID Connect with PKCE. We only accept an address the provider vouches for: on Google a Gmail or Google Workspace address, on Microsoft an address whose domain the organisation has verified, or a personal Microsoft account. We ask only for your name and email address, and we store no Google or Microsoft tokens It uses time-based one-time passwords (TOTP, RFC 6238), which are 6-digit codes that change every 30 seconds. Any authenticator app works, like 1Password, Google Authenticator, Microsoft Authenticator or Authy. We don't use SMS or email codes as a second factor With 2FA on, a password alone doesn't start a session. We only issue the session cookie once the code is accepted, and a password reset doesn't get around that Sensitive actions ask you to verify again if more than two hours have passed. Sessions last 7 days at most, and we track them on our server so we can revoke them centrally Turning 2FA off needs a valid code, and it signs out every other session on the account. Changing your password does the same Workspace owners can require 2FA for every member of a workspace. We check it on our server on every request in the app, so members who haven't set it up can't get into the workspace until they do If you sign in with Google or Microsoft and have 2FA on, we still ask for your code We don't offer recovery codes, SAML single sign-on (like Okta), enforcing SSO for a domain, SCIM provisioning, or hardware security keys today. If you lose the device your authenticator is on, email support@fibbler.co and we'll help you get back in Where your data lives Both the app and the database run on Google Cloud Platform, in one EU region (Belgium). We also use a cache from Redis Cloud on AWS in Frankfurt, which is in the EU too. Google Cloud is SOC 2 Type II and ISO 27001 certified, and you can read how Google handles security here. Transfer Impact Assessment (TIA) We've done a Transfer Impact Assessment, as GDPR Article 46 requires. All customer and application data is hosted in the EEA. That means your ad platform data, your CRM data and everything you see in the Fibbler app. Four things do run outside the EEA: support conversations in Intercom, marketing email in Loops, payment and billing details in Stripe, and the advertising tags on our marketing website. Each of them is covered by Standard Contractual Clauses or the EU-US Data Privacy Framework. Some of the business tools we run the company on are based in the US and covered the same way, but none of them hold your ad platform or CRM data. Sub-processors We publish every vendor that processes customer data on your behalf, with its location and what we use it for, in our Trust Center sub-processor list. We keep one list in one place, so you never have to work out which of two pages is the current one. Customer and application data stays in the European Economic Area, apart from the four things in the Transfer Impact Assessment above. Every sub-processor is covered by GDPR data processing terms. For one of them that's an agreement we signed separately, and for the rest it's the provider's own terms, which we accepted. We tell customers at least 14 days before we add or replace a sub-processor that processes customer data, and you can object to the change. We do that by publishing the change in the Trust Center list and by email. This commitment applies to changes made after 28 September 2026, the date we set the baseline for the list, and the change log below records every change since then. The ad platforms and CRMs you connect are not sub-processors (we explain why under Integrations you connect). Separately, we run the company on ordinary business tools: email, chat, source control, dependency scanning and issue tracking. None of them hold your ad platform or CRM data. Where they hold personal data at all, like your email address in a message you sent us, we're the controller of it. That's covered by our Privacy Policy rather than a data processing agreement, so the 14 days' notice commitment doesn't apply to those vendors. They're all in our vendor register, and we'll name them if you ask. Website visitor identification Leadfeeder is the one sub-processor that can run on your own website instead of ours. It works out which companies visit from their IP addresses, and it's only on your site if you install the Fibbler tracking script for the Google Ads or Meta Ads add-on. Leadfeeder is ISO 27001 and ISO 27701 certified, hosts all data in the EU, and we have a Data Processing Agreement with them. In the app at app.fibbler.co we use PostHog to see how people use the product. It's tied to the signed-in user and stores nothing in the browser. The app sets no advertising or analytics cookies, only the ones needed for sign-in and security. The analytics and advertising tags on our marketing website are a separate thing. They never touch your ad platform or CRM data, and they're covered by our Privacy Policy. Change log Date Change 15 November 2026 The Privacy Policy now lists everything product analytics records in the app: approximate location derived from the IP address (which is not stored), campaign parameters and ad click IDs from sign-up, and the inputs to MCP tool calls. It previously said MCP query content was not logged. 12 November 2026 The app no longer stores anything in the browser for product analytics: PostHog runs without cookies or local storage, and usage is tied to the signed-in user. The only cookies left are those needed for sign-in and security. 12 November 2026 Added sign-in with Google and Microsoft accounts to Authentication, and a question on single sign-on. SAML SSO, enforced SSO and SCIM are still not offered. 11 November 2026 Corrected what the app sets. This page said app.fibbler.co sets no tracking cookies, but the app uses PostHog, with a first-party cookie, to see how customers use the product. It sets no advertising cookies. The Privacy Policy now describes this as its own purpose. 5 October 2026 The sub-processor list now lives in the Trust Center only. We didn't add, remove or replace any vendor with this change. We had the same list in two places, and two lists end up disagreeing. We still tell you about changes in that list and by email, as before. This page keeps what the Trust Center does not cover, which is what Fibbler reads and writes in the platforms you connect. 28 September 2026 Baseline. We reviewed our sub-processors against our full vendor register and corrected the list. We added Intercom and Amazon Web Services. We already used both, we just hadn't listed them. We took Google Workspace, Slack, GitHub, GitHub Dependabot, npm and Jira off the list, because they don't process your data for you. Removed the Google Ads API entry, because the ad platforms and CRMs you connect are data sources rather than sub-processors. Recorded that Intercom's AI agent, Fin, is in use, and that support conversation content therefore reaches the model providers Intercom lists as its own sub-processors. Named a country and region for every vendor that previously said only "EU-hosted", and corrected the Stripe entry, which claimed no personal data was shared. The 14 days' notice commitment applies to changes made after this date. Technical & Organizational Measures (TOMs) We publish the full list of controls from our SOC 2 audit, and the policies behind them (access control, vendor management, incident management, risk assessment, endpoint security and the rest), in our Trust Center. These are the measures customers ask us about the most: Our controls were independently audited for SOC 2 Type 1 (Sensiba, September 2026) All data transfers and database connections are encrypted with TLS, and everything we store is encrypted at rest with AES-256 The database accepts no direct connections from the internet. It has no allowed networks, access goes through Google's IAM-authenticated proxy, and only encrypted connections are accepted Database accounts hold the minimum permissions they need Two-factor authentication (TOTP) is available to all users, and owners can enforce it across a whole workspace. Passwords are hashed with bcrypt and checked against the Have I Been Pwned breached-password database, and every authentication endpoint is rate limited We back up stateful systems every day, encrypted and stored separately, and we can restore customer-critical infrastructure within 24 hours We have real-time monitoring and alerting, we patch regularly for security, and we scan dependencies with vulnerability alerts A third party runs a penetration test every year (the latest was by Atoro, June 2026) Production access requires two-factor authentication, enforced across our Google Cloud organization Every code change is reviewed by a second person before it ships Only people we've authorized, with a specific business need, can reach the database or production infrastructure. They may look at production data to investigate a fault or answer a support request. Documents, DPA and NDA Trust Center (SOC 2 Type 1 report, controls, policies and sub-processors) Privacy Policy Terms of Service We also have a standard DPA (aligned with GDPR) and a Mutual NDA for when you're evaluating us. Email support@fibbler.co to get either one, or with any other security question. Found a security issue? Email support@fibbler.co. We don't run a paid bug bounty, but we read and answer every report.