Friday, 20 November 2026
/
10 min read

SOC 2 for SaaS: What Buyers Want and How to Build for It

Scroll ↓
SOC 2 for SaaS: What Buyers Want and How to Build for It
Friday, 20 November 2026
/
10 min read
by Format-3

Share article


    {
    "@graph": [
    {
    "@type": "Article",
    "image": {
    "url": "https://media.babylovegrowth.ai/blog-images/organization-13731/1795039278007_Security-reviewer-reading-a-SOC-2-report.jpeg",
    "@type": "ImageObject",
    "caption": "Security reviewer reading a SOC 2 report"
    },
    "author": {
    "url": "https://format-3.co",
    "name": "Format-3",
    "@type": "Organization"
    },
    "headline": "SOC 2 for SaaS: What Buyers Want and How to Build for It",
    "publisher": {
    "url": "https://format-3.co",
    "name": "Format-3",
    "@type": "Organization"
    },
    "inLanguage": "en-GB",
    "description": "SOC 2 isn't a law, yet enterprise buyers treat it as one. Learn what it covers, Type 1 vs Type 2, the timeline, and how to build a SaaS product that can pass.",
    "dateModified": "2026-11-20T00:00:00.000Z",
    "datePublished": "2026-11-20T00:00:00.000Z"
    },
    {
    "@type": "BreadcrumbList",
    "itemListElement": [
    {
    "item": "https://format-3.co",
    "name": "Format-3",
    "@type": "ListItem",
    "position": 1
    },
    {
    "name": "SOC 2 for SaaS: What Buyers Want and How to Build for It",
    "@type": "ListItem",
    "position": 2
    }
    ]
    }
    ],
    "@context": "https://schema.org"
    }

    SOC 2 for SaaS: What Buyers Want and How to Build for It

    SOC 2 is not required by any US law, and yet a SaaS company without a report will find enterprise deals stalling in security review. The uncomfortable part is that most teams treat it as paperwork to be assembled just before an audit. In practice a SOC 2 report is a verdict on how your product was built and how your team runs it. If the product was designed without that in mind, you will pay for it twice: once in the audit, and again in the rework.

    This guide explains what SOC 2 is, what it demonstrates, how Type 1 and Type 2 differ, how long it takes, and why the decisions that matter most are made in product design and engineering, long before an auditor appears.

    Format-3

    Build security into your SaaS product

    Format-3 designs and develops digital products through end-to-end strategy, design and engineering, helping teams build with purpose from the start.

    Visit

    TL;DR

    • SOC 2 is a voluntary attestation, not a certification. Contracts and customer security reviews, not statutes, are what make it feel mandatory.
    • Security is the only criterion every report must include. Availability, Confidentiality, Processing Integrity and Privacy are added only when your commitments to customers justify them.
    • Type 1 tests control design at one date. Type 2 tests operation over a period and is the report most enterprise buyers ask for.
    • A first Type 2 is a matter of months, mostly because controls must run before they can be tested. Start the clock by building the controls into the product and the workflow.
    • Products designed as systems, with access, change and logging built in, produce audit evidence as a by-product of normal work.

    Table of Contents

    What SOC 2 actually is

    SOC stands for System and Organization Controls. A SOC 2 examination reports on controls at a service organisation that are relevant to security, availability, processing integrity, confidentiality or privacy, giving your customers detailed information and assurance about the systems that process their data. The framework belongs to the American Institute of CPAs (AICPA), the professional body for US certified public accountants.

    Two points are worth getting right, because the vocabulary around them is sloppy.

    First, SOC 2 is an attestation, not a certification. The report is an independent CPA firm’s opinion, and management must also provide a written assertion that the system description is fair and that the controls were suitably designed and, in a Type 2, operated effectively. There is no badge, no pass mark and no certificate from the AICPA. There is a report, and your customer’s security team reads it.

    Second, the report does not say you are unhackable. It says that a described set of controls existed, and, for a Type 2, that an auditor tested whether they worked. Deficiencies found during testing are expected to be reflected in the report, which is exactly why buyers trust it more than a questionnaire you filled in yourself.

    Is SOC 2 legally required for SaaS?

    No. No federal or state law names SOC 2 as mandatory, and the AICPA does not regulate who must hold one. Nobody is fined for lacking a report.

    The requirement is nonetheless real, and its source is the contract. A typical enterprise security review asks for a Type 2 report first, and answering that you do not have one commonly stalls the deal or pushes you into a longer questionnaire process. Regulated customers add their own pressure: US banks, for instance, may review third-party SOC reports under interagency guidance on third-party relationships, and the FTC Safeguards Rule obliges covered financial companies to monitor their service providers, although neither names SOC 2.

    That matters for healthcare, energy, media and igaming companies selling software into larger organisations. Your buyer’s procurement team has its own obligations to demonstrate vendor oversight, and your report is how they discharge them. A practical first move is to ask the requesting customer which report it needs, which categories it expects and whether Type 1 or Type 2 will satisfy it.

    The five Trust Services Criteria, and which ones you need

    The report is organised around five categories. Crowe’s SOC 2 FAQ sets them out like this:

    Category: Security | What it covers: Systems and information protected from unauthorised access. Required in every SOC 2 report. | Typical SaaS relevance: Always

    Category: Availability | What it covers: Systems available as needed to meet commitments | Typical SaaS relevance: Common, especially where you promise uptime in contracts

    Category: Confidentiality | What it covers: Confidential information protected through its life cycle | Typical SaaS relevance: Common where customers upload sensitive business data

    Category: Processing Integrity | What it covers: Processing is authorised, valid, complete, accurate and timely | Typical SaaS relevance: Relevant for payments, calculations, data pipelines

    Category: Privacy | What it covers: Personal information collected, used, retained and disposed of per your objectives | Typical SaaS relevance: Relevant where you handle consumers’ personal data

    Crowe notes that security, availability and confidentiality apply to most organisations and services, while processing integrity and privacy may not apply to every one. Its auditors also warn that the choice is not static: relevant risks shift over time, and each extra category should be weighed for its relevance, value and incremental effort.

    The sensible reasoning runs from commitments, not from ambition. Read your contracts and your public promises. If you guarantee uptime, Availability belongs in scope. If customers store proprietary files, Confidentiality does. If you process health or consumer data, Privacy enters the conversation, alongside whatever other privacy law applies to you. Choosing criteria according to the service, the data it handles and the risks around it is the same discipline that underpins more broadly: the controls follow what you have promised.

    A caution for the ambitious: adding every category to look thorough creates more controls to maintain and more evidence to produce. Add one because a buyer needs it or because it truly describes the service.

    Type 1 versus Type 2

    The distinction is simple to state and consequential in a sales cycle. A Type 1 report addresses the design of controls as of a point in time, while a Type 2 addresses their operating effectiveness over a period. Baker Tilly describes Type 1 reports as providing less comfort to the intended audience, and the minimum reporting period for a Type 2 as typically three months, with recurring reports typically twelve months long.

    The contents differ accordingly. A Type 1 holds management’s system description, management’s assertion on design and the auditor’s design opinion. A Type 2 adds an operating-effectiveness assertion and opinion, plus a description of the auditor’s tests and their results. That last element is what a security reviewer actually reads.

    The realistic path for a young company is often a Type 1 first, to unblock early conversations, followed by a Type 2 once controls have operated for a while. KPMG’s whitepaper on SOC 2 planning likewise suggests a readiness assessment and a Type 1 report for the first instance. Treat the Type 1 as a milestone, not a destination: the buyers with the biggest contracts will ask for the Type 2.

    How long it takes, and why the clock starts late

    Teams ask how hard SOC 2 is. The honest answer is that the difficulty is mostly duration. A Type 2 takes longer because controls must operate for a defined period before they can be tested. Deloitte’s 2026 readiness guide puts that period at commonly three to twelve months for SOC 2, with six months as a leading-practice target.

    The period cannot begin until controls are really running. That is where schedules slip: a company decides to pursue SOC 2, then discovers it has no access review, no formal change approval and no tested restore procedure. Each missing control has to be designed, built, adopted by the team and then observed. Add the readiness work beforehand and the auditor’s fieldwork afterwards, and a first Type 2 becomes a project measured in many months, not weeks.

    Costs depend heavily on scope, auditor and company size, and published figures vary widely, so get quotes from licensed CPA firms for your own scope rather than planning from a headline number.

    A preparation path that works for SaaS teams

    The readiness literature from the large audit firms converges on a similar sequence. Deloitte’s guide says to define scope, map data flows, interview control owners, find and remediate gaps, and draft the system description, including service commitments. Here is that path adapted for a product team, with the places it tends to go wrong.

    1. Decide the scope from your commitments. List the systems that touch customer data: production infrastructure, the customer-facing application and APIs, identity management, databases, logging and the deployment pipeline. Marketing sites and isolated development environments generally sit outside. It goes wrong when scope is drawn from the org chart instead of the data flow, so a forgotten integration surfaces during fieldwork.
    2. Map the data. Trace where customer data enters, where it lives, who can reach it and where it leaves, including third-party services. This is a product question as much as a compliance one, and it is far cheaper to answer on a diagram than in a codebase.
    3. Run a readiness assessment. Evaluate whether the controls you have are suitably designed, list the gaps and prioritise remediation. It goes wrong when the gap list is treated as a to-do for the security lead alone. Most gaps are product and engineering work: role-based access, audit trails, environment separation, backup and recovery.
    4. Build and adopt the controls. Change management, access reviews, incident response, risk assessment and training must exist as live routines, not as documents. A policy nobody follows becomes a finding.
    5. Observe and collect evidence. In a Type 2, the auditor samples across the whole period: access approvals, change tickets, provisioning records, backup confirmations. Evidence that the system generates itself is far easier to produce than screenshots assembled in a panic.
    6. Undergo fieldwork and receive the report. An independent CPA firm tests the controls. This step is reserved to licensed auditors by the attestation standards, so it is part of the process for every company, whoever supports the preparation.

    The part most guides skip: SOC 2 is a design problem

    Most advice treats SOC 2 as a documentation exercise bolted onto a finished product. We think that framing is the reason so many first audits hurt. Consider what an auditor wants to see: that access to customer data is restricted and reviewed, that changes to production are approved and traceable, that you can detect and respond to incidents, and that you can recover. Each of those is a property of the system.

    A product built with roles and permissions as a first-class concept can show who can see what. A delivery pipeline that records approvals and deployments leaves a change trail without anyone filling in a spreadsheet. Logging designed into the architecture answers the monitoring question before it is asked. Retrofit any of these onto a mature product and you are re-engineering core behaviour under an audit deadline, with customers waiting.

    There is also a recent signal that the direction of travel favours systems. One 2026 guide reports that late-2025 AICPA guidance asks auditors to assess the reliability of system-generated evidence independently, giving automated logs and audit trails more weight than screenshots. We would treat that as a trend to watch rather than settled doctrine, but it fits the logic: evidence a system produces on its own is more convincing than evidence a person curates.

    The same applies to AI features. If your roadmap includes AI implementations, the data those features consume and the third parties they call belong in the data map from day one, not as a surprise during scope review.

    Perspective

    Most founders ask when they should “get SOC 2”. The better question is what their product would have to be like for SOC 2 to be boring.

    Where Format-3 fits

    If you are planning a SaaS product, or rebuilding one that must now satisfy enterprise buyers, we can start with Digital Product Discovery and Design Sprints to settle the data flows, roles and architecture that your controls will rest on, then carry them through full-stack development with a team embedded alongside yours. We have worked across healthcare, energy, igaming, SaaS and entertainment, where security reviews are routine. For a sense of how we approach product work, see our case studies for TripleKey and Bentkey, or read why products don’t scale, systems do.

    What changed recently, and what stays stable

    The core framework is steadier than the commentary suggests. The AICPA’s authoritative SOC 2 guide, updated as of October 2022, applies the 2017 trust services criteria with revised 2022 points of focus, which are implementation guidance rather than new requirements. Later changes to the underlying attestation standards have clarified matters such as materiality and fraud considerations, but the substance of what is tested has not changed. Plan around the five categories and the Type 1 and Type 2 structure, and confirm with your auditor which guidance applies to your reporting period.

    FAQ

    Is SOC 2 for SaaS companies only?

    No. SOC 2 applies to any service organisation that holds or processes customer data, including data centres and managed service providers. SaaS companies are its most frequent users because their customers hand them sensitive data and need independent assurance about how it is protected.

    Is SOC 2 legally required in the United States?

    No statute or regulator requires it, and no fines attach to lacking a report. It becomes a practical requirement when customers write it into procurement and vendor reviews, which is common in enterprise B2B sales.

    Do we need Type 1 or Type 2?

    Ask the customer who is requesting it. A Type 1 shows controls were designed properly at a date and can open early conversations, while a Type 2 shows they operated over a period and is the report larger buyers usually expect.

    Which Trust Services Criteria should a SaaS company include?

    Security is mandatory in every report. Add Availability, Confidentiality, Processing Integrity or Privacy according to the commitments you make to customers and the data you handle, since each addition brings more controls and evidence to maintain.

    Sources

    Recommended

    Format-3 profile picture
    By Format-3
    Press & Media
    Share article

      More thoughts

      Thought leadership creates value, builds knowledge and takes a stand, bridging the gap between traditional and digital platforms

      Need help with your project, career or want to interview us?

      SayHello!

      • 14:47:39
        Nashville
        USA
      • 15:47:39
        New York
        USA
      • 20:47:39
        London
        UK
      • 21:47:39
        Katowice
        Poland
      • 21:47:39
        Bratislava
        Slovakia
      • 22:47:39
        Plovdiv
        Bulgaria
      • 00:47:39
        Dubai
        UAE
      0
      %
      F
      o
      r
      m
      a
      t
      –
      3