Find and fix what could reach your client

Continuous threat exposure management (CTEM) is a repeating cycle of five stages: scope, discover, prioritize, validate and mobilize. Four modules run all five for every client you add. Discover maps your client before anything is touched. Remediate ranks every finding by blast radius, the damage it could do, confirms what’s real and sends the fix to your PSA. Validate adds penetration tests by a trusted network of third-party, AI-assisted, human-led testers. Suppliers finds and watches the suppliers behind every client.

The chart shows which module covers each stage. Active checks start only after your client approves.

Harbor Lending, from first look to fixed

Harbor Lending is a fictional client. Choose a stage to jump to it, or pause at any point.

An illustration with a fictional client, Harbor Lending, through the five stages of continuous threat exposure management. Scope: before you add the client, the Research engine has already mapped it from public sources. Discover: when you add its domain, the Research engine shows what Harbor knew about and what it didn’t: an unlisted staging site, four unlisted suppliers, two lookalike domains and three threats in play, before any scan. Prioritize: Outerwall ranks everything by blast radius, with the known-exploited flaw on the VPN gateway first. Validate: after Harbor approves, a non-destructive check confirms that flaw, and Outerwall simulates how far an attacker could get from it: sign-in, file shares and the payroll system. Mobilize: the findings go to your PSA as tickets ranked by blast radius, and one supplier breach is traced across 14 of your 40 clients.

Ranked by blast radius: how much damage a failure there could do to your client.

Module 1 of 4

Discover

Before you add a client, the fabric already holds its internet-facing systems, its suppliers and the rules it answers to. Discover turns that into a picture your client can confirm, before any scan touches it.

Covers
Scope and Discover
Needs
One domain. Nothing is scanned until your client approves
Lands in
Your partner portal and your client’s report

CTEM stage 1 · Scope

We map each company before you add it.

The Research engine maps each company from public sources and keeps it current: its internet-facing systems, its suppliers and the rules it answers to. Most of a new client’s scope is ready within the hour of adding it, and your client confirms what’s theirs. Every fact comes with where and when we found it. Anything we can’t confirm is labeled likely, or thrown out.

How the Research engine works for one fictional client, Harbor Lending. We read five public surfaces: infrastructure, records, events, workforce and vendor references. Each fact we already hold is labeled firm, likely, thrown out or can’t see, with its source and date. Firm and likely facts then feed what Harbor knew about, what it didn’t know, the rules it answers to, and what we can’t see.

1Where we look

InfrastructureCertificate logs, DNS, internet scan dataRecordsRegisters, certificationsEventsBreach notices, known-exploited flawsWorkforceJob ads, staff emails in breachesVendor referencesCase studies, procurement notices

2Already mapped · harborlending.example

Aug 12, 2026staging.harborlending.example, a staging siteUnlistedCertificate logFirm
LiveThe VPN gateway, with its make and versionInternet scan dataFirm
LiveLicensed mortgage broker, so the FTC Safeguards Rule appliesState license registerFirm
Sep 3, 2026A payroll provider and an endpoint agent, named in a job adUnlistedJob adLikely
2025A loan platform that names Harbor as a customerUnlistedVendor case studyLikely
9 days agoA flaw in that VPN make added to CISA’s known-exploited listUnlistedCISA KEVLikely
2019An FTP serverOld DNS record, no longer resolvesThrown out
Looked forThe internal payroll systemKnown to exist, unseen from outsideCan’t see

3What happens next

Known

What Harbor knew about: its domain, hosts and VPN, kept current.

Unlisted

What it didn’t know: a staging site, unlisted suppliers and a known-exploited flaw.

Rules

The FTC Safeguards Rule applies, so these become evidence Harbor will need.

Coverage

What we can’t see is listed first, never hidden.

FirmSeen directly in a source we trust, with its date.

LikelyInferred, and labeled likely until a check or your client confirms it.

Thrown outToo old, or contradicted by a newer source.

Can’t seeKnown to exist, unseen from outside. Listed, never hidden.

CTEM stage 2 · Discover

You see what each client knew about, and what it didn’t.

Within the hour of adding a domain, before any scan, the Research engine shows you the hosts, suppliers and lookalike domains each client knew about next to the ones it didn’t, and any threats already in play.

Harbor Lending · within the hour · public sources only
WhatThey knewOuterwall found
Domains13
Host names1238
Suppliers26
Lookalike domains02
Staff emails in breaches011
Threats in play03

0 active checks sent to Harbor so far.

Flagged for you

  • Unlisted staging siteCertificate log · Unlisted
  • A flaw in its VPN is on CISA’s known-exploited listCISA KEV and internet scan data · Likely
  • Its payroll provider filed a breach noticeBreach notice · Unlisted supplier
  • A lookalike domain that can receive mailCertificate log and DNS · Likely

Module 2 of 4

Remediate

Remediate puts everything in order: ranked by what an attacker could reach, confirmed on approved assets, and sent to your PSA as tickets that close only when the fix is proven.

Covers
Prioritize, automated validation and Mobilize, plus the read-only inside view
Needs
Proof of control (a DNS record or a file), a signed scope and 24 hours’ notice
Lands in
ConnectWise, Autotask, HaloPSA, ServiceNow or Jira, as ranked tickets

CTEM stage 3 · Prioritize

Everything is ranked by the damage it could do.

Outerwall ranks every finding by blast radius: how much damage it could do to your client if an attacker used it. A threat already in play moves a finding up. Severity alone never sets the order.

  • Blast radius

    How much damage a finding could do to your client: from one host to the payroll system, or from one supplier to every client that shares it.

  • Live threats

    A flaw on CISA’s known-exploited list, a supplier’s breach notice or a ransomware mention moves a finding up the list the day it appears.

  • How sure we are

    Every finding says whether it’s confirmed or likely, and coverage shows what we couldn’t see. Likely findings at the top get a confirming check first.

Every rule is tested code, so the same data always gives the same answer. No AI model decides a finding.

CTEM stage 4 · Validate, automated

Then we confirm what’s real, and how far it would reach.

A likely finding stays labeled likely until a check proves it. On approved assets, non-destructive checks confirm a finding or drop it. Then we trace the path an attacker could take from it, and what that path would reach.

An illustration of a traced attack path at a fictional client: a confirmed known-exploited flaw on the VPN gateway gives an attacker a foothold on the internal network, from there staff sign-in, and from there the payroll system with staff pay and bank details. Blast radius: it could reach payroll.

Traced attack path · illustration

  • Confirm it’s real

    Non-destructive checks on approved assets turn a likely finding into a confirmed one, or drop it. A version number alone is never enough.

  • Trace the path

    Findings are joined into the paths an attacker would take, including paths that start at a supplier, and ranked by what they would reach.

  • Measure it from inside

    With approval, Outerwall reads your client’s Microsoft 365, Entra and Defender. A flaw is checked against the patch state Defender reports and the accounts Entra shows it could reach. Blast radius becomes a measurement: which admins, which devices, which shares.

  • Then a person, if your client needs it

    Penetration tests from the Validate module show how far an attacker could really get. Findings land here as they’re confirmed.

CTEM stage 5 · Mobilize

Ranked tickets land in your PSA.

Each finding is escalated as a ticket in ConnectWise, Autotask, HaloPSA, ServiceNow or Jira, ranked by blast radius: how much damage it could do to your client. Before a ticket closes, we check the fix twice.

Your PSA · Harbor LendingRanked by blast radius
  1. 1Exploited flaw on the VPN gatewayConfirmed · could reach the payroll systemTicket #4821
  2. 2Payroll provider breachUnlisted supplier · severeTicket
  3. 3Invoice fraud from a lookalike domainLikely · the lookalike can receive mailTicket
  4. 4Unlisted staging siteThey didn’t know it existed · owner neededTicket

Each ticket carries the evidence and a plain-language note for your client.

After onboarding, Outerwall keeps watching

Fast-moving changes, in under 6 hours

A supplier’s breach, outage or bankruptcy. A new flaw on CISA’s known-exploited list that matches software a client runs. Staff credentials in a breach or infostealer log. A lookalike domain. A ransomware group naming a client.

Daily changes, within 24 hours

New subdomains and addresses, DNS and email records. A change triggers a recheck of only the asset that changed.

Module 3 of 4

Validate

Validate brings penetration testing into the same workflow, ordered and tracked inside Outerwall, by your own testers or a trusted network of third-party, AI-assisted, human-led testers. Their findings arrive with everything else.

Covers
Penetration testing, the tester-led part of the Validate stage
Needs
A scope agreed in writing
Lands in
Outerwall and your service desk, as each finding is confirmed

CTEM stage 4 · Validate, with testers

Penetration tests by a trusted network of testers

A tester shows how far an attacker could really get, and fixes are retested free. The findings sit alongside everything else Outerwall has found for your client.

Every engagement follows one methodology aligned to CREST’s, drawing on PTES, NIST SP 800-115, OWASP, OSSTMM and MITRE ATT&CK, whichever tester delivers it. AI handles the repetitive analysis at scale. A certified tester leads every engagement, decides what to try, carries out the exploitation and signs off each finding.

Where your client needs it, an engagement can follow CREST’s Defensible Penetration Test specification, with a written statement of scope and coverage, a timeline of the testing and a formal sign-off by a named, certified tester.

You can also use your own penetration testers. Their findings land in Outerwall the same way, ranked and ticketed with everything else.

One methodologyCREST-aligned

  1. Scopeagreed in writing, nothing outside it touched
  2. Reconnaissancemapped by tools, every result checked by a tester
  3. Exploitationhow far an attacker could really get; criticals reported at once
  4. ReportingCVSS 4.0, grouped by root cause, with the fix
  5. Clean-upevery account, tool and file removed
  6. Retestfixes retested free

Testers hold certifications such as OSCP, OSEP and CREST CRT.

  • Scoped from the fabric

    The map is the starting point, so the scope covers the whole attack surface, including the parts your client forgot. Nothing outside the signed scope is touched.

  • Findings as they’re confirmed

    Each finding is scored with CVSS 4.0, grouped by root cause and sent to your service desk with everything else, instead of waiting for a PDF.

  • Safe by rule

    A critical finding is reported the moment it’s found, and exploiting one needs your client’s approval first. Nothing that could take a service down is attempted.

Module 4 of 4

Suppliers

Suppliers runs across every stage. Because the fabric holds your whole book, it sees which suppliers your clients share, and a breach at one reaches every client it touches.

Covers
Suppliers, across all five stages
Needs
Nothing to start. Your client confirms the list
Lands in
One view across your book, and a ticket per client

Find the suppliers your clients didn’t list

Ask a client for its supplier list and it will forget the ones that matter most: the identity provider, and the security agent that updates itself on every laptop. Suppliers finds them from DNS records, job ads and vendor case studies, ranks them by the damage a failure would do, and watches them.

Because the fabric holds every client you protect, it sees which suppliers they share. When one is breached, every client that depends on it is flagged in one view within hours, with a ticket for each, so you raise it with all of them the same morning.

The suppliers it finds become the proposed supplier register in Predictive GRC, for your client to confirm.

Suppliers in depth

Harbor Lending · supplier register6 found, 2 listed

  • Identity and mailCatastrophicClient to confirm
  • Endpoint agentCatastrophicClient to confirm
  • Payroll providerSevereBreach notice today
  • Loan platformSevereClient to confirm
  • Cloud hostHighReviewed
  • E-signature and CRMModerateReviewed

Found from the traces each supplier leaves, including the four nobody listed.

Where this fits in CTEM

Gartner set out CTEM in 2022 as a five-stage program to find what matters, rank it, prove what’s real and get it fixed. Outerwall runs all five stages for every client you add, from one view, and the chart at the top of this page shows which module covers each one.

What is CTEM?

The same evidence runs your client’s program

Everything exposure management finds, confirms and fixes becomes evidence. Predictive GRC uses it to draft your client’s supplier register, risk register, controls and evidence packs, keeps them current, and proposes a risk when something in the fabric touches your client.

It works the other way too. What your client’s risk register says matters to the business sets what Discover and Remediate look at first.

GRC: Predictive GRC and Trust Center

See it on one client.

A short read: what Outerwall finds on a fictional mortgage broker, step by step.