Arctiq Main Blog

Why Identity Has to Come Before Zero Trust | Arctiq

Written by Eddie Skey | Nov 19, 2026, 3:11:49 PM

Key Takeaways

  • Identity is a practice, not a project. It means defined ownership, documented processes, identity data you can trust, working governance, and metrics someone answers for.
  • Zero Trust, passwordless, and SSO multiply whatever is already there, including the mess. Without named app owners, reliable lifecycle processes, and accurate identity attributes, they make wrong decisions faster.
  • Build the roadmap as a progression, not a shopping list: single source of truth, lifecycle and ownership, access and governance, then advanced capabilities. Protections with immediate risk, like phishing-resistant MFA for privileged accounts, run alongside all four steps.
  • Measure the practice, not the deployment. Track how long leavers keep access, how many apps have a named owner, and whether access reviews end in real remediation.


Nobody gets a standing ovation for fixing joiner, mover, and leaver processes. Nobody throws a party for a clean application inventory or a named owner on every app.

Announce a Zero Trust program, an enterprise single sign-on (SSO) rollout, or a passwordless initiative and the room lights up. Those projects get sponsors and budget.

I'm not against any of them… I'm against doing them first.

I keep seeing the same thing. An organization wants the outcomes of a mature identity program without the practice behind it. Risk based access decisions are the goal, but nobody can say who has access to what today. Passwordless is on the roadmap while the HR system and Active Directory still disagree about who works there. Automation gets funded even though the approval process lives in one person's head.

The project gets expensive and slow, and everyone blames the technology. Usually, the technology is fine. The order is the problem.

Identity is a practice, not a project

A project has an end date. A practice is an ongoing responsibility.

You can deploy an identity governance and administration (IGA) platform, roll out multi-factor authentication (MFA), and move to a new identity provider, and each one can be called a success on its own terms. None of that tells you whether you have a mature identity practice.

A practice is defined ownership, documented processes, identity data you can trust, working governance, and metrics somebody answers for. HR, IT, security, and the application owners all have a part in it. Technology supports the practice. It doesn't replace it.

What Zero Trust, SSO, and passwordless depend on

Take enterprise SSO. It centralizes authentication, and that's all it does. It doesn't decide who should have access. Onboard an application with no owner and nobody is deciding who gets a key, so over-provisioned users now reach more apps with less friction.

Passwordless has a similar problem. A strong authenticator is only as trustworthy as the enrollment behind it. Who verifies identity at enrollment, and what happens at account recovery? If your leaver process is inconsistent, you end up with a stale account holding a strong credential. Recovery becomes the new password, and the help desk becomes the attack surface.

Zero Trust makes every access decision on attributes like role, department, manager, device, and application criticality. When those attributes are wrong, the policy engine makes wrong decisions, just faster and with more confidence. It doesn't remove the foundational problems. It exposes them.

These capabilities multiply whatever is already there, including the mess.

What you risk if you skip the foundation

The stakes are not abstract. In Sophos's State of Identity Security 2026 research, 71% of organizations reported at least one identity-related breach in the past 12 months, and the average cost to rectify a successful one was $1.64 million.

In my 30+ years of experience, a significant number of projects built around advanced identity capabilities have stalled. Some stall early, the first time the team runs into bad data or an application nobody owns. Others make it midway, after the budget is committed and the pressure to show progress is highest. The cause is usually the same. The substructure a mature identity practice provides was never built to support the advanced features sitting on top of it.

The pattern is predictable. Projects run long because the team ends up solving data, ownership, and process problems in the middle of a deployment. Integration costs climb because each system was built on its own. Operations inherits manual workarounds. Security can't prove controls are working because the identity data underneath can't be trusted, and the business sees a big investment with little to show for it.

Here's what that looks like in practice:

A client I worked with had done the hard planning work. They had architected and designed the approach, built a reference architecture, and put together implementation plans. They were about 98% of the way to purchasing a Zero Trust technology.

Then it became clear that their identity program was nowhere near mature enough to make the initiative succeed. The biggest gap was the most basic one. They had no reliable source of truth for identities. Zero Trust makes decisions based on who someone is and what role they hold, and without an authoritative source, they had nothing trustworthy to base those decisions on. It wasn't a close call.

The conclusion changed the plan, and the next job was building the identity practice the initiative depends on.

Look at when they found out. It was before the purchase, not midway through a deployment with the money already spent. Plenty of organizations learn the same lesson later and at a much higher price. The design work wasn't wasted either, because it gave them a clear picture of what the foundation has to support.

The tool works as designed. The program around it has to be ready. Chase the shiny new object before the house is in order, and you end up paying to fix the same foundation twice.

"We can't wait for perfect data before we ship MFA"

Fair, and I'm not arguing for a two-year pause. Some work should run in parallel where the risk is immediate. Phishing-resistant MFA for privileged accounts is the obvious one. Do that now.

What matters is knowing which pieces are true prerequisites. Reliable identity data, lifecycle processes, and application ownership come first. They don't have to be perfect. They have to be dependable enough that the next layer isn't built on guesses.

A roadmap should be a progression, not a shopping list

Don't build the roadmap as a ranked list of products. Build it as a progression of capabilities.

  1. Get authoritative identity data in place, so lifecycle management has something solid to run on.
  2. Build consistent lifecycle processes and application ownership, so you can govern access.
  3. Define access models and approval responsibilities, so automation has something to automate.
  4. Then add advanced authentication, adaptive access, and Zero Trust.

Here is what each step means in practice:

Step 1: Single source of truth

Start with consolidated directory services and an authoritative record of who works for you, typically HR for employees and a governed record for contractors and third parties. Everything after this assumes the identity record is right. Done looks like every identity tracing back to one record you trust.

Step 2: Lifecycle and ownership

Build joiner, mover, and leaver processes on that source, and give every important application a named owner. Without owners, nobody can approve access. Without lifecycle processes, access just piles up. Done looks like access that changes when roles change and disappears when people leave, with an owner who answers for each application.

Step 3: Access and governance

With reliable data and clear owners, define access models, approval responsibilities, and review cycles. Then automate what is now well defined. Automating earlier only speeds up the inconsistency. Done looks like requests, approvals, and reviews that end in real decisions and remediation.

Step 4: Advanced capabilities

Now Zero Trust, passwordless, and SSO have what they need: attributes you can trust, owners who approve access, and lifecycle controls that hold. Done looks like rollouts that stick because the foundation underneath them does.

Here is the full sequence at a glance. Run protections with immediate risk, like MFA for privileged accounts, alongside all four steps without waiting.

What good looks like

When the foundation is in place, identities live in consolidated directory services backed by a single source of truth. Every important application has a named owner. Access is removed promptly when someone leaves, and role changes update access instead of letting it pile up. When the Zero Trust policies go live, they run on identity data the team trusts, and the rollout holds.

Measure the practice, not the deployment

Activity is not maturity.

Stop reporting how many applications you've integrated. A long list of completed projects doesn't tell you whether the practice is getting more reliable.

Ask these instead:

  • How long does a leaver keep access after termination?
  • What percentage of applications have a named owner and a documented approval process?
  • When someone changes roles, does their access change, or does it just pile up?
  • Are access reviews ending in real remediation, or in rubber stamps?
  • How many privileged and non-human identities are inventoried and governed? Service accounts, API keys, bots, that kind of thing.

Where to start

Start with three questions:

  1. Can you list every employee, contractor, third party, and service account in your environment?
  2. Do you know what each one can reach, and why?
  3. Who approves access to your top 20 applications?

If the answers take more than a day or two to pull together, that's your starting point.

None of this is glamorous. Leadership has to fund it anyway, because everything exciting depends on it.

Think of it like football. Blocking and tackling aren't glamorous, but they decide the game. Until your line holds and your tackles get made, the advanced defensive formations and the well-designed pass and run sequences won't save you. In identity, the blocking and tackling is your single source of truth, lifecycle processes, application ownership, and governance. The advanced plays are Zero Trust, passwordless, and adaptive access. Run them in that order.

Doing the right things matters. Doing them in the right order is what makes them work.

What the identity workshop delivers

Our identity workshop runs 8 to 10 weeks. You leave with four things.

  • A visual of your current maturity across the identity program, so you can see which areas need the most improvement.
  • The order of operations: what to address first, and in what sequence.
  • A reference architecture that shows what the target state should look like.
  • A phase plan across the next 12, 24, and 36 months to get you there.

If you're planning a Zero Trust, passwordless, or SSO initiative and you're not sure the foundation is ready, that's the conversation to have first. Connect with us and we'll talk through where you are.