An AWS landing zone setup gives your organization a governed, multi-account foundation before workloads multiply: separate accounts for production, staging and security, central logging, single sign-on, preventive guardrails and spend visibility per team. Done early, it spares you a difficult untangling exercise later. This guide explains what a good landing zone contains and how to roll one out.
If you already know you need one and want a second opinion on your current account structure, share your AWS setup with our cloud team and we will map out the first milestone with you.
Why a single AWS account stops working
Most companies start with a single AWS account. Developers, production databases, test environments and experiments all share the same IAM boundary and the same bill. That is fine for a prototype, but a liability once you have paying customers, enterprise security questionnaires or a growing team.
The symptoms are familiar: unclear ownership, test scripts that can touch production, CloudTrail logs in the account an attacker would compromise, and one bill nobody can break down. Accounts are the strongest isolation boundary AWS offers, so the fix is structural.
What an AWS landing zone setup actually includes
A landing zone is not a single product. It is a set of decisions, implemented as code, about how accounts are created, secured and connected.
Account and OU structure
AWS Organizations groups accounts into organizational units (OUs). A common pattern is a Security OU (log archive and audit accounts), an Infrastructure OU (shared networking, CI/CD tooling), a Workloads OU split into production and non-production, and a Sandbox OU for experiments with tighter limits. Policies attach to OUs, so every new account inherits the right controls automatically.
Identity and access
IAM Identity Center provides single sign-on into every account with permission sets mapped to job roles. Long-lived IAM user keys disappear and offboarding happens in one place.
Guardrails and controls
Service control policies (SCPs) set hard limits, such as denying disabled regions, blocking the deletion of CloudTrail or preventing public S3 buckets in production. AWS Control Tower adds a library of preventive, detective and proactive controls you can enable per OU.
Preventive versus detective
Preventive controls stop an action before it happens; detective controls, usually built on AWS Config rules, flag drift after the fact. A mature landing zone uses both: preventive for non-negotiables, detective for good practice you want visibility on without blocking delivery.
Central logging and security tooling
An organization-wide CloudTrail trail, AWS Config and VPC flow logs are delivered to a dedicated log archive account that workload teams cannot modify. GuardDuty and Security Hub are enabled across all accounts with a delegated administrator in the audit account, which gives your security reviewer one place to look. For ongoing posture monitoring after go-live, see our guide to cloud security posture management.
Networking
Shared networking usually lives in its own account, with a Transit Gateway hub, an IP plan that will not collide with office or partner networks, and central egress where compliance requires it.
Control Tower, Terraform or both?
AWS Control Tower is the managed way to stand up a landing zone. Launched from your organization's management account, it creates the shared log archive and audit accounts, applies a baseline of controls, and offers Account Factory for vending new accounts. For most growing companies it is the right starting point because AWS maintains the underlying orchestration.
Teams that already manage infrastructure in Terraform often add Account Factory for Terraform (AFT), which lets new accounts be requested through a pull request and customized with Terraform modules. Others use Customizations for Control Tower to deploy CloudFormation stacks into new accounts. A fully custom landing zone without Control Tower shifts long-term maintenance onto your team, so reserve it for unusual regulatory or network needs. If your infrastructure code needs attention first, our Terraform consulting guide covers how we structure modules and state.
Cost awareness built into the foundation
A landing zone is also the easiest moment to get financial visibility right, because structure decides how clearly you can see spend later. We do not quote figures here, since every environment differs, but the design levers are consistent:
- One account per workload and environment, so spend is attributed without relying on perfect tagging.
- Tag policies enforced at the OU level for owner, cost center and environment.
- AWS Budgets and anomaly detection per account, with alerts routed to the team that owns it.
- Sandbox limits through SCPs that restrict large or GPU instance families and unused regions.
- Consolidated billing so savings plans and reserved capacity are shared across the organization.
Engineers see what their account consumes, and finance gets a breakdown by product.
How we deliver a landing zone, milestone by milestone
At iJurug Soft, cloud management is one of our core engineering services, and landing zone work follows the same Discover, Design, Build, Launch and grow process we use on every engagement, with senior engineers doing the work and fixed milestones you can track.
- Discover: inventory existing accounts, IAM users, workloads, networking and compliance obligations such as SOC 2, ISO 27001 or the DPDP Act.
- Design: agree the OU tree, account naming, guardrail set, IP plan and identity model. This becomes a short written design you sign off.
- Build: deploy Control Tower or the agreed equivalent as code, configure Identity Center, central logging and security services, and set up account vending. Existing workloads then move into the new structure in waves, starting with the least risky, with rollback steps written down for each wave.
- Launch and grow: hand over runbooks, train your team on requesting accounts and reading findings, and stay available for ongoing support.
Before migrating production, it is worth running an AWS Well-Architected review so reliability and security gaps are fixed during the move rather than carried across.
Questions to ask any landing zone provider
- Will everything be delivered as code in our repository, or configured by hand in the console?
- How will existing accounts be enrolled without downtime for running workloads?
- Which controls are preventive, which are detective, and why?
- What documentation and training do we receive at handover?
Frequently asked questions
How long does an AWS landing zone setup take?
It depends on how many existing accounts and workloads need to move, your compliance requirements and how much networking is shared. A greenfield setup is much faster than enrolling a mature estate. We confirm a timeline after the discovery milestone.
Can we set up a landing zone if we already run production on AWS?
Yes. Existing accounts can be invited into the organization and enrolled into Control Tower, or workloads can be moved into newly vended accounts in planned waves. The approach depends on how tangled the current account is.
Does a landing zone lock us into Control Tower?
No. Control Tower builds on AWS Organizations, SCPs, Config and CloudTrail, which remain usable on their own. Keeping customizations in Terraform or CloudFormation keeps your options open.
How much does a landing zone project cost?
We do not publish prices because scope varies widely. The main drivers are the number of accounts and workloads, compliance needs, network complexity and whether migration is included. Share your details and we will give you a clear quote.
Ready to put a proper foundation under your AWS estate? Tell us about your accounts, workloads and compliance goals through our project form, or write to info@ijurugsoft.com. A senior iJurug Soft cloud engineer will review what you share and come back with a recommended first milestone.