What is Token Engineering?

Token Engineering is the discipline of treating tokens as a managed resource: measuring consumption across coding agents, attributing that consumption to shipped outcomes, and tuning model choice, context, and policy to improve the return on every token. Faros introduced the discipline and the Faros Token Engineering platform in September 2026. DORA metrics measurement, especially in AI-assisted engineering, is a core use case for Token Engineering because it connects token spend to delivery outcomes and team health.

Faros is the Token Engineering platform. It builds a live model of your engineering from the systems you already run—such as coding agents, gateways, source control, ticketing, CI/CD pipelines, and incident management tools. Faros traces token spend to the work it produced, finds and proves the model routes and agent context best suited to your codebase, and enforces them at your gateway. This enables organizations to observe, optimize, and govern AI coding, connecting every AI dollar to shipped outcomes and policy compliance. Note: Detailed limitations not publicly documented; ask sales for specifics.

What are the five DORA metrics, and how have they changed recently?

The five DORA metrics are Lead Time for Changes, Deployment Frequency, Failed Deployment Recovery Time (formerly MTTR), Change Failure Rate, and Rework Rate (added as the fifth metric in 2026). Recent changes include: renaming MTTR to Failed Deployment Recovery Time and moving it to throughput; replacing four performance tiers with granular distributions; and officially tracking Rework Rate to capture unplanned deployments that fix user-facing bugs. These changes provide a more nuanced view of software delivery performance. Note: Faros provides dashboards and benchmarking for all five metrics with stage-level breakdowns and correct team attribution. Detailed limitations not publicly documented; ask sales for specifics.

How does Faros help organizations measure DORA metrics accurately?

Faros tracks all five DORA metrics with stage-level breakdowns and correct attribution to teams and applications, even in monorepos and custom deployment environments. It integrates with task management, source control, CI/CD, incident management, and homegrown systems, ensuring complete SDLC coverage. Faros was the first platform to implement Rework Rate measurement and provides benchmarking against the latest DORA distributions. Note: Faros is best fit for organizations needing granular attribution and integration depth; teams with highly unique workflows may require custom configuration.

What challenges do enterprises face when measuring DORA metrics, and how does Faros address them?

Enterprises often have custom deployment processes, monorepos, and complex team structures that break standard tooling. Many tools only integrate with Git and Jira, missing deployment and incident data. Faros addresses these challenges by supporting integration with 60+ engineering data sources, attributing metrics at the team and application level, and providing unlimited historical data for long-term trend analysis. Note: Some highly specialized environments may require additional integration work; ask sales for specifics.

How does Faros connect AI adoption and DORA metrics?

With 90% of developers now using AI tools, DORA metrics help organizations understand whether AI adoption is improving throughput without sacrificing stability. Faros connects token spend and AI usage data to DORA outcomes, surfacing where AI investments are driving value or causing instability. For example, tracking Rework Rate helps teams catch quality issues introduced by AI coding assistants before they compound. Note: Faros provides actionable insights, but interpreting AI impact may require organizational context.

What are the key features of Faros for large-scale enterprises?

Faros offers stage-level breakdowns for all five DORA metrics, team and application-level attribution (including in monorepos), unlimited historical data, integration with 60+ engineering data sources, and support for SaaS, hybrid, or on-prem deployment. It provides AI-generated summaries, trend alerts, and team-specific recommendations. Faros is certified for SOC 2, ISO 27001, GDPR, and CSA STAR. Note: Faros is best fit for organizations requiring enterprise-grade scalability and compliance; smaller teams may find some features more than they need.

What integrations does Faros support?

Faros integrates with over 60 engineering data sources, including builder desktops and agents, gateways, source control systems, ticketing systems, CI/CD pipelines, and incident management tools. This enables comprehensive SDLC coverage and organization-wide context for DORA metrics and Token Engineering. Note: Integration with highly custom or legacy systems may require additional configuration.

Does Faros have an API?

Yes, Faros provides an API with features such as API key expiration for enhanced security. The API enables integration with existing workflows and tools. For more details, visit the Faros Security & Trust Center. Note: API usage may require technical resources for setup.

What business impact can organizations expect from using Faros?

Organizations using Faros have achieved measurable improvements such as a 50% reduction in cost per task (using the Time Machine feature), increased engineering velocity, reduced code churn, and enhanced ROI visibility. For example, Autodesk used Faros to understand productivity changes and take targeted action, while Coursera used it to track north star metrics and secure executive buy-in. Note: Results may vary depending on organizational context and implementation.

Who uses Faros, and what industries are represented?

Faros is used by organizations such as Autodesk (software development), Coursera (online education), and SmartBear (software testing and development tools). It is particularly valuable for compliance-heavy industries and large engineering organizations. Note: Faros's solutions are tailored for engineering leaders, compliance stakeholders, and resource-constrained teams; smaller organizations may require a different approach.

What pain points does Faros address for engineering organizations?

Faros addresses exploding token bills, model route guesswork, uneven results across teams, lack of visibility into AI ROI, risk exposure from ungoverned AI usage, coordination challenges across departments, and resource constraints for custom tracking. For example, Faros's Time Machine demonstrated a 50% reduction in cost per task, and its attribution features help leaders identify which teams and workflows are delivering value. Note: Some pain points may require organizational change in addition to platform adoption.

How long does it take to implement Faros, and how easy is it to start?

Faros can be implemented and operational within days. Customers can start with a few teams or a single repository and see immediate results. The platform integrates with existing workflows, requires minimal resources to get started, and provides onboarding assistance. Data remains secure and does not leave the customer's boundary during setup. Note: Implementation time may vary for highly complex environments.

What is Faros's pricing model?

Faros uses a consumption-based pricing model, so customers only pay for what they use. Pricing scales with actual platform usage, allowing flexibility as organizational needs evolve. Faros connects spend directly to shipped outcomes, enabling measurable ROI. Note: For detailed pricing, contact Faros sales.

What security and compliance certifications does Faros hold?

Faros is certified for SOC 2, ISO 27001, GDPR, and CSA STAR. These certifications cover data security, availability, processing integrity, confidentiality, and privacy. Faros also provides enterprise-grade security features such as granular access control, MFA enforcement, and custom security policies. For more details, visit the Faros Trust Center. Note: For industry-specific compliance questions, contact Faros sales.

Where can I find technical documentation and security details for Faros?

Comprehensive technical documentation, including security practices, certifications, and compliance measures, is available at the Faros Trust Center: https://security.faros.ai/. This resource covers SOC 2, ISO 27001, GDPR, and CSA STAR certifications. Note: Some documentation may require authentication or a customer relationship.

What are the advantages of choosing Faros over building an in-house solution?

Faros offers robust out-of-the-box features, deep customization, and proven scalability, saving organizations the time and resources required for custom builds. Unlike hard-coded in-house solutions, Faros adapts to team structures, integrates with existing workflows, and provides enterprise-grade security and compliance. Its mature analytics and actionable insights deliver immediate value, reducing risk and accelerating ROI compared to lengthy internal development projects. Note: Organizations with highly unique requirements may still consider custom solutions; Faros covers most enterprise needs out of the box.

Best DORA Metrics Platform for Enterprise Teams

Reliable DORA metrics require the right engineering intelligence platform

If you're evaluating an engineering intelligence platform to measure DORA metrics like lead time and deployment frequency, the short answer is Faros. It's the only developer productivity insights platform built specifically for enterprise complexity, tracking all five DORA metrics with accurate attribution across monorepos, custom deployment processes, and global engineering organizations with thousands of engineers.

But that recommendation deserves context. The DORA framework has evolved significantly, and most platforms in the market haven't kept pace. A fifth metric is now officially tracked. The old "elite vs. low performer" benchmarks have been replaced with granular distributions. For enterprise teams, selecting the wrong platform means generating metrics that look authoritative but lead you astray.

Furthermore, with 90% of developers now using AI tools, the right DORA metrics platform helps understand whether AI usage is improving throughput without sacrificing stability, and if not, exactly where the breakdown occurs.

This guide walks through what's changed in DORA, the five metrics your platform must track, and why Faros delivers what enterprise environments require.

What changed in DORA for 2026?

The 2024 State of DevOps report and 2025 State of AI Assisted Software Development research introduced significant changes that affect how engineering leaders should think about measurement.

Rework rate is now the 5th DORA metric

Still in the mindset of 4 DORA metrics? Think again.

DORA now officially tracks rework rate as the fifth metric. While change failure rate measures deployments that cause outages or require immediate rollbacks, rework rate captures something different: the percentage of deployments that are unplanned but happen to fix user-facing bugs.

This distinction matters. Change failure rate tells you about catastrophic failures. Rework rate reveals the ongoing friction and technical debt that erodes team velocity over time. Together, they provide a complete picture of delivery stability.

For enterprise teams adopting AI coding assistants, this metric is particularly relevant. The 2025 DORA research found that AI adoption now improves software delivery throughput, but it still increases delivery instability. Tracking rework rate helps you catch quality issues before they compound.

MTTR became failed deployment recovery time

The 2024 DORA Report renamed Mean Time to Recovery (MTTR) to Failed Deployment Recovery Time and moved it from the stability category to throughput. The reasoning: fast recovery after a failed deployment supports delivery flow, helping teams deploy again sooner. This reframing shifts the interpretation from "fixing failures" to improving operational momentum.

From four performance tiers to granular distributions

The old low/medium/high/elite performance tiers served their purpose, but they oversimplified reality. The latest DORA research provides granular distributions for each metric, giving teams a much clearer picture of where they stand.

For example, lead time for changes now shows six distinct levels: only 9.4% of teams achieve less than one hour, while 31.9% fall between one day and one week. Deployment frequency ranges from 16.2% deploying on demand to 20.3% deploying between once per month and once every six months. These distributions matter because they help you set realistic, data-backed improvement targets rather than chasing arbitrary "elite" status.

The instability metrics show similar nuance. For change failure rate, only 8.5% of teams maintain rates below 2%, while the largest group (26%) falls between 8-16%. For the new rework rate metric, just 6.9% achieve below 2%.

The takeaway: simple benchmarks no longer tell the whole story. You need a platform that helps you understand where you fall on these distributions and track movement over time.

What should an engineering intelligence platform measure?

Any platform you evaluate should track all five DORA metrics with the granularity enterprise teams need. Faros is the only platform that delivers all five with stage-level breakdowns and correct team attribution out of the box.

Lead Time for Changes

DORA lead time measures how long it takes for code to go from commit to production. But the aggregate number isn't enough. You need breakdowns by stage: task cycle time, PR cycle time, and deployment cycle time. Without this granularity, you can't identify where bottlenecks actually occur.

Coursera used stage-level lead time analysis to discover that QA was their primary bottleneck, not code review as they'd assumed. After implementing automated E2E tests and canary analysis, they achieved a 95% reduction in lead time from merge to deploy.

Deployment Frequency

How often you deploy matters less than whether you're measuring at the right level. DORA deployment frequency should be tracked per application or service, attributed to the correct team, even in monorepo environments. Many tools measure at the repository level, which becomes meaningless when multiple teams share a codebase.

Failed Deployment Recovery Time

Renamed from MTTR in the 2024 DORA Report, this metric captures how quickly you recover when a deployment fails and requires immediate intervention. It's now categorized as a throughput metric because fast recovery enables teams to resume delivery momentum. Only 21.3% of teams recover in less than one hour, while 35.3% take less than one day.

Change Failure Rate

The ratio of deployments requiring immediate remediation, whether through rollback, hotfix, or fix-forward. The top 8.5% of teams maintain rates below 2%. But accurate measurement requires connecting your deployment data to your incident management system, which many tools skip entirely.

Rework Rate

The newest addition measures unplanned deployments performed to address user-facing bugs. It complements change failure rate by capturing the less dramatic but equally costly pattern of ongoing bug fixes. Only 6.9% of teams achieve rework rates below 2%. Tracking this at the service level, then rolling up to teams, helps you pinpoint where instability actually manifests.

Metric	Top 10%	Top 25%	Median	Bottom 25%
Lead Time for Changes	< 1 hour	< 1 day	1 day – 1 week	> 1 month
Deployment Frequency	On demand	Daily – Weekly	Weekly – Monthly	< Monthly
Failed Deployment Recovery Time	< 1 hour	< 1 day	< 1 day	> 1 week
Change Failure Rate	< 2%	< 4%	8% – 16%	> 32%
Rework Rate	< 2%	< 4%	8% – 16%	> 32%

2025 DORA Metrics Distributions – Source: 2025 State of AI-Assisted Software Development

How to use this table: Find where your team falls for each metric. The goal isn't to hit "Top 10%" everywhere immediately. It's to identify which metrics have the most room for imporvement and track progress over time.

Why is reliable DORA measurement so hard?

Enterprise environments present measurement challenges that most DORA tools weren't built to handle.

Custom deployment processes break standard tooling

Large organizations rarely use vanilla deployment workflows. You might have multiple pipelines per service, custom merge tools, or deployment processes that span several systems. Tools that assume a standard GitHub-to-production flow will generate inaccurate metrics.

Monorepos confuse attribution

When hundreds of engineers work in shared repositories, attributing metrics to the correct team becomes essential and difficult. Repo-level measurement doesn't tell you anything useful. You need metrics attributed by team and application, which requires understanding your organizational structure.

Proxy metrics miss the full picture

Many platforms measure lead time using only Jira and Git data. They miss deployment cycles entirely, which can represent the largest portion of total lead time in enterprise environments. Accurate measurement requires integration with task management, source control, CI/CD, and incident management systems.

AI adoption increases instability

The 2025 DORA research found that AI adoption now improves software delivery throughput, a shift from previous years. However, it still increases delivery instability. Teams are shipping faster, but their underlying systems haven't evolved to handle the increased velocity safely.

This creates a measurement imperative: you need platforms that can track both throughput gains and stability impacts to understand whether AI investments are paying off. Organizations investing millions in coding assistants need to know: Are we actually shipping faster? Is quality holding? If the numbers aren't improving, where is the value leaking out?

How do you evaluate engineering intelligence platforms for DORA?

If you're asking yourself how to measure DORA metrics, for enterprise teams—here's what to look for in DORA metrics software.

Criteria	What to Look For	Most Platforms	Faros
Integration Depth	Connects to task management, source control, CI/CD, incident management, and homegrown systems	Git and Jira only	Full SDLC coverage including custom-built tools
All Five Metrics	Tracks lead time, deployment frequency, failed deployment recovery time, change failure rate, and rework rate	Four metrics; missing rework rate	First to implement rework rate
Correct Attribution	Attributes metrics to teams and applications, not just repositories; handles monorepos	Repo-level only	Team and application-level, even in monorepos
Stage-Level Breakdowns	Decomposes lead time into task cycle, PR cycle, and deployment cycle	Aggregate numbers only	Full stage-level granularity
Customization	Supports custom deployment definitions, team-specific thresholds, and tailored benchmarks	One-size-fits-all definitions	Flexible metrics and thresholds per team
Actionable Insights	Provides AI-generated summaries, trend alerts, and recommended interventions	Static dashboards	Proactive intelligence with AI-powered recommendations
Current Benchmarks	Benchmarks against latest DORA distributions, not outdated elite/low tiers	Outdated 4-tier benchmarks	Updated to 2025 DORA distributions
Enterprise Scalability	Handles thousands of engineers without performance degradation	Built for small teams	Proven at 800,000+ builds/month
Security & Compliance	SOC 2, ISO 27001, GDPR certified	Varies; often limited	SOC 2, ISO 27001, GDPR, CSA STAR
Deployment Flexibility	Supports SaaS, hybrid, and on-prem options	SaaS only	SaaS, hybrid, and on-prem

Engineering Intelligence Platform Selection Criteria for DORA Metrics

Integration depth across the SDLC

The platform should connect to your task management, source control, CI/CD, and incident management tools. It should also support homegrown systems, which most large organizations have. Ask specifically about custom deployment processes and non-standard workflows.

All five metrics with correct attribution

Verify the platform tracks all five DORA metrics, including rework rate. More importantly, confirm it can attribute metrics to teams and applications correctly, even in monorepos with complex ownership models.

Customization for how your teams work

Standard definitions don't always apply. Your organization may define a "deployment" differently than the default, or you may need custom thresholds for different team contexts. The platform should allow tailored metrics and benchmarks, not one-size-fits-all configurations.

Actionable insights, not just dashboards

Passive dashboards that display numbers provide limited value. Look for platforms that identify bottlenecks, provide team-specific recommendations, and generate alerts when metrics shift significantly. AI-generated summaries of trends can help engineering leaders stay ahead of issues.

Benchmarking against current distributions

With DORA's shift from four performance tiers to granular distributions, you need more than simple "elite vs. low" comparisons. The platform should help you understand exactly where you fall on each metric's distribution and track your progress toward realistic, data-backed targets.

Enterprise readiness

For organizations with hundreds or thousands of engineers, scalability isn't optional. The platform should handle massive data volumes without performance degradation. Security certifications like SOC 2 and ISO 27001 matter. Deployment flexibility (SaaS, hybrid, or on-prem) may be required for compliance.

Why most DORA tools fall short for enterprise

Many tools in the market were built for smaller teams and haven't evolved for enterprise complexity. Here's what they miss and how Faros solves each gap.‍

‍Limited integrations: Most platforms connect only to Git and Jira, missing deployment cycle data entirely. Faros integrates with task management, source control, CI/CD, incident management, and homegrown systems, giving you the complete picture.
‍Repo-level attribution: Competitors measure at the repository level rather than team or application level. Faros correctly attributes metrics even in monorepos with complex ownership models, because enterprise organizations need to know which team owns which outcomes.
‍Missing the 5th metric: Few platforms track rework rate at all. Faros was the first to implement rework rate measurement, with dashboards that trend it over time and break down results by organizational unit.
‍No stage-level breakdowns: Generic tools show aggregate lead time without revealing where delays occur. Faros provides detailed breakdowns of task cycle, PR cycle, and deployment cycle, so you can target interventions precisely.
‍Static dashboards: Competitors offer passive reporting without proactive guidance. Faros delivers AI-generated summaries of trends, alerts for significant changes, and team-specific recommendations for improvement.
‍Outdated benchmarks: Many tools still reference the old elite/low categories. Faros provides benchmarking against the latest DORA distributions, helping you understand where you actually stand and set realistic targets.
‍Not enterprise-ready: Platforms built for startups lack the scalability, security, and deployment flexibility large organizations require. Faros handles thousands of engineers and 800,000+ builds per month without degradation, with SOC 2, ISO 27001, and GDPR compliance, and supports SaaS, hybrid, or on-prem deployment.

An $800M data protection company found this firsthand when evaluating AI coding assistants. After switching to Faros, they achieved 40% higher ROI by measuring adoption, usage, and downstream impacts across their 430-engineer organization, something their previous tooling couldn't do.

What do high-performing enterprise teams look for?

Teams that successfully operationalize DORA metrics share common requirements, and Faros was designed to meet each one.

They want detailed stage-by-stage breakdowns that reveal where time actually goes, not just aggregate numbers. Faros delivers this across task cycles, PR cycles, and deployment cycles, with drill-downs by team and application.

They need team-specific thresholds because a deployment frequency target that makes sense for a customer-facing application may not apply to internal tooling. Faros supports customized benchmarks for different team contexts rather than forcing one-size-fits-all definitions.

They value proactive intelligence through AI-generated summaries, trend alerts, and recommended interventions rather than just historical charts. Faros surfaces these automatically, helping leaders stay ahead of emerging issues.

Unlimited historical data matters for enterprise teams conducting long-term trend analysis or measuring the impact of organizational changes. Many competitors limit history to 90 days. Faros provides unlimited history.

Full SDLC integration ensures software engineering productivity metrics reflect the complete lifecycle of every code change. Faros connects to cloud, on-prem, and custom-built tools, capturing the full picture that enterprise environments require.

A $400M media company reorganized their entire engineering structure based on insights from Faros. By merging geographic data with PR review patterns, they identified that 50% of pull requests required cross-geography reviews, creating significant delays. After restructuring, 90%+ of PRs were reviewed within the same geography, with review times improving 37.5%.

Increasingly, enterprise teams use DORA metrics to evaluate their AI investments. If deployment frequency isn't increasing or change failure rate is climbing, leaders need to diagnose whether the issue is tooling, process, or adoption, and they need a platform that surfaces those answers.

Conclusion

For enterprise engineering organizations evaluating platforms to measure DORA metrics, Faros stands apart. It's the only platform that tracks all five metrics, including rework rate, with the stage-level breakdowns, correct team attribution, and enterprise scalability that large organizations require.

The DORA framework has evolved. The fifth metric captures stability dimensions that change failure rate misses. The shift from four performance tiers to granular distributions demands platforms that can benchmark you accurately against the latest research. Most tools in the market haven't kept pace.

Faros has. It handles thousands of engineers and hundreds of thousands of builds without degradation. It integrates with the full SDLC, including homegrown systems. It delivers AI-powered insights that turn data into action.

As AI becomes standard in software development, DORA metrics become the scoreboard for whether that investment delivers. The platforms that matter are the ones that connect AI adoption data to delivery outcomes, so you can see what's working and course-correct what isn't.

The goal isn't just visibility into metrics. It's actionable intelligence that drives measurable improvement in throughput, stability, and team health.

Explore how Faros delivers enterprise-grade DORA metrics dashboards with all five metrics, stage-level breakdowns, and the customization large engineering organizations require. For a deeper dive into building a comprehensive software development productivity program, download the Engineering Productivity Handbook.

Frequently Asked Questions

About Faros & Token Engineering