Why Engineering Has the Most Complex SaaS Stack
Every department buys software. But engineering teams buy more of it, across more categories, with more overlap, and with less finance visibility than any other function in the company.
Based on Cledara platform data, engineering-tagged tools represent roughly 23% of all software vendors a company uses, and 88% of companies on the platform are paying for at least one engineering tool. The average company manages about 8 dedicated engineering tools on top of the broader company-wide stack. That number grows fast: when you factor in infrastructure, monitoring, CI/CD, developer productivity tools, and the newest category (AI coding assistants), engineering teams often account for the single largest share of a company's SaaS spend.
There are structural reasons for this complexity. Engineers are power users who evaluate tools deeply before committing. They have strong preferences (ask any team whether they prefer GitHub or GitLab, and you will get opinions). They work across multiple layers of the stack simultaneously: code, infrastructure, testing, deployment, observability, and collaboration. Each layer has its own best-of-breed tool, and each tool has its own pricing model, billing cycle, and admin interface.
The result: engineering SaaS stacks grow organically, often without centralised oversight, until someone in finance or IT asks, "Why are we paying for three different monitoring tools?" According to BetterCloud, the average company now uses 106 SaaS applications. Engineering's share of that number is disproportionately large, and growing.
The Typical Engineering SaaS Stack (Mapped by Function)
Before you can rationalise your engineering tool stack, you need to see it clearly. Here is how the typical engineering SaaS stack breaks down by function, based on what companies on the Cledara platform actually pay for. Each category has its own cost profile, growth dynamics, and management challenges.
Source Control and CI/CD
This is the foundation. GitHub dominates: it is the most widely adopted engineering tool on the Cledara platform, used by more companies than any other single engineering vendor. GitLab, Bitbucket, CircleCI, Buildkite, Semaphore, and Bitrise fill out the category. Companies in this segment spend an average of roughly $9,600 per year on source control and CI/CD tools.
The hidden cost here is overlap. A company might use GitHub for source control and CircleCI for CI/CD, then discover that GitHub Actions could replace half of what CircleCI does. Or an engineering team migrates from Bitbucket to GitHub but keeps both subscriptions alive during a "transition period" that never ends.
Infrastructure and Monitoring
This is typically the most expensive category. Companies on the Cledara platform spend an average of roughly $14,500 per year on infrastructure and monitoring tools. The list includes hosting platforms (Vercel, Netlify, Fly.io), databases (Supabase, Neon, PlanetScale), monitoring and observability (Datadog, Sentry, New Relic, Grafana, Elastic), incident management (PagerDuty), feature flags (LaunchDarkly), and CDNs (Cloudflare).
Infrastructure spend is hard to predict because many of these tools use consumption-based pricing. A Datadog bill can double in a quarter if your team ships a feature that generates more metrics. Sentry costs scale with error volume. PlanetScale charges by storage and reads. Without proactive monitoring, these subscriptions can quietly become your largest line items.
Developer Productivity and Collaboration
This category covers everything engineers use to plan, communicate, and ship: IDEs (JetBrains, with roughly 400 companies on the Cledara platform paying for it), project management (Linear, Jira, Atlassian, ClickUp, Asana), documentation (Notion, Confluence), design collaboration (Figma, Miro, Whimsical), communication (Slack, Loom), API development (Postman), and testing (BrowserStack, LambdaTest, Cypress).
Developer productivity tools are the broadest category and the one most likely to overlap with tools other teams already use. Does engineering really need Linear if the rest of the company is on Asana? Does the design team's Figma subscription overlap with engineering's Whimsical usage? These are the questions that surface during a proper tool audit. (For a deeper look at how to run one, see our guide to SaaS consolidation.)
AI Coding Assistants (the Newest Category)
This is the fastest-growing segment in engineering SaaS. On the Cledara platform, 34% of companies now pay for at least one AI coding assistant. The growth trajectory is steep: unique AI tool subscriptions went from 18 in 2022 to over 1,650 in 2025, and 2026 is on track to exceed that.
The major players are Cursor (averaging roughly $5,900 per company per year on the Cledara platform), GitHub Copilot, Windsurf, CodeRabbit, Tabnine, and Supermaven. Broadly, the AI coding assistant market hit $7.4 billion in 2025. Cursor alone crossed $2 billion in annualised revenue by early 2026, while GitHub Copilot holds roughly 42% market share with over 20 million users.
The management challenge with AI coding assistants is threefold. First, many use consumption-based pricing (API calls, tokens), making costs unpredictable. Second, adoption often starts with individual developers paying with personal credit cards, creating shadow IT. Third, the category is evolving so fast that what you evaluate in January may be obsolete by June.
Common Problems: Redundancy, Sprawl, and Shadow Dev Tools
Engineering tool stacks do not become messy because engineers make bad decisions. They become messy because engineers make local decisions: each team picks the tool that is optimal for their immediate problem, without visibility into what other teams already use or what the company already pays for.
Here are the patterns that show up repeatedly.
Redundancy across teams. One team uses Datadog for monitoring while another uses New Relic. Both are solving the same problem. Neither team knows the other's tool exists because the subscriptions were set up independently, often on different credit cards. The company pays for two monitoring platforms when one would suffice.
Zombie subscriptions. An engineer signs up for a tool during a proof of concept. The POC ends, the team moves on, but nobody cancels the subscription. The credit card keeps getting charged. Multiply this across a 100-person engineering org, and you can easily find 10 to 15 subscriptions that nobody is using.
Shadow IT (and shadow AI). Developers are the number one shadow IT offenders. Research shows that 52% of developers do not use IT-approved tools, and engineering teams have the highest shadow AI adoption rate at 79%. Shadow AI tool usage increased 156% between 2023 and 2025, and 43% of companies still have no policy on AI tool usage. This is not just a cost problem; it is a security and compliance risk. When an engineer signs up for an AI tool using a personal email and feeds proprietary code into it, the company has zero visibility into what data is leaving the building. Our SaaS governance guide covers how to build policies that address this.
Approval bottlenecks that backfire. Some companies respond to sprawl by adding heavy procurement processes. This creates a different problem: engineers bypass the process entirely because waiting three weeks for approval on a $20/month tool is not compatible with shipping software. The governance system fails because it was designed for enterprise procurement, not developer workflows. (For more on building processes that actually work, see our IT Manager's SaaS Management Playbook.)
A Framework for Engineering Tool Governance
Good engineering tool governance balances two things: visibility and control for finance and IT, with speed and autonomy for developers. Get the balance wrong in either direction, and you end up with either unchecked sprawl or bureaucratic friction that slows down your best people.
Here is a practical framework that works for engineering teams at the 30 to 500 employee stage, where formal IT governance is often still maturing but the tool stack is already complex.
1. Build your engineering tool inventory. You cannot rationalise what you cannot see. Start by cataloguing every tool your engineering teams use, including the ones that are not on official procurement lists. Browser-level discovery tools (like Cledara's Engage extension) catch shadow IT that invoice audits miss. Tag each tool by category (source control, CI/CD, monitoring, productivity, AI), owning team, cost, billing cycle, and contract renewal date.
2. Set tiered approval thresholds. Not every tool purchase needs the same level of scrutiny. A practical approach: tools under $50/month get auto-approved by the engineering manager; $50 to $500/month requires IT and finance sign-off; above $500/month requires VP-level approval with a business case. The key is that the low tier is genuinely fast, so developers do not feel tempted to go around the process.
3. Create a preferred tool stack. For each category, designate a preferred tool. This does not mean mandating one option and banning everything else. It means making the default path clear: "We use GitHub for source control, Datadog for monitoring, Linear for project management." Engineers can request exceptions, but the burden of justification shifts to the person proposing the new tool, not the person defending the status quo.
4. Review usage, not just spend. A tool that costs $500/month and is used daily by 20 engineers is a great investment. A tool that costs $200/month and was last logged into three months ago is waste. Usage analytics are more important than spend analytics for engineering tools, because the question is not "are we paying too much?" but "are we getting value?"
5. Manage AI tools as a distinct category. AI coding assistants are too new, too fast-moving, and too consumption-driven to fit neatly into existing tool governance. Give them their own budget line, their own review cadence (quarterly, not annually), and their own monitoring. Track API spend daily, not monthly, because a single runaway integration can blow through a budget in days.
6. Consolidate at renewal time, not before. Ripping out tools mid-contract is disruptive and expensive. Instead, build your rationalisation plan around renewal dates. When a Sentry renewal comes up, evaluate whether Datadog's error tracking covers the same use case. When a Jira renewal hits, assess whether Linear has enough adoption to justify a full migration. Renewal dates are the natural decision points. Build a calendar that tracks every engineering tool renewal, and start the evaluation process 60 to 90 days before each one.
How Cledara Helps Engineering Organisations
Engineering teams present a specific set of challenges: high tool volume, consumption-based pricing, fast-moving categories (especially AI), and a culture that resists heavy procurement processes. Cledara is built to handle exactly this combination.
Virtual cards per subscription. Cledara issues a unique virtual card for every engineering tool. Each card has its own spend limit, which means a runaway Datadog bill cannot affect your GitHub subscription. If a tool needs to be cancelled, freeze the card. No vendor runaround, no waiting for a contract clause to kick in. For engineering teams with dozens of subscriptions, this granularity is the difference between controlled spend and chaos.
AI Dashboard. Cledara connects to AI providers (OpenAI, Anthropic, Cursor) via API keys to track usage-based AI spend with daily visualisation and budgets. Given that AI coding assistant adoption grew from near-zero to 34% of companies on the platform in just three years, having dedicated monitoring for this category is not optional; it is essential.
Approval flows designed for speed. Cledara's approval workflows are configurable by spend threshold: low-cost tools can be approved quickly (single manager approval), while larger commitments route through IT and finance. Engineers request tools through a simple workflow instead of filing procurement tickets. The result is compliance without bottlenecks.
Engage (usage analytics). Cledara's Engage browser extension tracks which developer tools are actually being used, not just which ones are being paid for. It deploys across Chrome, Safari, and Firefox, and it is privacy-first: it only tracks SaaS provider URLs, never browsing history. For engineering managers, this answers the critical question: "Which of our tools are delivering value, and which ones are zombie subscriptions?"
Accounting automation. Every engineering tool subscription gets mapped to the correct GL code, department, and cost centre automatically, with native integrations for Xero, QuickBooks, and NetSuite. For finance teams reconciling engineering spend, this eliminates the monthly scramble of matching credit card charges to invoices to expense codes.
The net effect: engineering teams keep the autonomy they need to move fast, while finance and IT get the visibility and control they need to manage spend and risk. Cledara customers see an average 23% reduction in SaaS costs and save over 13 hours per month on SaaS administration. For engineering organisations where the tool stack is large, fast-changing, and mission-critical, that combination of savings and time is significant.
Rationalise your engineering tools without slowing down your developers. See how Cledara works.
