The Top Benefits of Declarative ETL for Enterprise Teams

That sinking feeling when you see the monthly cloud bill from Snowflake or Splunk is all too common. Costs spiral not just because of data volume, but because traditional ETL jobs are incredibly inefficient, moving and processing far more data than necessary. This is a direct result of their imperative, step-by-step nature, which lacks built-in optimization. A declarative model offers a powerful solution by focusing on the end state, not the process. By defining your desired outcome, the system can automatically determine the most efficient execution path, minimizing data movement and resource consumption. Understanding the benefits of declarative ETL is the first step toward regaining control over your data infrastructure costs and turning your data platform into a predictable, high-ROI investment.

Key Takeaways

• Define the 'what,' not the 'how': Shift your team's focus from writing complex, step-by-step instructions to simply defining the desired data outcome. This approach reduces code, minimizes maintenance, and frees up your engineers to work on high-impact projects.
• Build compliance and quality into your pipelines: Embed data quality rules and governance policies directly into your pipeline definitions. This automates enforcement, creates clear audit trails, and ensures your data is trustworthy and compliant by design, not by chance.
• Reduce costs with smarter resource management: A declarative system automatically optimizes how your data is processed, minimizing resource consumption and unnecessary data movement. This leads to direct savings on cloud infrastructure and brings much-needed predictability back to your data platform budget.

What is Declarative ETL?

If you've ever felt like your data engineers spend more time wrestling with brittle pipelines than delivering actual insights, you're not alone. The traditional way of building data pipelines is often complex, manual, and a significant source of technical debt. Teams write detailed, step-by-step instructions for every part of the process, which creates rigid systems that are difficult to maintain and scale. When something breaks—and it always does—engineers have to manually trace the problem through lines of procedural code.

Declarative ETL offers a more streamlined and resilient approach. Instead of writing instructions on how to move and transform data, you simply define what you want the final result to look like. You specify the sources, the required transformations, and the target schema, and the system figures out the most efficient way to get it done. This fundamental shift from focusing on the process to defining the outcome is key to building more manageable and scalable data workflows for your enterprise.

Declarative vs. Traditional ETL: What's the Difference?

In a traditional, or imperative, ETL process, your engineers code every single step: extract data from this source, apply this specific transformation, handle this potential error, and load it into that destination. It’s a rigid script. A declarative approach, on the other hand, lets you focus on the business logic. You define the transformations and the structure of the final dataset, while the platform automatically handles the orchestration, infrastructure management, and error handling. Think of it less as a strict divide and more of an evolution—moving from giving detailed commands to simply describing the desired data outcomes.

How Does Declarative ETL Actually Work?

So how does the system manage all that complexity? A declarative ETL framework acts as an intelligent manager for your data pipelines. Once you’ve defined your desired end state, the system takes over the heavy lifting. It automatically orchestrates the workflow, making sure every step runs in the correct sequence. It also builds in data quality checks and monitors the pipeline's health to ensure everything is running smoothly. This means your team can stop micromanaging the process and instead trust the system to handle the intricate details of creating reliable data pipelines, freeing them up to focus on higher-value work.

Why Enterprise Teams Are Adopting Declarative ETL

If you've ever felt like your data engineering team is stuck on a hamster wheel of building and fixing pipelines, you're not alone. Traditional ETL processes, where you have to manually script every single step of the data journey, are becoming a major bottleneck for large organizations. This imperative, step-by-step approach creates complex and brittle systems that are expensive to run and a nightmare to maintain. Every new data source or schema change threatens to break the entire chain, forcing your best engineers to spend their days firefighting instead of innovating.

This is precisely why so many enterprise teams are shifting to a declarative model. Instead of telling the system how to transform data, you simply declare the what—the final state you want your data to be in. The platform handles the rest, optimizing the execution plan, managing dependencies, and ensuring data quality along the way. This fundamental shift allows teams to build more resilient, scalable, and cost-effective data infrastructure. It’s about trading manual toil for intelligent automation, which is critical for any organization looking to get faster, more reliable insights from its data. Expanso's approach to distributed data processing aligns with this modern paradigm, enabling right-place, right-time compute without the operational overhead.

Uncovering the Hidden Costs of Complex Data Pipelines

Complex data pipelines often come with a shocking price tag, and it’s not just about the cost of your data platform license. Traditional, imperative ETL jobs can be incredibly inefficient, moving and processing massive volumes of data unnecessarily. This directly translates into higher cloud compute and storage bills, causing those unpredictable spikes in your Snowflake or Splunk costs. A declarative approach tackles this head-on by automatically optimizing transformations. Think of it as a smart system that finds the most efficient path for your data, reducing the volume that needs to be moved and processed. This streamlined execution means you can achieve the same outcomes with a fraction of the resources, bringing much-needed predictability back to your budget.

Solving Pipeline Reliability and Maintenance Headaches

Pipeline fragility is a constant source of stress for data teams. When engineers spend most of their time on pipeline prep, cleaning, and fixing brittle connectors, they have little time left for building new analytics or AI projects. Declarative ETL helps break this cycle by abstracting away the complex implementation details. You define your desired data outcomes, and the system automatically handles the underlying tasks, like managing schema changes, checking data quality, and mapping data flows. This built-in automation means you write less code, which in turn means there are fewer things that can break. The result is a more resilient data infrastructure that requires far less manual intervention, freeing your team to focus on work that actually drives business value.

How Declarative ETL Improves Developer Productivity

Your data engineering team is under constant pressure to deliver reliable data faster. But traditional ETL pipelines often become a bottleneck, bogged down by complex, hand-written code and endless maintenance cycles. This is where a declarative approach changes the game. Instead of forcing developers to specify every single step of the data transformation process—the how—declarative ETL lets them simply define the desired outcome—the what.

This fundamental shift frees your engineers from writing and maintaining thousands of lines of boilerplate code for tasks like orchestration, error handling, and performance tuning. The system handles the execution details, allowing your team to focus on higher-value work, like designing better data models and delivering insights to the business. By abstracting away the low-level complexity, you empower your developers to build, test, and deploy pipelines more quickly and with greater confidence. This not only accelerates project timelines but also improves job satisfaction by letting talented engineers solve interesting problems instead of just keeping the lights on. Expanso’s distributed computing solutions are built on this principle, helping teams streamline their data operations from edge to core.

Write Less Code, Develop Faster

Imagine your developers could describe the final state of the data they need, and the system would figure out the most efficient way to produce it. That’s the core benefit of declarative ETL. Because so many features are built-in—like automatically handling schema changes, running data quality checks, and mapping data flows—your team writes significantly less code. This reduces the surface area for bugs and inconsistencies, leading to more robust pipelines from the start. By focusing on the business logic rather than the implementation details, developers can move from concept to production in a fraction of the time it would take with an imperative, step-by-step approach.

Automate Optimization and Performance Tuning

One of the biggest time sinks for data engineers is manually tuning pipelines for performance. With traditional ETL, developers often have to become experts in the underlying infrastructure to optimize resource allocation and ensure jobs run efficiently. Declarative systems take this burden off your team. The framework automatically handles complex tasks like orchestration, resource management, and monitoring. It can intelligently plan execution, parallelize tasks, and scale resources up or down based on the workload. This built-in automation means your team can specify their desired data outcomes without getting bogged down in the weeds of performance tuning, ensuring your data processing is both fast and cost-effective.

Simplify Debugging and Troubleshooting

Troubleshooting a failing data pipeline can feel like searching for a needle in a haystack. When a job breaks, tracing the issue back through multiple stages of complex code is a tedious and frustrating process. Declarative frameworks make debugging much simpler. By defining pipelines as code in a clear, high-level language, it becomes easier to write unit and integration tests, catching bugs before they ever reach production. Furthermore, features like automated data lineage and built-in quality checks provide clear visibility into how data is transformed at each step. This makes it much faster to pinpoint the root cause of an issue and resolve it, improving the overall reliability of your data infrastructure.

How Does Declarative ETL Improve Pipeline Reliability?

Fragile data pipelines are a constant source of stress for enterprise teams. When a job fails, it triggers a frantic scramble to find the root cause, delaying critical analytics and eroding business trust in the data. Traditional, imperative pipelines are often the culprit; their step-by-step instructions create complex dependencies that can break in countless ways. A single change can have unforeseen ripple effects, and troubleshooting feels like untangling a massive knot. This constant firefighting pulls your best engineers away from innovation and into maintenance mode.

Declarative ETL offers a more stable and resilient alternative. Instead of dictating how to process data, you simply define the what—the final state you want your data to be in. The system then figures out the most efficient and reliable way to get there. This approach builds resilience directly into your architecture. Declarative systems are designed to be scalable, handling growing data volumes with ease, and can recover quickly when things go wrong. This shift from manual intervention to automated management is key to building data systems you can actually depend on. Expanso’s approach to distributed computing is built on this principle of creating intelligent, self-managing systems that reduce manual overhead and increase reliability.

Validate Data Quality Automatically

Poor data quality is one of the biggest threats to any analytics or AI initiative. With traditional ETL, data validation is often a separate, downstream step. This means bad data can flow deep into your systems before it’s caught, corrupting reports and models along the way. Cleaning it up is a painful, retroactive process.

Declarative ETL flips the script by letting you embed data quality rules directly into your pipeline definitions. You can specify constraints, such as "this column cannot be null" or "this value must be within a certain range," as part of the desired outcome. The system then automatically enforces these rules during processing. If a record violates a rule, the framework can be configured to quarantine it, alert your team, or halt the pipeline altogether. This proactive approach ensures that only clean, trusted data makes it to your end-users, building a reliable foundation for decision-making.

Handle Errors and Recover with Ease

In a complex data environment, failures are inevitable. A network hiccup, a temporary API outage, or a malformed file can bring a traditional pipeline to a grinding halt, often requiring a data engineer to manually restart the job. This process is time-consuming and prone to human error.

Declarative frameworks, however, are designed for graceful error handling. Because the system understands the target state, it can automatically manage retries and recovery. If a transient error occurs, the system can simply try the operation again without any manual intervention. For more persistent issues, like a batch of corrupted data, it can isolate the problematic records and continue processing the valid ones. This self-healing capability makes your pipelines far more robust, ensuring that minor issues don’t cascade into major outages and that your data flows remain consistent and timely.

Get Real-Time Monitoring and Alerts

With traditional ETL scripts, gaining visibility into pipeline performance can be a challenge. You often have to build custom logging and monitoring solutions just to understand what’s happening under the hood. This lack of built-in observability makes it difficult to proactively identify bottlenecks or troubleshoot issues when they arise.

Declarative ETL solutions provide monitoring and alerting capabilities out of the box. Since the framework manages the entire execution plan, it has a complete picture of every step in the process. It can automatically track data lineage, monitor processing times, and log key events. This information is typically presented in an intuitive dashboard, giving your team real-time visibility into pipeline health. You can easily set up alerts for failures or performance degradation, allowing you to address problems before they impact the business. This level of transparency is one of the key features that makes modern data platforms so powerful.

How Declarative ETL Improves Team Collaboration

Data pipelines are often seen as a purely technical challenge, but they’re really a team sport. When data engineers, analysts, and business leaders can’t speak the same language, projects stall, requirements get lost in translation, and the final data product misses the mark. This friction is a major source of the pipeline fragility and delays that plague so many enterprise projects. The core issue is that traditional, imperative ETL code is written for machines, not for people. It details every step of how to transform data, creating a complex, rigid script that’s difficult for anyone but its original author to understand.

Declarative ETL changes this dynamic by shifting the focus from how to what. Instead of writing step-by-step instructions, your team defines the desired end state of the data. This approach creates a common ground where technical and non-technical team members can collaborate effectively. It turns the data pipeline from a black box into a clear, shared blueprint. This transparency is crucial for building reliable, business-aligned data products, especially in complex environments that require robust security and governance. By making the pipeline’s logic accessible to everyone, you reduce misunderstandings and empower your entire team to contribute to data quality and strategy.

Bridge the Gap Between Technical and Business Teams

One of the biggest hurdles in data projects is ensuring the technical implementation matches the business requirements. Declarative ETL helps clear this hurdle by using a more human-readable format. Because it allows teams to define desired data outcomes without writing extensive code, it’s much easier for non-technical stakeholders to understand and contribute to the process. A marketing analyst, for example, can review a declarative pipeline definition and confirm that the logic for customer segmentation is correct without needing to be a SQL or Python expert. This shared understanding ensures that the data products your engineers build are the ones the business actually needs to drive decisions.

Improve Documentation and Version Control

If you’ve ever inherited a complex data pipeline with little to no documentation, you know how challenging it can be. Declarative code is often described as self-documenting. Since the focus is on what the solution should achieve, the intent of the code is more explicit and easier to grasp. This clarity simplifies maintenance and makes onboarding new team members much faster. When combined with version control systems like Git, declarative definitions make it easy to track changes and understand their impact over time. This creates a transparent and auditable history of your data logic, which is a fundamental part of building a reliable distributed data warehouse.

Create a Shared Understanding of Pipelines

Ultimately, better collaboration comes from a shared understanding. Declarative ETL frameworks provide a higher-level abstraction that allows all team members, regardless of their technical expertise, to see the complete data flow and transformations. It’s like having a universally understood blueprint for your data architecture. This clear, unified view fosters better communication, reduces the risk of misinterpretation, and helps teams identify potential issues before they become major problems. When everyone from the CDO to the data engineer is looking at the same picture, you can build more resilient, efficient, and valuable data solutions for the entire organization.

How Declarative ETL Reduces Costs

When you’re managing enterprise-scale data, the costs can feel like a runaway train. Between unpredictable cloud consumption bills, massive storage fees, and the engineering hours spent just keeping brittle pipelines running, budgets are constantly under pressure. This is where a declarative approach to ETL can make a significant impact. By shifting the focus from writing complex, step-by-step instructions to simply defining the desired outcome, declarative ETL cuts costs across the board. It reduces the manual effort required from your team, optimizes the use of expensive infrastructure, and dramatically lowers the long-term burden of maintenance. Let's break down how these savings add up.

Save on Development Time and Resources

Declarative ETL allows your data teams to define the "what" instead of the "how." Instead of hand-coding every transformation, error-handling routine, and dependency, engineers simply state the final form the data should take. The platform handles the rest. This means less code to write, test, and debug. With built-in features for handling data quality checks and schema changes automatically, your team can stop spending the majority of their time on pipeline prep and cleaning. This frees up your most valuable technical talent to focus on higher-impact projects that drive business insights, rather than just keeping the lights on. This efficiency is key to scaling your log processing and other data-intensive workflows without scaling your headcount.

Optimize Infrastructure and Storage Costs

One of the biggest sources of runaway spending is inefficient use of compute and storage. Traditional ETL jobs often require you to over-provision resources to handle peak loads, meaning you’re paying for capacity you don’t always need. Declarative systems, on the other hand, can automatically optimize the execution plan. By analyzing the entire workflow, the platform can determine the most efficient way to process data, minimizing resource consumption and reducing processing times. This leads to direct savings on your cloud bills and helps you avoid the sticker shock of unpredictable consumption charges from data warehouses. It’s a smarter way to manage your data infrastructure and ensure you’re only paying for what you actually use.

Lower Long-Term Maintenance Spending

Brittle, imperative pipelines are a constant maintenance headache. A small change in a source system can cause a cascade of failures that requires hours of troubleshooting to fix. Because declarative ETL abstracts away the underlying complexity, it makes pipelines far more resilient and easier to manage over time. The system handles many of the operational aspects automatically, so when changes occur, the pipeline can often adapt without manual intervention. This significantly reduces the complexity and cost of long-term maintenance. Your team spends less time fighting fires and more time improving and extending your data warehouse capabilities, leading to a lower total cost of ownership and more reliable data delivery.

How Declarative ETL Simplifies Compliance and Governance

Meeting compliance and governance standards isn't just a box to check—it's a fundamental requirement that can feel like a constant battle. For global enterprises, navigating regulations like GDPR, HIPAA, and various data residency laws with traditional ETL is a complex, manual process that’s prone to costly errors. Declarative ETL changes the game by allowing you to build your compliance rules directly into your data pipelines. Instead of reacting to compliance issues, you can design your pipelines to be compliant from the start, turning governance into an automated, reliable part of your workflow. This approach provides a clear, auditable framework that ensures your data handling practices meet the strictest security and governance standards. By defining what you want the outcome to be—compliant, secure data—you let the system figure out how to do it, embedding your policies into the very fabric of your data operations. This shift from procedural to outcome-based design is what makes robust governance achievable at scale, without slowing your teams down or requiring an army of compliance specialists to review every line of code. It transforms compliance from a bottleneck into a built-in feature.

Automate Data Lineage and Audit Trails

If you’ve ever been asked to prove where a piece of data came from and every transformation it went through, you know how difficult that can be with traditional ETL. Because you’re defining every step, tracing data lineage often involves manually piecing together complex scripts and logs. Declarative ETL simplifies this by focusing on the outcome. Since you define the end state, the framework can automatically generate clear data lineage and audit trails. This creates a transparent record of data movement and transformation, making it much easier to demonstrate compliance and respond to regulatory audits without weeks of forensic work.

Validate Compliance Rules Automatically

Ensuring that sensitive data is properly masked or that data quality meets specific standards is often a separate, manual step in traditional pipelines. This introduces a significant risk of human error. Declarative ETL frameworks allow you to embed validation directly into your pipeline logic. You can define rules—like ensuring all personally identifiable information (PII) is anonymized—as part of the desired outcome. The system then automatically validates the data against these rules during processing. This proactive approach helps you catch and fix issues instantly, ensuring that only compliant, high-quality data reaches its destination.

Enforce Data Residency and Cross-Border Controls

For global organizations, data residency is a major compliance hurdle. Manually ensuring that data from a specific region is processed and stored within that region is incredibly complex and fragile. With declarative ETL, you can define data residency requirements as a core part of your pipeline's configuration. For example, you can specify that all European customer data must be processed on servers within the EU. The system automatically enforces these cross-border controls, simplifying how you manage a distributed data warehouse and reducing the risk of accidentally violating regulations like GDPR. This lets you operate globally while adhering to local rules.

What Are the Challenges of Adopting Declarative ETL?

Adopting a declarative approach to ETL is a smart move, but let's be real—it’s not a magic wand. Like any significant shift in technology and process, it comes with its own set of hurdles. Being aware of these challenges upfront is the best way to create a smooth transition for your team and set your project up for success.

The main obstacles aren't always technical. They often involve your team's existing skills, your company's established processes, and the legacy systems you've relied on for years. Slow or inflexible data pipelines can delay critical business decisions, so it's important to address these potential roadblocks head-on. Planning for the learning curve, system integrations, and data dependencies will help you get the full benefits of declarative ETL without the friction.

Managing the Learning Curve and Organizational Change

Shifting from an imperative to a declarative mindset is the biggest adjustment your team will face. Engineers who are used to writing step-by-step instructions for data transformations now have to learn to define the desired outcome and trust the system to figure out the "how." This requires a different way of thinking and a new set of skills.

This transition isn't just a technical training issue; it involves realigning organizational processes. Initially, you might see a temporary dip in productivity as the team gets comfortable with new tools and workflows. Getting buy-in from everyone, from data engineers to business analysts, is key. You need to clearly communicate why the change is happening and what the long-term benefits are for everyone involved.

Integrating with Legacy Systems

One of the most common misconceptions about modernizing your data stack is that you can simply "lift and shift" old logic into a new framework. Unfortunately, it's rarely that simple. Your legacy ETL jobs are likely intertwined with older systems, each with its own specific data formats, APIs, and operational quirks.

Rewriting these pipelines for a declarative model requires more than just a direct translation. You have to untangle years of custom scripts and business logic built for a different architecture. This process of rewriting legacy ETL demands a careful migration strategy that accounts for every dependency and ensures data integrity is maintained throughout the transition.

Handling Dependencies and Schema Changes

Data is never static. Source systems change, schemas evolve, and new data fields are added—sometimes without warning. Managing these changes is a persistent challenge in any data pipeline, and declarative systems are no exception. While they are designed to be more resilient, you still need to define and manage the dependencies between different stages of your pipeline.

Creating accurate source-to-target data mappings is often a time-consuming but critical step. When a schema changes upstream, it can have a ripple effect that breaks downstream processes. A solid declarative ETL solution should help you visualize these dependencies and automate parts of the impact analysis, but it still requires careful oversight from your team to handle unexpected changes gracefully.

How to Choose the Right Declarative ETL Solution

Switching to a declarative model is a big move, so picking the right solution is critical. You’re not just buying a tool; you’re adopting a new way of managing your data pipelines. The ideal platform should feel less like a smart partner that understands your goals and more like a rigid set of instructions. It needs to handle the heavy lifting of optimization, scaling, and error handling so your team can focus on delivering value from your data.

As you evaluate your options, think about your entire data ecosystem. The right tool will integrate smoothly with your existing systems, from data warehouses like Snowflake to observability platforms like Datadog. It should also provide clear visibility into your pipelines, making it easier to maintain data quality and meet strict compliance standards. Look for a solution that offers both powerful automation and the flexibility to handle your unique, complex use cases without forcing you into a corner.

Key Features to Look For

When you’re vetting declarative ETL solutions, start by looking for a platform that lets you define the what, not the how. A great tool allows you to specify your desired final data state, and it intelligently figures out the most efficient way to get there. This approach drastically reduces the amount of code your team has to write and maintain. Look for key features like automatic optimization, which streamlines transformations to reduce data movement and lower processing costs. Your chosen solution should also include built-in data quality checks and validation rules, ensuring that only clean, reliable data makes it to its destination.

Evaluating Scalability and Performance

Your data volume isn't shrinking, so your ETL solution needs to scale effortlessly. Assess how each platform handles large datasets and complex workloads. Can it process data in real time or near-real time to support time-sensitive analytics? For example, companies like Volvo use declarative pipelines to get real-time inventory processing and end-to-end order tracking. A strong solution should leverage distributed computing to handle massive throughput without bottlenecks, whether you're managing a distributed data warehouse or processing logs from thousands of sources. Performance isn't just about speed; it's about maintaining that speed as your business grows.

Assessing Security and Integration Capabilities

For any enterprise, especially in regulated industries like finance or healthcare, security and governance are non-negotiable. A robust declarative ETL solution must provide comprehensive security and governance features. This includes granular access controls, data masking for sensitive information, and automated lineage tracking for auditability. Ensure the tool can enforce data residency rules to meet compliance standards like GDPR and HIPAA. Equally important is its ability to integrate with your existing security infrastructure and monitoring tools. Poor data quality can lead to serious compliance issues, so look for a platform that makes it easy to monitor performance and maintain data integrity throughout the entire ETL process.

Related Articles

• What Are Declarative Data Pipelines? A Guide | Expanso
• What is a Declarative Pipeline? A Simple Breakdown | Expanso
• The Data Journey | 5 Stages from Source to AI-Ready | Expanso

Frequently Asked Questions

My team is full of experienced data engineers who are experts at writing ETL scripts. Will they feel like a declarative approach takes away their control? That's a common concern, but it's really about shifting where your team applies its expertise. Instead of controlling the low-level mechanics of data movement, a declarative model lets your engineers control the high-level architecture and business logic. They can focus their skills on designing efficient data models and ensuring data quality, rather than getting bogged down in manual performance tuning and error handling. It empowers them to solve more strategic problems.

This sounds like a big change. Can we apply declarative principles to our existing pipelines, or is this only for brand-new projects? You don't have to rip and replace everything at once. Many teams start by identifying a specific pain point, like a particularly brittle or expensive pipeline, and rebuilding it using a declarative approach. This allows you to demonstrate the value and learn the new workflow on a smaller scale. Over time, you can gradually migrate more of your legacy jobs as they need to be updated or retired, making the transition manageable.

How does declarative ETL actually reduce my cloud data warehouse or SIEM costs? Can you give a practical example? It primarily saves money by being more efficient. Imagine you have a pipeline that processes raw logs. A traditional script might pull all the data, then filter and transform it in several steps, using significant compute resources along the way. A declarative system can analyze the entire workflow and optimize it, perhaps by pushing a filter directly to the source to avoid moving unnecessary data in the first place. This reduction in data movement and processing directly translates to lower consumption costs on platforms like Snowflake or Splunk.

What's the biggest hurdle teams face when they switch to a declarative model? The biggest challenge is usually the mental shift, not the technology itself. Engineers who are accustomed to writing detailed, step-by-step instructions have to learn to trust the system to handle the execution. It requires moving from thinking about the process to thinking about the outcome. Getting comfortable with this new way of defining work takes a bit of time, but once it clicks, teams find they can build and iterate much faster.

How does this approach help when business requirements inevitably change? This is where declarative ETL really shines. When a business requirement changes—say, you need to add a new field to a report—you simply update the definition of your final dataset. You're not hunting through lines of procedural code to figure out where to insert a new transformation step. Because the code is more readable and focused on the outcome, making changes is faster and far less likely to break something downstream. This makes your entire data infrastructure more agile and responsive to business needs.