---
title: "Performance and cloud cost optimization | WIP"
description: "Improvements to response time, resource usage, database queries, architecture, and cloud costs, measured before and after."
url: https://workinprogress.company/services/performance-and-costs
updated: 2026-10-27
lang: en
---

WIP Optimize · Engineering project

# Better performance and less waste, measured before and after

WIP Optimize improves systems that work but could be faster or cheaper. We tackle slow queries, resource usage, architectural bottlenecks, and operating costs, always measuring before and after each change. Scope and price are set after the diagnosis, once it is clear where the potential gains are.

quoted after assessment

Updated on October 27, 2026

 ![A data pipeline where the narrow section, the bottleneck, was replaced by a wider one; the old section lies aside.](https://workinprogress.company/_astro/servico-optimize.Bar0IZMr_2qTaMH.png)

## How does the optimization work?

Optimization follows a simple cycle: measure, change one thing, and measure again. Only what improves the agreed metric stays; anything that doesn’t is rolled back.

Before any change, we record the baseline: response times for the main pages and APIs; error rate; CPU, memory, and disk usage; database query times; and the monthly infrastructure cost. Without that snapshot, there is no way to prove that anything improved, or by how much.

From there, every change goes through four steps:

1. **Hypothesis:** where the waste is and what gain is expected.
2. **Isolated change:** one change at a time, to know the effect of each.
3. **Measurement:** comparison with the baseline, under the same conditions.
4. **Decision:** keep, adjust, or roll back.

This cycle guards against two mistakes: changing several things at once without knowing which one helped, and gaining speed at the cost of stability without noticing. It also protects your budget, because the work stops when further gains are no longer worth the effort.

The result is a report with the before and after of each change, so the gains are visible to decision-makers and reproducible for whoever maintains the system.

## Where should you look for performance and cost gains?

Gains show up where the system does unnecessary work or pays for capacity it doesn’t use. The investigation usually starts with the database and the cloud bill, because that is where the signals are easiest to measure.

| Where to look | What to look for |
| --- | --- |
| Database | Queries without indexes, queries repeated inside loops, too many connections, tables that grow without cleanup |
| Application | Heavy work done during the request that could go to a queue, and external calls with no timeout |
| Cache | Responses that rarely change but are recalculated on every request, static files not cached in the browser |
| Infrastructure | Oversized servers, unused disks and IP addresses, forgotten test environments left running |
| Traffic and logs | Expensive data transfer between regions, logs kept with no retention limit, uncompressed images |
| Architecture | Components that could be simpler, and managed services that cost more than they deliver |

Not every performance gain reduces cost, and not every cost reduction improves performance. Sometimes the two goals go hand in hand, as when a cache takes load off the database and allows a smaller server. Sometimes there is a trade-off, and it has to be explicit: the proposal states what you gain, what you pay, and the risk of each option.

## Why is the price set only after the diagnosis?

The price comes after the diagnosis because, before measuring, nobody knows where the gains are or how much work they require. A fixed price given in the dark would have to include a margin for the unknown, and you would pay for that uncertainty.

The way around this is to split the work into two stages. First, a diagnosis with a defined scope, which can be [WIP Check](https://workinprogress.company/services/technical-diagnosis), measures the system and identifies the opportunities. Then the implementation proposal comes with:

- the recommended changes, in order of impact;
- the target for each one, measured against the baseline;
- the effort and price of the implementation;
- what is out of scope, and why.

With that, you decide with numbers in hand: you can hire WIP for all of it, for only the highest-return part, or have your own team carry out the plan.

Infrastructure, licenses, and third-party services remain separate from WIP’s fees, before and after the optimization. That makes it clear how much of the savings came from the infrastructure and how much the work to get there cost.

## What does WIP Optimize deliver?

- Baseline measurement of response time, error rate, resource usage, and cost.
- Optimization of queries, indexes, and database usage.
- Tuning of caching, resource usage, and infrastructure sizing.
- Review of operating costs, with idle resources identified.
- Report with before-and-after measurements for each change.

## Who is WIP Optimize for?

- Systems that slow down at peak times or as their data grows.
- Businesses whose cloud bill has grown faster than their usage.
- Teams that want to lower the cost per operation without losing reliability.

## What is not included in WIP Optimize?

- A promised percentage of improvement before anything is measured: targets come out of the diagnosis.
- A complete rewrite of the application, which is assessed under WIP Modernize when warranted.
- Infrastructure, license, and tool costs, which are kept separate from WIP's fees.

## How much does WIP Optimize cost?

WIP Optimize is quoted after a technical assessment. Price set after the diagnosis and scoping, based on the size of the system and the agreed targets. Infrastructure, licenses, and third-party services are charged separately.

## Frequently asked questions about WIP Optimize

### How much does it cost to optimize a system's performance?

WIP Optimize is priced after the diagnosis, because the price depends on the size of the system, where the bottlenecks are, and the agreed targets. The proposal sets out scope, targets, and price before work starts. Infrastructure, licenses, and third-party services are charged separately.

### Does WIP guarantee a cost reduction?

We don't promise a percentage before measuring. The diagnosis shows where the waste is and what gain is realistic, the targets go into the proposal, and the result is measured before and after each change.

### Does optimizing mean switching cloud providers?

Not necessarily. The gains may lie in queries, caching, sizing, or forgotten resources, without changing providers. When a switch makes sense, it is recommended with a comparison of costs and risks, because WIP is vendor-independent.

### Do I need to take the system offline to optimize it?

Usually not. Changes are made gradually, with testing and a way back. When a step requires a maintenance window, it is agreed with you in advance.

## Related services

Technical diagnosis

### [WIP Check](https://workinprogress.company/services/technical-diagnosis)

You don't know if your system is secure, if your backups work, or where to start improving it.

starting at US$ 150

Preparing for growth

### [WIP Scale](https://workinprogress.company/services/scale-readiness)

A campaign, a big client, or a launch is coming, and you don't know if your system can handle it.

quoted after assessment

Troubled software

### [WIP Rescue](https://workinprogress.company/services/software-rescue)

The system crashes, slows down, or keeps costing more, and nobody knows exactly why.

starting at US$ 470

## Start with a diagnosis

Before investing in changes, understand the risks, bottlenecks and waste in your system. WIP Check delivers an executive report and a prioritized action plan.

[Request a technical diagnosis](https://workinprogress.company/diagnosis)