Web consultancy based in Tel Aviv, Israel. About us.
Read about how we work with our clients.
Contact us, we are friendly. And we're hiring.
Read our blog.

launch review

What's it all about

When you think your application is ready to launch, or as soon as it's launched, we recommend that you take one last tiny step to ensure your success. We know that the modern approach to web applications is to "ship, then test", but often in the mayhem following shipping the "then" never actually happens. We're not talking about testing frameworks, nor about QA. This is about having a second opinion. We know your R&D team is doing its best to deliver the greatest product ever, but a second opinion always a good thing. We're sure your R&D team is doing its best to deliver the greatest product ever, but sometimes they miss something they shouldn't really have to.

We've been asked many times to perform code reviews, security reviews (with or without access to the code), deployment reviews, usability reviews, and check out many other kinds of issues before a project launch or during beta/production stages. So we came up with a service that we call Launch Review.

Why do you need this

Bugs vs Problems

We're about to say something you may not agree with, but we'll try to convince you that it's true ;-) It's actually a good thing when your application throws an exception and users see some "Error 500, we're sorry but something went wrong" message.

Sounds weird, right? Well, it's not.

This is a sign that you have a bug, and it can be fixed. Set up some sort of notification system or do whatever it takes to ensure that you are informed as early as possible about any exceptions that occur, write a regression test reproducing the exception, check whether it actually fails, and then fix the bug and make sure the test is successful.
Quite simple, right?

But what happens when you're not aware that something is not working properly in your application?

You won't get an exception if someone gains unauthorized access to your application, for example. You have a problem, and you're not aware of it. That's why we've said that bugs are good; at least you know they exist.

Give us a chance to check your application. All the projects we've reviewed so far have had security problems. Sometimes developers are simply not aware of security issues, sometimes third party libraries or plugins have security bugs, sometimes both. It's dangerous to leave these problems unsolved, especially if your application deals with sensitive private data.

Usability and graphics.

You and your team all work hard to create the best look and feel for your application. You and your team work several hours a day, every day, looking at the application's design, filling out forms and clicking links.

You know the navigation hierarchy because, well, you created it. You definitely know what needs to be done to add a product to the shopping cart, or create a new item, or whatever your application does. You know what you're going to see when you click that tab, what kind of menus will appear, and so on. Sure you know, because you've been working very hard to create this project.

But what about your users?

They will arrive at your site for the first time. Will your application behave as they expect? Will they understand what's going on when they click the Submission link? Are the tab captions readable and obvious, or will users have to work hard to understand what's going on?

Are you making your users think?

Because most of them don't really like to think about those things. And if they feel uncomfortable and lost, if you don't do the thinking for them where you can and make their life easier, they're eventually going to leave.

It doesn't actually matter what your application does if the users' expectations are not met.
Let us check this out.

What we do

Here is a brief list of things we can do to make your application better

  • Security. There are a lot of possible ways to compromise your application: XSRF , XSS , SQL injection , the list goes on and on. We try to find out if your project is vulnerable to these attacks and suggest possible solutions
  • Scalability. Scalability is not the target, it is a process that allows a swift response to a growing number of requests. Every project should have a scalability plan. The current implementation should be able to deal with a growing number of request until the next step of the plan is implemented. The next step of the plan should be known at every stage of development. We check that the current implementation is able to deal with a growing number of requests, and that the next step of the plan is theoretically implementable under the current approach.
  • Performance. Please note that scalability and performance are not the same thing. Scalability issues are about a growing of number of requests (with solutions such as adding more front ends, splitting databases, etc.), while performance is a function of the time it takes to process one request, with possible solutions such as query optimization, caching, etc.
  • Technology. We check that your chosen technologies match the project requirements and meet industry standards such as full-text search implementation, AJAX and Javascript usage, etc.
  • Code Review. We check that your code is written "the right way" (i.e., it is readable, maintainable, maintains separation of concerns, etc.)
  • Usability Assessment. We look out for usability issues, perform user testing and prepare a prioritized list of things that can be improved, together with possible solutions.
  • Graphic Design. We work with the best designers, who know a lot more than just how to make a good looking button in Photoshop. They know how to develop a graphic language for your project and focus attention on the most important information on the page, while maintaining a uniform style across all pages. They know what users expect and how to avoid disappointing them.
  • Deployment. Deployment is about sticking to a checklist. We have outlined it in a separate blog post .

Sounds interesting? Talk to us at we@astrails.com