Two of the most common questions we get also happen to be contradictory:

• How can you charge so little?
• How can you charge so much?

To be fair, depending on your perspective these are both very legitimate questions.

The short answers to these questions are:

• We can charge as little as we do because we keep our costs very low. We don’t need 95% profit margins and we’re quite happy building a small-but-efficient business.
• When you factor in the total cost of a load test (man hours, training, scripting, hardware set up), we think we are actually a great deal when compared to even Do-It-Yourself options.

To review our pricing, click here. If you’re interested in learning more, read on.

Understanding Traditional Costs

To understand, allow us to explain the traditional load testing options and their pricing. Usually there are three options people face:

• Use open source software and do it yourself
• Use commercial software and do it yourself
• Use an external load testing service and/or consultants

Each of these options has different pricing implications and associated costs – to you and to the provider.

Generally open source solutions have a low investment cost – often no money for the software and very little for hardware, especially if you can “borrow” it from your IT lab. You’ll then also need to pay for the time your personnel dedicate to setup, scripting, execution, and reporting.

The commercial software, on the other hand, have extremely high up-front fees. For example, you can buy products from Mercury/HP or Borland, but they often can cost $20K or more for even a moderate level of virtual users. They charge based on concurrent users, with 250 users quickly becoming extremely expensive. You’ll also need your own hardware and personnel costs.

Load testing services fall somewhere in the middle. You don’t need your own hardware, but you end up paying for that in the fact that the price per concurrent user per hour is, at a minimum, $3 and often much higher (we’ve seen $6-$10 quoted before). This means a 250 user test for an hour could cost $1500 or more!

Comparing Pricing to Commercial Options

We don’t believe in complex price curves or charging more because you need less right now. Our costs are mostly fixed (with the exception of customer acquisition and support), so we made our prices fixed as well. We do have one lower price point in our subscription plan, which we can afford to do when we know we have a committed customer and don’t have to spend more money finding new ones.

But what about our core infrastructure/load generation costs? Because we use real browsers, they do indeed cost significantly more than any other load testing service. And yet somehow our prices are still cheaper. That’s because we keep our overall operating costs extremely low. We also aren’t trying to become a large business that tries to employ thousands of people, so our margins can be thinner.

Of course we make money on every load test, but we aren’t greedy. By getting a larger quantity of customers we can turn even these tight margins in to a viable business.

Comparing Pricing to Open Source Options

If you compare the cost of a 200 user BrowserMob test for an hour ($200-$400) to the cost of using JMeter, we might look more expensive. And that is why we sometimes get questions about why we’re so expensive compared to Open Source Tool X.

But when you look at the total cost, we often are actually much cheaper. The primary reason is the cost to get started. Using JMeter, OpenSTA, or other open source tools is not easy. They require you to set up hardware and install agents.

Once you have the infrastructure ready, scripting is the next challenge. Because these tools don’t run real browsers (if they did you’d need a LOT more hardware), creating a good script is very difficult. You have to make sure that your cookies, AJAX headers, and HTTP parameters all get handled properly.

On the other hand, recording a BrowserMob script is painless. If you can record a Selenium script in Selenium IDE, you’re done. No need to worry about adjusting it to work in a load test.

If you consider the setup and scripting costs, we think our product ends up being comparable to Do-It-Yourself options. And for the same price, you get to rent a ton more hardware, use real browsers, get simple scripting, great error reporting, and access to expert advice.

That’s why we think BrowserMob is easily the lowest cost choice when compared to any other option: commercial, open source, or external service.

Tweet This Post 

No tags

BrowserMob officially opened for business today. While there are still several last-minute items to do (our FAQ is still not fully answered), we’ve decided that after months of polish and preparation, we’re now ready to open up BrowserMob to the public.

So what is it?

BrowserMob is the world’s first load testing service that uses real browsers (Firefox 3.0.3 to be exact). It’s on-demand, low-cost, and super easy to get started with. And best of all: you only have to pay for what you use.

What real browsers?

Our company is named BrowserMob. We named it that because we believe that as web applications get more complex, testing them with incomplete simulators does not properly prepare you for the real world.

For example, other load testing tools simulate HTTP GETs and POSTs but require painful scripting efforts to make sure that the the simulated traffic handles cookies, parameters, and other dynamic capabilities. Similarly, when something goes wrong they can’t show you what a user would have seen, since they don’t run a browser a user would use.

BrowserMob can.

By leveraging the popular Selenium project, it makes recording painless. It also generates the same traffic characteristics real users would and reports errors in a way that make sense to you: with screenshots of every individual browser error.

What about the cost?

While it’s true that using real browsers is in some cases 100 times more computationally expensive than simulating traffic, we set out to create a product that was not only better, but also more affordable.

We’ve achieved that by leveraging open source and other modern infrastructure technology, such as Amazon Web Services, to keep our overall costs low. Our goal isn’t to milk our customers for everything they have, but instead to provide a great product and build an efficient small business in the process.

What’s next?

Now that we’ve “gone public”, our attention will turn to two things: getting the word out and continually improving BrowserMob to meet the needs of our customers.

We’ll use this blog as a vehicle to communicate the progress of both of these tasks, as well as to discuss general load testing advice, industry news, and anything else that seems relevant. So please, subscribe to our RSS and contact us to continue the discussion.

Thanks, and welcome to BrowserMob!

Tweet This Post 

No tags

We’re only a day or two from officially launching BrowserMob. The only holdups are documentation and our new logo. The logo is out of our hands (we have an excellent graphic designer helping us out), but the documentation is definitely something we can wrap up on our own.

Since this blog serves as both our blog and our documentation, you can find the documentation as we wrap it up today and tomorrow right here.

Tweet This Post 

No tags

Welcome to the BrowserMob blog. BrowserMob is a new, low-cost service that offers on-demand load testing using real web browsers. This blog is where new features and announcements about BrowserMob are made.

But more than that, this blog is a place where load testing and performance testing tips and tricks are discussed, dissected, and disseminated. If you’re in the business of developing, testing, or deploying web applications that must meet today’s high expectations of richness and performance, we hope that this blog serves as a useful guide for producing high performance applications.

This blog also serves as the online documentation for BrowserMob, so please feel free to reference it for help when using the product or simply browse around if you’d like to learn more about this exciting new testing solution.

And of course, if you ever have any questions, we encourage you to leave a comment here or contact us. We love to hear from our customers, prospective customers, or anyone else who is interested in making quality software.

Tweet This Post 

No tags