Mercury Blogs

A new white paper has been added to Mercury.com:   Best Practices for Implementing Mercury Quality Center 9.

To download this whitepaper,  visit:

http://download.mercury.com/cgi-bin/portal/download/loginForm.jsp?id=235#d235

- Matt

On April 3rd, Mercury will announce the largest release of Mercury Quality Center in our history.  

Mercury Quality Center 9.0 is an integrated solution for managing the quality of business applications to ensure functionality is as the business requested..  The 9.0 release was purpose built to ensure the needs of ensuring quality in IT organizations of the next decade – addressing applications from a Risk-Control perspective.  As such, Quality Assurance departments have become the gate-keeper to ensuring that risk is controlled for new releases.

Highlights of Quality Center 9.0

Quality Assurance for Distributed Teams or Off-shored Environments

Mercury TestDirector 9, Mercury Business Process Testing 9, and Mercury Quality Dashboard 9 have been revamped to ensure Quality Assurance with distributed, global teams.    Features that support this:

-               Ensuring business test-requirements are clearly defined and leveraged, Mercury TestDirector 9’s new requirements management functionality includes more powerful definition functionality and enables traceability to more elements of testing process (including test runs, executions, and defects).  TestDirector 9 also supports new reporting mechanisms (like Live Reporting and export to Excel.)

-               Mercury Business Process Testing 9 has been revamped to support Manual Testing, reducing manual-test creation time by up to 50%,  decreasing time to execute Manual testing (through self-guided wizards), and automating the test plan documentation (through BPT’s Auto-Documentation).

-               Mercury QuickTest Professional 9 provides the industries first distributed QA team workgroup support with function library sharing features, a new XML based Object Repository, revamped “Application Area” support for Keyword Management,  and concurrent test asset management.

-               The new Mercury Quality Dashboard 9 provides new Key Performance Indicators (KPIs) for robust visibility for distributed quality initiatives.

SOA, Webservices, Web 2.0 Testing Initiatives

Mercury QuickTest Professional 9 introduces technology support for testing the next-generation of development architectures.  This includes support for testing Webservices and SOA, Macromedia FLEX,  Eclipse JRE Java applications, Firefox, and Wireless applications.    Users can now build automated tests of WSDL files, auto-generating test cases to ensure highly-leveraged webservices are functioning correctly.

Introducing Change-Impact Testing for SAP

Quality Center 9.0 introduces a new Mercury technology to ensure users can test for changes within their application.  Change-Impact Testing for SAP includes a tight integration with SAP Solution Manager that enables real-time notification of test assets which need to be run to validate hot-packs, application changes, or new releases.  Features supporting this include:

-               Mercury TestDirector 9 Requirements Module’s new integration with SAP Solution Manager flags changes and inherits SAP application definitions.

-               Mercury Business Process Testing 9 provides traceability between SAP application components and the Business Component repository within Quality Center.

-               Reporting functionality that groups testing assets which need to run (or updated) to validate changes within an SAP Instance.

Plus Much More

Quality Center 9.0 release includes all new versions of Mercury QuickTest Professional, Mercury Business Process Testing, Mercury TestDirector, and Mercury Quality Dashboard. As part of BTO Enterprise, all of these new releases are fully integrated into Mercury IT Governance and Application Management products, supporting the Mercury Application Change Lifecycle.

For more information, look for product information to be available on Monday (April 3^rd) at http://www.mercury.com/us/products/quality-center/.

-   Matt

Join us for an exclusive live seminar, and discover how our innovative Application Performance Lifecycle solution will enable you to ensure application performance at every stage of development and deployment.

During this informative session, you'll learn how Mercury Application Performance Lifecycle solutions help:  

• Mitigate risk with testing and monitoring from the end-user perspective
• Increase IT efficiency in requirements gathering and asset reuse between development and production
• Accelerate time to resolution with increase collaboration between disparate teams
• Optimize business performance

I would love to meet the folks who read this blog on a regular basic....   Register below:

http://mercury.regsvc.com/Reg/register.asp?site=61

- Matt

Welcome to 2006!

The new year always ushers in promises of new beginnings and new-year's resolutions.  In managing the business of IT, 2006 is expected to be a big year for the new paradigm of Services Oriented Architecture (SOA) systems.  The concept behind an application that leverages the SOA paradigm is that multiple systems and processes can re-use modular software components to create composite application processes.

This offers many exciting benefits:  including reducing the level of work - and re-work - associated with building redundant components of multiple applications.  It also centralizes ownership and control for specific application modules within larger IT organizations.

But with any "centralization" paradigm, SOA introduces a new level of IT and business risk into the software development equation.   Quality and Performance validation must now account for the "single-point-of-failure" concern from leveraging shared systems.  In addition, change and configuration management has become significantly more complicated.

While Mercury's Quality Center and Performance Center software suites have addressed this emerging market for sometime, you will see some exciting new features and product announcements around this paradigm. 

In fact, Mercury started the year off with a bang:  announcing recent acquisition of Systinet.   (read the press release).  WIth this acquisition, Mercury now offers a SOA governance solution which helps manage change, control, and registry of SOA systems.   Combined with Mercury's optimization centers, we offer an unparalleled solution for managing SOA systems. 

SOA is here to stay... and Mercury is aligned to ensure success. 

...watch this space!

- Matt

Does Testing equal Quality?

Have you noticed how the definition of quality has shifted in the QA Industry? 

Over the past 15 years, Mercury has been focusing on ensuring the application that business processes function and perform as expected.   In 1989, when Mercury was founded, the largely understood solution to these quality problems was “more testing.”  This mindset carried on for sometime, as the waves of testing followed technology innovation (ie:  Mainframe to Client Server, Client Server to Web, Web to .NET, etc)

Looking back over the years, the Industry has really evolved in this line of thought.  In today’s world of compliance regulation and substantially more complex software applications, Quality is more of a management paradigm.

Looking at test automation, businesses must have the flexibility to build test-cases off of design documents, early in the life cycle.  Quality Assurance departments must have the capability to report, in real time, the status of the quality of business processes which are being implemented. 

In other words, traditional record/replay tools can not scale to the requirements of modern Quality Assurance.  These previous generation of solutions do not enable test creation early in the software-development-lifecycle and the programatical scripts are “separate and distinct” from design specification.

Looking beyond test automation, this new quality paradigm requires new thinking in terms of test management, quality control, as well as the overall QA planning process. 

On this topic,  there will be some very exciting sessions at Mercury World next week.   

In count-down to Mercury World 2005,   I will have frequent blog entries this week covering the sessions that I am looking forward to, and a couple of interesting technology previews that Mercury will be conducting.

I hope to see you at the show!

- Matt

We have a customer, a very large medical equipment manufacturer, who has standardized on Oracle ERP suite for Financials, Manufacturing and Order Management. Most CEO's and CFO's will tell you that the ability to "close the books" or make quarterly forecasts is probably one of the most important things for the business. (For most of us balancing checkbooks takes 7 days, so imagine the magnitude of the problem for this large enterprise).

Like most large enterprises, the last few days of the quarter are absolutely critical to help bring together the key elements of accounts - payable, receivable, general ledger all together. This customer had a particular backlog of orders (orders too close to the last few days of the quarter). The accounting personnel depended on the performance of the Oracle Financials 11i system to ensure they could perform their business processes on time and schedule to report the numbers to the executives.

Turns out the Oracle application was tested and designed for peak load but not for the volume of transactions that suddenly spiked up due the back end loaded nature of the quarter. The system would take normally 4 seconds to respond to queries and return results, now suddenly took over 29 seconds. Orders got batched up and the Order entry personnel were had a long backlog.

End result: Several transactions could not be completed before the specified time which delayed filing earnings for the quarter by 2 days.

The CFO and CIO put together a task force to ensure this problem was never replicated again. They deployed User monitoring and put together a "War Room" that monitors the performance for the users of the Financial system during the last 3 weeks of the quarter. The "War Room" comprised of several personnel from the various Operations teams with the skills and capability to resolve performance problems fast.

Over the last few years they have been doing this diligently each quarter and never missed a beat.

I on the other hand have never balanced my checkbook in years. My wife claims its because of the lack of funds to balance.

The Accelerators for Business Process Testing are on fire.

Last week, I had a follow-up conversation to one of our Business Process Testing customers in the West who have completed a successful patch upgrade to his Oracle 11i implementation. This customer had been trying to automate the testing of each-and-every patch cycle for the past two years, but simply could not get the resources and time to build the automated tests.

To quote: "Automation is great. Everyone knows this. Our problem is that the patch cycle was too frequest for us to build automated test cases... we simply did not have enough time. It makes us very nervous, as our business depends on our Oracle systems."

The Business Process Testing accelerators helped change the game for this customer. With the Business Process Testing Accelerator for Oracle 11i, there is no automated test creation from scratch. Instead, an entire suite of completed test cases come pre-packaged in terms of the automated testing components, the data management, and the business process flows.

Since Business Process Testing features auto-documentation, the test-plan documents are auto-generated by the system.

Available with a professional services engagement for the customization of the accelerator, the BPT Accelerator for Oracle dramatically reduces the implementation cost and overhead of test automation. For more information on the Business Process Testing Accelerator for Oracle applications, visit:

http://www.mercury.com/us/solutions/erp-crm/oracle/business-process-testing-accelerator/

 - Matt

We have a customer that is a State Penitentiary. Before you read any further, I want to dispel any rumours that I actually visited their site.

They do have an administrative office I visited over 2 months ago. They had a unique business (social) problem. Many of their inmates were from other countries. For visitations to these inmates that had to be scheduled, they developed a web based reservation system that was integrated with the scheduling system.

Since most of the relatives of these inmates were from other countries, after US VISIT program (post 9/11) they had to obtain daily visas to visit them. These visas would only be valid for the day of their visit to the penitentiary.

The reservation system was performing very poorly and often experienced outages. Since each relative had to obtain a reservation to apply for a visa, they would come and wait for days for the application to perform and access available schedules.

The IT organization located multiple End User Monitoring transactions using our Managed services (Global Points of Presence) to get a perspective of the "transaction breakdown" for each request. With this information they were able to identify, isolate, triage and diagnose the problem resulting in outage time reduction by 30%.

Visitors were able to schedule quicker, obtain their visas and visit inmates reducing backlog for other visas and saving time and resources for the Penitentiary.

The new release of Business Process Testing is making waves in the industry.  Yesterday, I received the following email from a Quality Assurance Manager of a large shipping company.

"We have been looking for a method to convert existing WinRunner scripts into QuickTest Professional assets for the last three months, and have evaluated a couple of tools on the market to do this.  I wanted to let you know that with this new release of Businesss Process Testing, this is NO LONGER AN ISSUE!

Without the expense of purchasing another tool, hiring a consultant, or taking on a massive internal project, we are using Business Process Testing to re-use our existing WinRunner scripts as Business Components.  We have now built a combined test framework of both WinRunner and QuickTest Professional assets, which allows us to test practically every environment from one solution.  It really works great guys."

What great news! 

When we planned the Business Process Testing 8.2.1 release, one of the top objectives were to give our existing WinRunner installed base the ability to re-use ANY existing WinRunner asset (scripts, functions, GUI-Maps) in the new system. 

For more information on using WinRunner assets in Business Process Testing, visit

Business Process Testing for WinRunner Users

- Matt

The news: Mercury acquired BeatBox (ClickCadence) for $14Million. BeatBox has over 100 customers thought OEM partnerships.

Why: Customers have mentioned 3 things to us consistently:

1. There are reasons why the need real user experiences (to complement their synthetic user monitoring) such as to diagnose problems for users experiencing them or compliance initiatives require them to avoid putting synthetic transactions in their enterprise systems. Example: One of our customers is a pharma giant, and for them HIPAA rules state they cannot create dummy purchase transactions even if they are deleted later.

2. The real user behavior can be captured, and it helps provide their testing them with real world scenarios on what customers are doing with the applications. This prevents the entire blame-game around why QA is testing what no one is really doing  Example: A trading exchange customer found out that QA was testing order fulfillment and order entry while 80% of real customers were just looking for stock quotes.

3. Customers want to able to automatically identify and track every unique visitor on their application.  This helps session management and Geo location. Example: One of our customers is a bank, which cannot (because of Patriot Act) allow transactions to be initiated from certain countries.

The product will be integrated with Business Availability Center and generally available in Q1 2006. Early access is happening right now!

Its here! SiteScope 8.0 is available to download!

http://download.mercury.com/cgi-bin/portal/download/loginForm.jsp?id=57#d57.

1. Better User Interface to reduce time and improve administration.

2. Global Search and Replace to configure and lower the time to make changes

3. New XML (BTW this is not my shirt size ) metrics monitor to help you get any XML feed into SiteScope.

4. Ability to create custom views and filter in the dashboard, so you can understand where problems are quicker.

In talking with customers and sales alike, we get the impression that Product Management makes decisions of product features purely in a vaccum. Here are the 3 most important strategies that are adopted by us during the process of feature and capability prioritization.

First we collect requirements from analyst firms (Gartner, Forrester, etc.) followed by detailed discussions with customers (the top 50-100 customers are polled frequently). Input from other customers not in the top 100 are usually  collected from our customer support and managed services teams. The sales and sales engineering teams are consulted next. After detailed reviews with executives on the strategic goal of the product, business and market metrics, there is a list of features that are collected and put into a product planning system we use internally.

Second all these features and capabilities are sorted, categorized, collated and deleted  (kidding on the deleted).

Finally there are 3 different approaches adopted by 3 of our Application Management Product team members. I highlight the 3 because they are indicative of the rest of the team, but these three do it with great style.

A) Ramin is a fan of the roshambo method. He spends enormous time with his son Carter (who is all of 5 weeks old) to get his input on the strategic focus required and also the features that he drools the most on.

B) Chris comes from a sales background and prefers good ole' cash bribes. If you are flush with some pieces of paper with dead presidents on them, there's all the reason to get the exact features you want. Forget strategy, the only goal in this method is cash flow maximization.

C) Finally Tim is a fan of the craps table. Odds are that he does not even care about the roll of the dice. He just wants to play and get a pair of sixes. If you beat him 4 rolls in a row you are never getting that feature built-in. So beware.

There you have it: scientific and measurable, creative and process driven. sophisticated yet simple.

These are our ways to get ahead of the feature request.

Tom Kucharvy, an industry analyst with Summit Strategies, talked about the game-changing nature of on-Demand software. Here are his requirements.

1. Black Box Effect: Provides services from a closed, shared infrastructure.
2. Stable Platform Model: Allows developers to write to a stable, standardized Internet platform, rather than to evolving "instances" of operating systems or application environments.
3. Real-Time Delivery: Enables a continuous feedback loop that puts on demand vendors constantly in touch with how their customers use their products.
4. Rapid Increment Adoption Model: Enables customers to start small and then easily step up...
5. Pay-as-You-Go-Model.
6. Continuous Configurability of On Demand Services: Provides some of the customization capabilities of on-remises software without relying on custom code.
7. Process Ownership Delivered to Users: Puts them in charge of business process automation.
8. Pluggable Integration: Enabled by a services architecture...

Mercury Managed Services (MMS) has been operational for the last 5 years as the leading on Demand operational service for Production applications. Over 500 customers depend on Mercury to manage, deploy and run their monitoring in an outsourced environment. In talking with customers here are the reasons they love the service:

1. Best practice learning and implementation: Since MMS has been working with 100's of customers, they know how production applications need to be managed across industries and have learned the best ways to do that from several customers.

2. Consistent and ongoing mentoring and training: Rather than attend annual classes to learn about the software, adminstration, monitoring their applications, they learn consistently by means of the Technical Account Managers (TAM).

3. Pre deployed and ready to use infrastructure: Saves time and gets the solution up and running in 20 days. Then customers can customize the solution to their needs. Saves waiting for 30 days for an order of 4 way CPU boxes to arrive and be provisioned.

4. Best answer for lack of resources internally: Show me one IT organization with too many resources and I'll show you the moon with blue cheese.

5. Full service at pay-as-you-go prices: Pay as you go - what a concept, but almost everyone expects to give up something in order to get it. With MMS you get the price break but with complete scripting, management, mentoring, support and service.

No wonder its one of the fastest growing portions of our Application Management business.

I was having lunch with a System Monitoring Manager the other day of a Large Insurance company ( a long term Mercury customer). The topic turned to competition. I was telling him how competitive the software industry is and the bare knuckle tactics that are adopted. After all terms such as Competitive Silver Bullet, Knockoffs, Laying traps, throwing FUD are all standard part of the repertoire.

He was smiled and said the that the competitive spirit is across industries not just software. Its just your tactics are more "new age" almost at times not as sophisticated as the more mature industries. Curious, I asked him to elaborate.

His company like most sells insurance by means of a broker dealer network, has some of their own employees selling their insurance and also wholesale agents. A decade ago, insurance agents would get price quotes either via phone or submitting faxes to the company based on the customer's needs. The company after a series of acturial calculations and a loan origination process give them back the estimate quote. This process for the end consumer would take nearly 2-3 days.

With the advent of the Internet, and instant 15 min quotes on websites, the industry obviously gets more competitive. 3rd party brokers (since they are not paid anything but comissions on their sales) have a great choice of offerings when their end consumer has insurance requirements.

The company found that for 2 months in a row the 3rd party quote requests dropped significantly enough to undertake a detailed review. Turns out since time is of essence for the broker, they would access the application from the company that was the easiest to use and the quickest for quotes. Most times since they either had a customer on the phone or with them in their office, they wanted to be as reponsive as possible. Since the insurance company's quoting tool was performing poorly (some quotes took over 120 seconds), they chose other (competitors) insurance companies.

So our customer deployed the end user managment on various locations within 3rd party broker's locations and tried to localize and isolate the problem. Having verifiable and repetable information about performance, time of day, locations, etc. they were able to make an informed decision to help defray the costs of high speed internet connections for a subset of their "million dollar club" brokers.

Quotes went back up, brokers were pleased and their key competitors were neutralized. All in 2 months. Speed is a wonderful thing.

We have a customer who is a Cruise company. I actually went on one of their cruises. Awesome experience and highly recommend it - pure decedance, everything at your beck and call, great service and the food - easiest and quickest way to gain 10 pounds in 2 days.  I firmly believe that the person that invented cruises is the same person who founded La-Z-Boy chairs.

Back to my point.

Our customer has multiple cruise ships. I learned from them that each ship really is a small business. Usually have over 800-1000 staff on board, and each is a datacenter with application such as payroll, reservations, casino management, Point of Sales terminals etc. The ship's data center connects via satellite link to the shore unit and the command center.

They contacted us since they were having problems with application performance for a set of cruise ships. Most ships completed transactions at all hours of the day but uploading of the transactions to the mainframe backend systems was done only at specific times as a batch process using a controlled UI by one of the ship's admins on board.

Point of sales transactions meant closing business for the day so these were mission critical. Trouble was some of the ships were having trouble uploading them in their allocated time window. The IT Operations team checked the batch logs and figured it was because of processing speeds for the application server were insufficient. So they upgraded the machine to a four bay, 8 CPU machine costing them $400,000.

The problem still was not solved.

They implemented End User Monitoring from 3 of the ships for 5 days. It gave them the perspective of the transactions from those machines and the transaction breakdown of the time it took the transactions to go from the User interface through the satellite link to the application server all the way to the database.

They isolated the problem to the satellite link NOT on the batch processing server.

Question still remained why some ships had the satellite link problems while others did not. With location analysis they found out that at the time some ships were trying to upload they were at locations on the sea where the satellite link had the weakest connection.

They changed the time of upload for a few ships to solve the problem and ensure that they still uploaded the transactions and ensured POS data was collected.

And let the vacationers worry about those few extra pounds!

During one of the meetings with a customer who was the Director of Operations at a Regional bank in North East, I was trying to get a sense for

a) Why are there so many banks in the US. I knew there were differences in Investment banks, Majors, Regionals and Credit unions, but the # of banks is just astonishing.

b) A Bank is a Bank is a Bank. What's the difference? They all keep my money and offer mostly the same services.

Turns out I got a better understanding of banking sub segments and also the value of Application Management for the Banking industry.

The first is that most of the banks subsegment and target different types of customers - consumers, prime consumers (better credit history), small business, medium corporations, industry specific banks etc.

This bank differentiated itself on primarily catering to small businesses. So their key differentiation was to focus on small business services such as loans, collection services, credit-card processing, lines of credit and consolidated billing and management. Most of their customers were companies with 5-25 employees.

Their customers kept a very tight look at the daily cash flow. Most of their customers would go online at between 3-5 pm and make money transfers for payment services to ensure their prime money segments were being paid on full. So the availability of the online funds transfer application at this time was absolutely critical, else it meant loss of thousands of dollars for their customers. These customers would log into the application, check their prime checking and make the daily cash flow transfer at days end to meet daily obligations for payment or collections.

The Bank IT operations team was initially measured on SLA of availability of systems, but after discussing with the business owners, they adopted End to End Monitoring to get a sense for the availability and performance from a customer (user) standpoint. This meant they were going to be measured on SLA that mattered to the business and ultimately their customers.

If a Finance head from a small business got online and had a problem they could identify them, quickly find their problem and respond quickly to address their issue.

That was a HUGE difference for their customers. They increased assets under management from $36 Billion to $50 Billion in less than a year and targetted the small medium customers with this unique ad campaign - "cash transfers done or you dont pay any fees for the month". Customer retention was up by 15% and they increased new customer intake by 20%.

Big difference for a small regional bank and a great story of the direct impact that IT Operations has on the business.

Customer stories of their success with Mercury Applicalition Management are always a great thing to write about. The best is if the story has a "human touch" unlike just an IT value angle to it. The kind I can tell my mom and not get a glazed over look with a "You want some more ketchup with your omlette?"

I had a chance to meet with some customers in Southern Europe, couple of months ago. One of our customers is a small regional airline. Their key goal is to emulate JetBlue and SouthWest here in United States and so "on-time" arrivals and departures are a key for their customer satisfaction.

On average their turnaround - "time when a plane comes in to the time its ready to take off again" is about 24 minutes (excluding passenger onboarding). Their stated business goal was to reduce that time to under 20 minutes. Which they translated into 2.4Million Euros each year of savings. Quicker turnarounds meant more time for the planes on the air, making money.

There are close to 28 people involved in each plane turnaround - from the gate agent, the mechanic, baggage handlers, cleaners, food stocking specialists, gate bridge co-ordinator, etc. The turnaround involves a smooth and finely tuned concert of all these people focused on their individual tasks to meet the turnaround time.

Each of them a computer system (applications) that gives them the details of the plane - arrival, expected # of bags, pilot instructions for the mechanic on issues they noticed, gate bridge time notice, etc.

The performance of these applications is obviously very critical to their keeping their "on-time record". So every minute or two we can help shave from them spending time in front of the computer instead of doing their work, is crucial.

Finding out that some applications took nearly 90-120 seconds to respond the IT staff implemented End User Management to get a user perspective on the applications. Finding out the problems that affected the mechanics looking up part order numbers or gate agents looking up flight schedule and travel logs, before they affected them shaved nearly 3 minutes from their response time for the users.

IT was able to deliver on the business promise by reducing the time these personnel spent in front of the computer and reduced turnaround time by 4 minutes.

Keeps planes flying on time, and irate passengers from complaining. Most importantly keeps the airline profitable - What a concept in these times.