CAT | ROI Case Studies
Unscheduled downtime happens all the time. In 2012 alone dozens of big-name websites went down for hours at a time, from GoDaddy.com to, well, almost anything running on Amazon. Every time an outage occurs, the Internet has a field day speculating about everything from the cause of the outage to who got fired for it. Some more transparent companies will publish a post-mortem on their blog with the details, which is nice, but there’s one question that’s almost never answered: how much did the downtime cost?
Various analysts & institutions have taken a stab at trying to answer this question with exhaustive research and surveys. The results are not very conclusive – it seems that the answer depends on the industry, the size of the company, the type of application, and even the year. We took a shot at breaking it all down for you in the infographic below: How much does downtime really cost you?
Embed this image on your site:
Summary: How much does downtime cost?
Downtime is really expensive – especially if you work at a financial services institution that relies on transaction-based fees like credit card transaction fees or trading fees. But there are costs associated with downtime that we don’t imagine – for example, the human cost of having to spend every day firefighting instead of focusing on other projects, or the damage to your company’s brand, which can impact customer loyalty. It may be worth the effort to find out what downtime costs your organization, and how much of it you can afford.
Today we announced a great milestone at AppDynamics: We have selected an application performance management (APM) solution to monitor our Drupal/WordPress website, and we decided to go with AppDynamics. In order to get a pretty sweet discount on the product we agreed to sell our souls and do a press release AND a video testimonial. Here are both of those (totally serious) marketing assets. Enjoy.
AppDynamics Selects AppDynamics to Monitor AppDynamics.com
Leading APM Company Chooses Leading APM Company to Ensure Performance of Company Web Site
Today AppDynamics announced that it selected AppDynamics to monitor the performance of its web site, AppDynamics.com. The company’s marketing team evaluated multiple potential vendors but chose AppDynamics due to its unmatched ability to identify bottlenecks and resolve problems in production web sites.
Chris Tiwald, Technical Operations Engineer at Conductor who uses AppDynamics, commented, “I don’t understand why the marketing team at AppDynamics didn’t talk to me for this press release. I’m an actual customer who loves AppDynamics, and I would have been happy to share that we’ve seen amazing uptime, performance gains, and ability to correct critical web site problems fast using their easy-to-use and deploy solution. But if they want to do the press release by themselves, whatever, I’m fine with that, I guess.”
The AppDynamics marketing team selected AppDynamics.com based on its ability to gain 100% visibility when monitoring their PHP environment, its rapid troubleshooting capability, and the solution’s overall ease of use.
“The AppDynamics.com web site drives a majority of the leads that ultimately become revenue for the company, and therefore it’s critical that we find and fix web site issues in record time,” said Greg Howard, Sr. Director of Marketing at AppDynamics. “We’ve found that AppDynamics for PHP does exactly that. Another plus was its simplicity and lack of required configuration—after all, I was a liberal arts major and I need my APM tool to be simple.”
“We evaluated other PHP solutions and found that they were hard to use, worked poorly in production, and had terrible looking dashboards,” said Stephen Burton, Director of Product Marketing and Technology Evangelism for AppDynamics. “Only AppDynamics met the requirements of the AppDynamics marketing team. And I play poker with the sales guy, so that was a plus.”
“If our solution is good enough for the likes of Netflix, Priceline, TiVo, AMICA Insurance, Hotels.com, StubHub, Staples, Insight Technologies, and Cornell University, it ought to be good enough for AppDynamics. Therefore, I’m pleased that the AppDynamics marketing team selected AppDynamics to monitor AppDynamics.com,” said Jyoti Bansal, Founder and CEO of AppDynamics. “As we continue to disrupt the market with our application management solution designed for production environments, we’re seeing companies flock to our new approach and throw out legacy vendors that are overly complex, expensive, and ill-suited for today’s highly distributed applications. Furthermore, this validates that the people I hire have very good taste.”Link to this post:
Everyone and their mother is talking about big data these days – how to manage it, how to analyze it, how to gain insight from it – but very few organizations actually have big data that they have to worry about managing or analyzing. That’s not the case for FamilySearch, the world’s largest genealogy organization. FamilySearch has 10 petabytes of census records, photographs, immigration records, etc. in its database, and its data grows every day as volunteers upload more documents. Ironically, this organization that’s tasked with cataloging our past is now on the forefront of the big data trend, as they’re being forced to find new and innovative ways to manage and scale this data.
From 2011 to 2012, FamilySearch scaled almost every aspect of their application, from data to throughput to user concurrency. According to Bob Hartley, Principal Engineer and Development Manager at FamilySearch, AppDynamics was instrumental in this project. Hartley estimates that FamilySearch saved $4.8 million over two years by using AppDynamics to optimize the application instead of scaling infrastructure. That’s a pretty big number, so we broke it down for you in this infographic:
Embed this image on your site:
How FamilySearch Scaled
- From 11,500 tpm to 122,000 tpm
- From 6,000 users per minute to 12,000 users per minute
- From 12 application releases per year to 20 application releases per year
- From 10 PB of data to approaching 20 PB of data
- No additional infrastructure
- Response time reduced from minutes to seconds
- 227 Severity-1 incidents/year took 33 hours each to troubleshoot
- 300 pre-production defects per year took 49 hours each to troubleshoot
- This amounts to a total of 36,891 man-hours spent on troubleshooting every year
FamilySearch estimates that they saved $4.8 million with AppDynamics in two years. That’s a huge number, so let’s break it down:
- FamilySearch would have had to purchase 1,200 servers at approx. $1,000 each, amounting to $1,200,215 in savings
- Those 1,200 servers would cost $2,064,370 in power and air conditioning
- Those 1,200 servers would cost $200,000 in administrative costs over two years
FamilySearch estimates that they’ve reduced troubleshooting time for both pre-production defects and production incidents by 45%, amounting to $885,170 in savings for pre-production and $460,836 in savings for production incidents (based on average salaries for those positions).
To learn more about what FamilySearch accomplished and how they use AppDynamics, check out their case study and Bob Hartley’s video interview on the FamilySearch ROI page.Link to this post:
Edmunds.com, the leading online automobile resource, is in the fast lane to achieve $1.2 Million in savings using AppDynamics to monitor their web application performance. John Martin, Senior Director of Production Engineering, opens up about his road trip through managing application performance. Martin’s operations team was able to reach goals and destinations that were not previously possible and turn “feelings into facts”, by using AppDynamics.
Driving a DevOps Culture
Martin’s goals as the leader of production engineering at Edmunds.com include maintaining a collaborative DevOps culture and keeping response times below 150 milliseconds. “Our development team is very data-driven”, said Martin, “so it is necessary that whenever an event occurs in production that the operations teams are able to feed back actual data to development.” With AppDynamics, the operations team is able to present “facts instead of feelings” to development.