SlideShare a Scribd company logo
1 of 85
Download to read offline
Convince Your CFO that #PerfMatters
@ColinBendell
©2015 AKAMAI | FASTER FORWARDTM
©2015 AKAMAI | FASTER FORWARDTM
• Convincing the Business (General)
• Save you $100k in 10 minutes
• COGS & CapEx Fin. Modeling
• Revenue Modeling
©2015 AKAMAI | FASTER FORWARDTM
Nobody likes waiting…
©2015 AKAMAI | FASTER FORWARDTM
Source: Equation Research on behalf of Compuware
©2015 AKAMAI | FASTER FORWARDTM
Source: Equation Research on behalf of Compuware
©2015 AKAMAI | FASTER FORWARDTM
Source: Akamai
©2015 AKAMAI | FASTER FORWARDTM
But does Perf. Investment
Business Sense?
©2015 AKAMAI | FASTER FORWARDTM
©2015 AKAMAI | FASTER FORWARDTM
• Use existing business metrics:
• Revenue
• Conversion Rate
• Don't forget in-direct metrics:
• Engagement
• Successful Logins
Lara Hogan
http://larahogan.me
Run A/B Tests; Compare results
©2015 AKAMAI | FASTER FORWARDTM
Create Pride: Show your users' experience
©2015 AKAMAI | FASTER FORWARDTM
Compare Yourself with Your Competition
Andy Davies
http://andydavies.me/
©2015 AKAMAI | FASTER FORWARDTM
©2015 AKAMAI | FASTER FORWARDTM
©2015 AKAMAI | FASTER FORWARDTM
Other Business'
Success Stories
©2015 AKAMAI | FASTER FORWARDTM
Performance Impacts Business Results
Source: Walmart, Torbit, Aberdeen Group
©2015 AKAMAI | FASTER FORWARDTM
Speed Matters: Conversion Rate vs. Load Time
Conversion  Rate  (%)
0-­1 1-­2 2-­3 3-­4 4-­5 5-­6 6-­7 7-­8 8-­9 9-­10 10-­11 11+
Population Conversion  Rate  (%)
Page  Load  Time  (Sec)
©2015 AKAMAI | FASTER FORWARDTM
Performance Impacts Engagement
Josh Fraser
@joshfraser
©2015 AKAMAI | FASTER FORWARDTM
2 Seconds Faster Boosted Conversions By 66%
http://www.conversionconference.com/blog/case-study-how-making-pages-2-seconds-faster-boosted-conversions-by-66/
Tammy Everts
@tameverts
©2015 AKAMAI | FASTER FORWARDTM
©2015 AKAMAI | FASTER FORWARDTM
©2015 AKAMAI | FASTER FORWARDTM
Spreadsheets:
the Language of the CFO
©2015 AKAMAI | FASTER FORWARDTM
©2015 AKAMAI | FASTER FORWARDTM
10 minute
MBA
©2015 AKAMAI | FASTER FORWARDTM
Money buys beer
©2015 AKAMAI | FASTER FORWARDTM
How do I get [more] Money?
©2015 AKAMAI | FASTER FORWARDTM
Long Answer: GDP
+ ++
GDP = C + I + G + (X-M)
©2015 AKAMAI | FASTER FORWARDTM
Short Answer:
Your Company Earnes
More Money
©2015 AKAMAI | FASTER FORWARDTM
Stock Price is What Others Think is the Company
Value
Company_Value =  Stock_Price *  Shares
©2015 AKAMAI | FASTER FORWARDTM
Discount Cash Flow // Common Valuation Technique1
𝐶𝑜𝑚𝑝𝑎𝑛𝑦_𝑉𝑎𝑙𝑢𝑒 =
𝐶𝑎𝑠ℎ_𝐹𝑙𝑜𝑤2
1 + 𝑊𝑒𝑖𝑔ℎ𝑡𝑒𝑑_𝐴𝑣𝑒𝑟𝑎𝑔𝑒_𝐶𝑜𝑠𝑡_𝐶𝑎𝑝𝑖𝑡𝑎𝑙 2
+
𝐶𝐹=
1 + 𝑊𝐴𝐶𝐶 =
+ ⋯ +
𝐶𝐹?
1 + 𝑊𝐴𝐶𝐶 ?
+
1This  is  one  of  many  techniques.  It  is  has  *many*  flaws.
©2015 AKAMAI | FASTER FORWARDTM
Discount Cash Flow //a common valuation technique
𝐶𝑜𝑚𝑝𝑎𝑛𝑦_𝑉𝑎𝑙𝑢𝑒 =
𝐶𝑎𝑠ℎ_𝐹𝑙𝑜𝑤2
1 + 𝑊𝑒𝑖𝑔ℎ𝑡𝑒𝑑_𝐴𝑣𝑒𝑟𝑎𝑔𝑒_𝐶𝑜𝑠𝑡_𝐶𝑎𝑝𝑖𝑡𝑎𝑙 2
+
𝐶𝐹=
1 + 𝑊𝐴𝐶𝐶 =
+ ⋯ +
𝐶𝐹?
1 + 𝑊𝐴𝐶𝐶 ?
+
©2015 AKAMAI | FASTER FORWARDTM
Profit & Loss Statement (or Income Statement)
How  Much  Beer  You  Sold
Cost  to  Make  Beer
Management
Savings  for  more  distilleries
Your  Money
©2015 AKAMAI | FASTER FORWARDTM
Three Big Leavers to Increase Profits
How  Much  Beer  You  Sold
Cost  to  Make  Beer
Savings  for  more  distilleries
Your  Money
©2015 AKAMAI | FASTER FORWARDTM
More Money. More Beer.
©2015 AKAMAI | FASTER FORWARDTM
We are at a Performance Conference…
… how does performance impact your P&L?
©2015 AKAMAI | FASTER FORWARDTM
Velocity SC 2015: Financial Impact
How LinkedIn used RUM to drive optimizations and make the site faster
Thursday @ 13:45pm - Ritesh Maheshwari (LinkedIn)
What are third-party components doing to your site?
Thursday 14:40 - Andy Davies & Simon Hearne (NCC Group)
Reaching everyone
Friday 11:50 - Tim Kadlec (Akamai)
How Go is making us faster
Friday @14:40 - Wilfried Schobeiri (MediaMath
PHP 7
Friday @16:10 - Rasmus Lerdorf (Etsy)
Speeding up the PlayStation 4 user interface
- web performance optimization outside the browser
Friday 17:05 - JP CASTRO (Sony)
Yesterday's perf best-practices are today's HTTP/2 anti-patterns
Friday 17:05 - Ilya Grigorik (Google)
Revenue    COGS    CapEx
✔ ✔ ✔
✔
✔ ✔
✔ ✔
✔ ✔
✔
✔ ✔
©2015 AKAMAI | FASTER FORWARDTM
Mapping performance to the Income Statement
Revenue    COGS    CapEx
Reduce bytes ✔ ✔ ✔
Reduce requests ✔ ✔ ✔
Expedite rendering ✔
Reduce time-to-last-byte ✔ ✔ ✔
©2015 AKAMAI | FASTER FORWARDTM
Financial Modeling
Performance
©2015 AKAMAI | FASTER FORWARDTM
What is Financial Modeling?
• Excel! Yay!
• Technique to build confidence in decision making
• Consistent comparison for project planning outcomes
• Good: consistent IRR comparison
• Bad: ignores Goodwill (Brand) and Job satisfaction (Culture)
• Two schools of thought:
• Precision: Consider every detail
• Directionality: Consider the big inputs
©2015 AKAMAI | FASTER FORWARDTM
ProTip 1: Avoid the wormhole of
Finance Precision. Use Directionality
©2015 AKAMAI | FASTER FORWARDTM
ProTip 2: Use conservative shortcuts
©2015 AKAMAI | FASTER FORWARDTM
Model 1:
CapEx / OpEx CashFlow
©2015 AKAMAI | FASTER FORWARDTM
Build-More or Optimize
©2015 AKAMAI | FASTER FORWARDTM
CashFlow Opportunities: Requests, Bytes, TTLB
Server Time
Handshake
©2015 AKAMAI | FASTER FORWARDTM
Pros & Cons of CashFlow Modeling
Use this model:
• When you are optimizing backend processes
• Adding caching layers (back-end, cdn, client)
• Optimizing user workflows
• Optimizing page layout
Don't use this model:
• Optimizing client compute (eg: javascript)
• Optimizing client rendering (eg: adding webp)
©2015 AKAMAI | FASTER FORWARDTM
Cash_Flow =   Capital_Exenses
+  Operational_Expenses
©2015 AKAMAI | FASTER FORWARDTM
Typical Year-over-Year Spend
©2015 AKAMAI | FASTER FORWARDTM
Questions
• Will your improvements slow the rate of new hardware purchased?
• Will it reduce the operating costs?
• Will you be able to turn down servers?
• What are the projected costs with and without improved performance?
©2015 AKAMAI | FASTER FORWARDTM
Avoid including "Funny Numbers"
Don't include:
• Productivity gains
• Revenue from selling old hardware
• Contract [re]negotiation time
• Software Licensing savings
This will allow you to bypass a 7 week discussion with your procurement
about the true cost of your enterprise agreement.
©2015 AKAMAI | FASTER FORWARDTM
This company tried computing the
true cost of the enterprise agreement.
©2015 AKAMAI | FASTER FORWARDTM
Two Models
• CoLo / OnPrem
• This years Capital Expense contributes to next years' Operating Expense
• You can theoretically extend the life of purchased capital
• IaaS / PaaS
• Everything is Operational Expenses
©2015 AKAMAI | FASTER FORWARDTM
Operating_Exepnse =  Number  of  Servers  *  Usage
OR
Capital_Expense
=  Number_of_New_Servers *  Average_Server_Cost
Operating_Exepnse
=  Number_of_Servers *  Average_Server_KVA
*  CoLo_Cost_per_KVA
©2015 AKAMAI | FASTER FORWARDTM
Calculating Capital Expenses (CapEx)
CapEx =  Number_of_Servers *  Average_Cost
• Use "Fully Loaded" cost (incl: cost to procure, security audits, colo
service tickets, rack & stack fees, etc)
• Use lowest-common-denominator to save inventorying your
infrastructure
• If in doubt:
$5,000
$100,000
©2015 AKAMAI | FASTER FORWARDTM
A Note about Virtual Machines
• Virtual Servers also require Capital
• Don't forget to include (vMotion) buffer
• Be consistent in calculation
©2015 AKAMAI | FASTER FORWARDTM
Plugging in the Numbers
https://goo.gl/6OHj9a
©2015 AKAMAI | FASTER FORWARDTM
Calculating Operating Expenses (OpEx)
Best Practice:
• Include all servers, regardless of role
• Assume that increase in user demand = increase in infrastructure use
Data-Center OpEx
• Most Co-Lo providers bill based on power usage
• Assume Co-Lo pricing includes cooling, floor space, and internet
• Each datacenter is a special snowflake: be conservative in your
calculations
©2015 AKAMAI | FASTER FORWARDTM
OpEx
OpEx =  Number_of_Servers *  KVA_per_Server *  KVA_Price
• Pick one hardware profile that is representative
• If in doubt, use the newest hardware’s power consumption
• Use 80% of the manufacturers reported fully loaded power draw
• Most hardware reports power as Watts and BTU. Assume a power-
factor of 0.9 and use:
• KVA  =  Watts  /  900
• When in doubt I use these approximation numbers:
• 0.5KVA for an average pizza box server
• 3.5KVA for a 6U high density compute chassis
©2015 AKAMAI | FASTER FORWARDTM
Colo's charge by electricity
OpEx =  Number_of_Servers *  KVA_per_Server *  KVA_Price
• Most CoLos bundle amenities and charge by electricity
• Ranges from $0.70/KVA/mo to $0.20/KVA/mo
• Running your own datacenter =~ colo space
• If in doubt, a conservative, after tax value: $0.50/KVA/mo
©2015 AKAMAI | FASTER FORWARDTM
©2015 AKAMAI | FASTER FORWARDTM
Plugging in the Numbers
https://goo.gl/6OHj9a
©2015 AKAMAI | FASTER FORWARDTM
Example
• 140 “pizza box” type servers ($5k, 0.5KVA)
• 400 Page View/s peak
• 30% YoY growth rate
• Has a problem with personalization tracking
©2015 AKAMAI | FASTER FORWARDTM
Before
©2015 AKAMAI | FASTER FORWARDTM
After:
©2015 AKAMAI | FASTER FORWARDTM
Ignore the size of the OpEx budget
Focus on user performance budget
©2015 AKAMAI | FASTER FORWARDTM
PageView and Interaction Cost
• A Page View requests that return Content-­‐Type:  text/html.
• Page View is not perfect: use a metric that represents user activity
• Conversions
• Checkout-Rate
• Logins
©2015 AKAMAI | FASTER FORWARDTM
Page View is a proxy for 'user interaction'
©2015 AKAMAI | FASTER FORWARDTM
Interaction Cost
Interaction_Cost =  (OpEx +  CapEx)  /  Page_View_per_Second
• Max page view per second tells us how much money is spent to
maintain this peak traffic
• You have built implicitly to meet peak traffic
• As user demand grows, you will build more hardware in lock step
• Interaction Cost to remain constant year over year
©2015 AKAMAI | FASTER FORWARDTM
Before (with $/PVs)
©2015 AKAMAI | FASTER FORWARDTM
After (with $/PVs)
©2015 AKAMAI | FASTER FORWARDTM
What about IaaS / PaaS?
• Model still applies with a few tweaks; ignore Capital costs
• Each IaaS / PaaS provider has different pricing levers
• Focus on cost-per-interaction model
©2015 AKAMAI | FASTER FORWARDTM
Please, tell me more
about your made-up
financial model
©2015 AKAMAI | FASTER FORWARDTM
User Performance Improvement (RUM Data)
©2015 AKAMAI | FASTER FORWARDTM
Origin Bandwidth Reduced to Below 100mbps
©2015 AKAMAI | FASTER FORWARDTM
Reduce Server Utilization
$-
$10,000
$20,000
$30,000
$40,000
$50,000
$60,000
$70,000
$80,000
$90,000
CapEx…
OpEx
Assumptions:  5yr  Amortization;;  0.5KVA/server;;  $0.50/VA/mo
$200,000
savings  over  5yrs
©2015 AKAMAI | FASTER FORWARDTM
Cashflow Summary
• Backend performance can impact end-user performance
• Performance can have financial impact
• Use CapEx and OpEx modeling to show directionality of impact
• Avoid the detailed nuances – use conservative generalizations
• You can use the same model to project negative financial impact (eg:
image responsive images)
©2015 AKAMAI | FASTER FORWARDTM
Revenue
Probabilistic Modeling
©2015 AKAMAI | FASTER FORWARDTM
Revenue Modeling
• Your Operations teams can control CapEx and OpEx spending
What about Revenue?
• How can performance improve total sales?
• How performance increases user growth (CAGR)
• How performance increases sales per user (ARPU)
©2015 AKAMAI | FASTER FORWARDTM
Revenue Modeling is Hard
©2015 AKAMAI | FASTER FORWARDTM
Applying Monte-Carlo Simulation
• Apply probability based input on an equation
• Run many times
• Evaluate the results
©2015 AKAMAI | FASTER FORWARDTM
Plugging in the Numbers
https://goo.gl/6OHj9a
©2015 AKAMAI | FASTER FORWARDTM
Notes about Revenue Modeling
• Use Revenue Modeling to show directionality
• Avoid absolute claims about performance
• Consider regression analysis to understand what probability inputs
• Know your user performance distribution first.
(You are using Real-User-Monitoring, right?)
©2015 AKAMAI | FASTER FORWARDTM
So yea, you're going to have to come in to
work over the weekend now…
©2015 AKAMAI | FASTER FORWARDTM
Summary
• Performance correlates to engagement
• Make performance part of your culture
• Understand the user experience and leverage empathy
• Project cash flow savings from performance
• Applies to OnPrem and IaaS/PaaS
• Project revenue impact from performance
©2015 AKAMAI | FASTER FORWARDTM
Performance can make happy users
and makes business sense!
Thank-You!
@ColinBendell

More Related Content

What's hot

Akamai: From Theory to Practice
Akamai: From Theory to PracticeAkamai: From Theory to Practice
Akamai: From Theory to PracticeLiz Bradley
 
Responsive Web Demo with Akamai
Responsive Web Demo with AkamaiResponsive Web Demo with Akamai
Responsive Web Demo with AkamaiFran Albaladejo
 
Asynchronous Programming
Asynchronous ProgrammingAsynchronous Programming
Asynchronous ProgrammingLearnNowOnline
 
Heartland Ovation Payroll
Heartland Ovation PayrollHeartland Ovation Payroll
Heartland Ovation PayrollRichard Dunn
 
Runtime Verification for the Web (RV 2010 Tutorial)
Runtime Verification for the Web (RV 2010 Tutorial)Runtime Verification for the Web (RV 2010 Tutorial)
Runtime Verification for the Web (RV 2010 Tutorial)Sylvain Hallé
 

What's hot (7)

Ap workflow what is it
Ap workflow   what is itAp workflow   what is it
Ap workflow what is it
 
Akamai: From Theory to Practice
Akamai: From Theory to PracticeAkamai: From Theory to Practice
Akamai: From Theory to Practice
 
Responsive Web Demo with Akamai
Responsive Web Demo with AkamaiResponsive Web Demo with Akamai
Responsive Web Demo with Akamai
 
Asynchronous Programming
Asynchronous ProgrammingAsynchronous Programming
Asynchronous Programming
 
Heartland Ovation Payroll
Heartland Ovation PayrollHeartland Ovation Payroll
Heartland Ovation Payroll
 
BrightPay: Employer duties under Auto Enrolment for Payroll Bureaus
BrightPay: Employer duties under Auto Enrolment for Payroll BureausBrightPay: Employer duties under Auto Enrolment for Payroll Bureaus
BrightPay: Employer duties under Auto Enrolment for Payroll Bureaus
 
Runtime Verification for the Web (RV 2010 Tutorial)
Runtime Verification for the Web (RV 2010 Tutorial)Runtime Verification for the Web (RV 2010 Tutorial)
Runtime Verification for the Web (RV 2010 Tutorial)
 

Similar to 2015 Velocity SC: Convince your CFO that #perfmatters

Advanced Caching Concepts @ Velocity NY 2015
Advanced Caching Concepts @ Velocity NY 2015Advanced Caching Concepts @ Velocity NY 2015
Advanced Caching Concepts @ Velocity NY 2015Rakesh Chaudhary
 
Con1169 office depot scm cloud mbx session
Con1169 office depot scm cloud mbx sessionCon1169 office depot scm cloud mbx session
Con1169 office depot scm cloud mbx sessionBharti Tiwary, PMP
 
Dev348 ReInvent Corteva Agriscience
Dev348   ReInvent Corteva AgriscienceDev348   ReInvent Corteva Agriscience
Dev348 ReInvent Corteva AgriscienceRandy Black
 
Sap Product Lifecycle Costing
Sap Product Lifecycle CostingSap Product Lifecycle Costing
Sap Product Lifecycle CostingJulien Delvat
 
CA Workload Automation ESP: Tips and Tricks for Getting the Most Out of Your...
CA Workload Automation ESP: Tips and Tricks for Getting the Most Out of Your...CA Workload Automation ESP: Tips and Tricks for Getting the Most Out of Your...
CA Workload Automation ESP: Tips and Tricks for Getting the Most Out of Your...CA Technologies
 
2015 akamai ir_summit_show_file_v6_unanimated
2015 akamai ir_summit_show_file_v6_unanimated2015 akamai ir_summit_show_file_v6_unanimated
2015 akamai ir_summit_show_file_v6_unanimatednoellefaris
 
Scribe online 01 best practices for sol performance
Scribe online 01   best practices for sol performanceScribe online 01   best practices for sol performance
Scribe online 01 best practices for sol performanceScribe Software Corp.
 
Optimize EC2 for Fun and Profit - SRV203 - Anaheim AWS Summit
Optimize EC2 for Fun and Profit - SRV203 - Anaheim AWS SummitOptimize EC2 for Fun and Profit - SRV203 - Anaheim AWS Summit
Optimize EC2 for Fun and Profit - SRV203 - Anaheim AWS SummitAmazon Web Services
 
Remove Undifferentiated Heavy Lifting from CI/CD Toolsets with Corteva Agrisc...
Remove Undifferentiated Heavy Lifting from CI/CD Toolsets with Corteva Agrisc...Remove Undifferentiated Heavy Lifting from CI/CD Toolsets with Corteva Agrisc...
Remove Undifferentiated Heavy Lifting from CI/CD Toolsets with Corteva Agrisc...Amazon Web Services
 
How Mozy Uses Gainsight for Collections and Billing
How Mozy Uses Gainsight for Collections and BillingHow Mozy Uses Gainsight for Collections and Billing
How Mozy Uses Gainsight for Collections and BillingGainsight
 
Cloudlets and DevOps - A Dangerously Powerful Combination to Extend Capabilit...
Cloudlets and DevOps - A Dangerously Powerful Combination to Extend Capabilit...Cloudlets and DevOps - A Dangerously Powerful Combination to Extend Capabilit...
Cloudlets and DevOps - A Dangerously Powerful Combination to Extend Capabilit...Akamai Developers & Admins
 
Site Managing Performance
Site Managing PerformanceSite Managing Performance
Site Managing PerformanceDesmond Tam
 
Well Architected: Cost Optimization
Well Architected: Cost Optimization Well Architected: Cost Optimization
Well Architected: Cost Optimization Amazon Web Services
 
Networks, Networks Everywhere, And Not A Packet To Drink
Networks, Networks Everywhere, And Not A Packet To DrinkNetworks, Networks Everywhere, And Not A Packet To Drink
Networks, Networks Everywhere, And Not A Packet To DrinkReadWrite
 
Akamai IR Summit 2013
Akamai IR Summit 2013Akamai IR Summit 2013
Akamai IR Summit 2013Liz Bradley
 
Sequencing work for maximum economic benefit using the Scaled Agile Framework...
Sequencing work for maximum economic benefit using the Scaled Agile Framework...Sequencing work for maximum economic benefit using the Scaled Agile Framework...
Sequencing work for maximum economic benefit using the Scaled Agile Framework...Association for Project Management
 
How to Improve agile team efficiency
How to Improve agile team efficiencyHow to Improve agile team efficiency
How to Improve agile team efficiencyHimanshu Desai
 
Metrics: Talking about Churn and Retention in the Board Room
Metrics: Talking about Churn and Retention in the Board RoomMetrics: Talking about Churn and Retention in the Board Room
Metrics: Talking about Churn and Retention in the Board RoomGainsight
 
Amplify_Session_1185_Final Version
Amplify_Session_1185_Final VersionAmplify_Session_1185_Final Version
Amplify_Session_1185_Final VersionGuruprasad Samaga
 
How Spyder Trap Transformed Its Business with QuickBase Sync
How Spyder Trap Transformed Its Business with QuickBase SyncHow Spyder Trap Transformed Its Business with QuickBase Sync
How Spyder Trap Transformed Its Business with QuickBase SyncQuickBase, Inc.
 

Similar to 2015 Velocity SC: Convince your CFO that #perfmatters (20)

Advanced Caching Concepts @ Velocity NY 2015
Advanced Caching Concepts @ Velocity NY 2015Advanced Caching Concepts @ Velocity NY 2015
Advanced Caching Concepts @ Velocity NY 2015
 
Con1169 office depot scm cloud mbx session
Con1169 office depot scm cloud mbx sessionCon1169 office depot scm cloud mbx session
Con1169 office depot scm cloud mbx session
 
Dev348 ReInvent Corteva Agriscience
Dev348   ReInvent Corteva AgriscienceDev348   ReInvent Corteva Agriscience
Dev348 ReInvent Corteva Agriscience
 
Sap Product Lifecycle Costing
Sap Product Lifecycle CostingSap Product Lifecycle Costing
Sap Product Lifecycle Costing
 
CA Workload Automation ESP: Tips and Tricks for Getting the Most Out of Your...
CA Workload Automation ESP: Tips and Tricks for Getting the Most Out of Your...CA Workload Automation ESP: Tips and Tricks for Getting the Most Out of Your...
CA Workload Automation ESP: Tips and Tricks for Getting the Most Out of Your...
 
2015 akamai ir_summit_show_file_v6_unanimated
2015 akamai ir_summit_show_file_v6_unanimated2015 akamai ir_summit_show_file_v6_unanimated
2015 akamai ir_summit_show_file_v6_unanimated
 
Scribe online 01 best practices for sol performance
Scribe online 01   best practices for sol performanceScribe online 01   best practices for sol performance
Scribe online 01 best practices for sol performance
 
Optimize EC2 for Fun and Profit - SRV203 - Anaheim AWS Summit
Optimize EC2 for Fun and Profit - SRV203 - Anaheim AWS SummitOptimize EC2 for Fun and Profit - SRV203 - Anaheim AWS Summit
Optimize EC2 for Fun and Profit - SRV203 - Anaheim AWS Summit
 
Remove Undifferentiated Heavy Lifting from CI/CD Toolsets with Corteva Agrisc...
Remove Undifferentiated Heavy Lifting from CI/CD Toolsets with Corteva Agrisc...Remove Undifferentiated Heavy Lifting from CI/CD Toolsets with Corteva Agrisc...
Remove Undifferentiated Heavy Lifting from CI/CD Toolsets with Corteva Agrisc...
 
How Mozy Uses Gainsight for Collections and Billing
How Mozy Uses Gainsight for Collections and BillingHow Mozy Uses Gainsight for Collections and Billing
How Mozy Uses Gainsight for Collections and Billing
 
Cloudlets and DevOps - A Dangerously Powerful Combination to Extend Capabilit...
Cloudlets and DevOps - A Dangerously Powerful Combination to Extend Capabilit...Cloudlets and DevOps - A Dangerously Powerful Combination to Extend Capabilit...
Cloudlets and DevOps - A Dangerously Powerful Combination to Extend Capabilit...
 
Site Managing Performance
Site Managing PerformanceSite Managing Performance
Site Managing Performance
 
Well Architected: Cost Optimization
Well Architected: Cost Optimization Well Architected: Cost Optimization
Well Architected: Cost Optimization
 
Networks, Networks Everywhere, And Not A Packet To Drink
Networks, Networks Everywhere, And Not A Packet To DrinkNetworks, Networks Everywhere, And Not A Packet To Drink
Networks, Networks Everywhere, And Not A Packet To Drink
 
Akamai IR Summit 2013
Akamai IR Summit 2013Akamai IR Summit 2013
Akamai IR Summit 2013
 
Sequencing work for maximum economic benefit using the Scaled Agile Framework...
Sequencing work for maximum economic benefit using the Scaled Agile Framework...Sequencing work for maximum economic benefit using the Scaled Agile Framework...
Sequencing work for maximum economic benefit using the Scaled Agile Framework...
 
How to Improve agile team efficiency
How to Improve agile team efficiencyHow to Improve agile team efficiency
How to Improve agile team efficiency
 
Metrics: Talking about Churn and Retention in the Board Room
Metrics: Talking about Churn and Retention in the Board RoomMetrics: Talking about Churn and Retention in the Board Room
Metrics: Talking about Churn and Retention in the Board Room
 
Amplify_Session_1185_Final Version
Amplify_Session_1185_Final VersionAmplify_Session_1185_Final Version
Amplify_Session_1185_Final Version
 
How Spyder Trap Transformed Its Business with QuickBase Sync
How Spyder Trap Transformed Its Business with QuickBase SyncHow Spyder Trap Transformed Its Business with QuickBase Sync
How Spyder Trap Transformed Its Business with QuickBase Sync
 

Recently uploaded

Streamlining Your Application Builds with Cloud Native Buildpacks
Streamlining Your Application Builds  with Cloud Native BuildpacksStreamlining Your Application Builds  with Cloud Native Buildpacks
Streamlining Your Application Builds with Cloud Native BuildpacksVish Abrams
 
Cybersecurity Challenges with Generative AI - for Good and Bad
Cybersecurity Challenges with Generative AI - for Good and BadCybersecurity Challenges with Generative AI - for Good and Bad
Cybersecurity Challenges with Generative AI - for Good and BadIvo Andreev
 
Deep Learning for Images with PyTorch - Datacamp
Deep Learning for Images with PyTorch - DatacampDeep Learning for Images with PyTorch - Datacamp
Deep Learning for Images with PyTorch - DatacampVICTOR MAESTRE RAMIREZ
 
OpenChain Webinar: Universal CVSS Calculator
OpenChain Webinar: Universal CVSS CalculatorOpenChain Webinar: Universal CVSS Calculator
OpenChain Webinar: Universal CVSS CalculatorShane Coughlan
 
JS-Experts - Cybersecurity for Generative AI
JS-Experts - Cybersecurity for Generative AIJS-Experts - Cybersecurity for Generative AI
JS-Experts - Cybersecurity for Generative AIIvo Andreev
 
Sales Territory Management: A Definitive Guide to Expand Sales Coverage
Sales Territory Management: A Definitive Guide to Expand Sales CoverageSales Territory Management: A Definitive Guide to Expand Sales Coverage
Sales Territory Management: A Definitive Guide to Expand Sales CoverageDista
 
Fields in Java and Kotlin and what to expect.pptx
Fields in Java and Kotlin and what to expect.pptxFields in Java and Kotlin and what to expect.pptx
Fields in Java and Kotlin and what to expect.pptxJoão Esperancinha
 
eAuditor Audits & Inspections - conduct field inspections
eAuditor Audits & Inspections - conduct field inspectionseAuditor Audits & Inspections - conduct field inspections
eAuditor Audits & Inspections - conduct field inspectionsNirav Modi
 
ARM Talk @ Rejekts - Will ARM be the new Mainstream in our Data Centers_.pdf
ARM Talk @ Rejekts - Will ARM be the new Mainstream in our Data Centers_.pdfARM Talk @ Rejekts - Will ARM be the new Mainstream in our Data Centers_.pdf
ARM Talk @ Rejekts - Will ARM be the new Mainstream in our Data Centers_.pdfTobias Schneck
 
Generative AI for Cybersecurity - EC-Council
Generative AI for Cybersecurity - EC-CouncilGenerative AI for Cybersecurity - EC-Council
Generative AI for Cybersecurity - EC-CouncilVICTOR MAESTRE RAMIREZ
 
Webinar_050417_LeClair12345666777889.ppt
Webinar_050417_LeClair12345666777889.pptWebinar_050417_LeClair12345666777889.ppt
Webinar_050417_LeClair12345666777889.pptkinjal48
 
Growing Oxen: channel operators and retries
Growing Oxen: channel operators and retriesGrowing Oxen: channel operators and retries
Growing Oxen: channel operators and retriesSoftwareMill
 
ERP For Electrical and Electronics manufecturing.pptx
ERP For Electrical and Electronics manufecturing.pptxERP For Electrical and Electronics manufecturing.pptx
ERP For Electrical and Electronics manufecturing.pptxAutus Cyber Tech
 
Big Data Bellevue Meetup | Enhancing Python Data Loading in the Cloud for AI/ML
Big Data Bellevue Meetup | Enhancing Python Data Loading in the Cloud for AI/MLBig Data Bellevue Meetup | Enhancing Python Data Loading in the Cloud for AI/ML
Big Data Bellevue Meetup | Enhancing Python Data Loading in the Cloud for AI/MLAlluxio, Inc.
 
How Does the Epitome of Spyware Differ from Other Malicious Software?
How Does the Epitome of Spyware Differ from Other Malicious Software?How Does the Epitome of Spyware Differ from Other Malicious Software?
How Does the Epitome of Spyware Differ from Other Malicious Software?AmeliaSmith90
 
20240319 Car Simulator Plan.pptx . Plan for a JavaScript Car Driving Simulator.
20240319 Car Simulator Plan.pptx . Plan for a JavaScript Car Driving Simulator.20240319 Car Simulator Plan.pptx . Plan for a JavaScript Car Driving Simulator.
20240319 Car Simulator Plan.pptx . Plan for a JavaScript Car Driving Simulator.Sharon Liu
 
Introduction-to-Software-Development-Outsourcing.pptx
Introduction-to-Software-Development-Outsourcing.pptxIntroduction-to-Software-Development-Outsourcing.pptx
Introduction-to-Software-Development-Outsourcing.pptxIntelliSource Technologies
 
AI Embracing Every Shade of Human Beauty
AI Embracing Every Shade of Human BeautyAI Embracing Every Shade of Human Beauty
AI Embracing Every Shade of Human BeautyRaymond Okyere-Forson
 
IA Generativa y Grafos de Neo4j: RAG time
IA Generativa y Grafos de Neo4j: RAG timeIA Generativa y Grafos de Neo4j: RAG time
IA Generativa y Grafos de Neo4j: RAG timeNeo4j
 
Enterprise Document Management System - Qualityze Inc
Enterprise Document Management System - Qualityze IncEnterprise Document Management System - Qualityze Inc
Enterprise Document Management System - Qualityze Incrobinwilliams8624
 

Recently uploaded (20)

Streamlining Your Application Builds with Cloud Native Buildpacks
Streamlining Your Application Builds  with Cloud Native BuildpacksStreamlining Your Application Builds  with Cloud Native Buildpacks
Streamlining Your Application Builds with Cloud Native Buildpacks
 
Cybersecurity Challenges with Generative AI - for Good and Bad
Cybersecurity Challenges with Generative AI - for Good and BadCybersecurity Challenges with Generative AI - for Good and Bad
Cybersecurity Challenges with Generative AI - for Good and Bad
 
Deep Learning for Images with PyTorch - Datacamp
Deep Learning for Images with PyTorch - DatacampDeep Learning for Images with PyTorch - Datacamp
Deep Learning for Images with PyTorch - Datacamp
 
OpenChain Webinar: Universal CVSS Calculator
OpenChain Webinar: Universal CVSS CalculatorOpenChain Webinar: Universal CVSS Calculator
OpenChain Webinar: Universal CVSS Calculator
 
JS-Experts - Cybersecurity for Generative AI
JS-Experts - Cybersecurity for Generative AIJS-Experts - Cybersecurity for Generative AI
JS-Experts - Cybersecurity for Generative AI
 
Sales Territory Management: A Definitive Guide to Expand Sales Coverage
Sales Territory Management: A Definitive Guide to Expand Sales CoverageSales Territory Management: A Definitive Guide to Expand Sales Coverage
Sales Territory Management: A Definitive Guide to Expand Sales Coverage
 
Fields in Java and Kotlin and what to expect.pptx
Fields in Java and Kotlin and what to expect.pptxFields in Java and Kotlin and what to expect.pptx
Fields in Java and Kotlin and what to expect.pptx
 
eAuditor Audits & Inspections - conduct field inspections
eAuditor Audits & Inspections - conduct field inspectionseAuditor Audits & Inspections - conduct field inspections
eAuditor Audits & Inspections - conduct field inspections
 
ARM Talk @ Rejekts - Will ARM be the new Mainstream in our Data Centers_.pdf
ARM Talk @ Rejekts - Will ARM be the new Mainstream in our Data Centers_.pdfARM Talk @ Rejekts - Will ARM be the new Mainstream in our Data Centers_.pdf
ARM Talk @ Rejekts - Will ARM be the new Mainstream in our Data Centers_.pdf
 
Generative AI for Cybersecurity - EC-Council
Generative AI for Cybersecurity - EC-CouncilGenerative AI for Cybersecurity - EC-Council
Generative AI for Cybersecurity - EC-Council
 
Webinar_050417_LeClair12345666777889.ppt
Webinar_050417_LeClair12345666777889.pptWebinar_050417_LeClair12345666777889.ppt
Webinar_050417_LeClair12345666777889.ppt
 
Growing Oxen: channel operators and retries
Growing Oxen: channel operators and retriesGrowing Oxen: channel operators and retries
Growing Oxen: channel operators and retries
 
ERP For Electrical and Electronics manufecturing.pptx
ERP For Electrical and Electronics manufecturing.pptxERP For Electrical and Electronics manufecturing.pptx
ERP For Electrical and Electronics manufecturing.pptx
 
Big Data Bellevue Meetup | Enhancing Python Data Loading in the Cloud for AI/ML
Big Data Bellevue Meetup | Enhancing Python Data Loading in the Cloud for AI/MLBig Data Bellevue Meetup | Enhancing Python Data Loading in the Cloud for AI/ML
Big Data Bellevue Meetup | Enhancing Python Data Loading in the Cloud for AI/ML
 
How Does the Epitome of Spyware Differ from Other Malicious Software?
How Does the Epitome of Spyware Differ from Other Malicious Software?How Does the Epitome of Spyware Differ from Other Malicious Software?
How Does the Epitome of Spyware Differ from Other Malicious Software?
 
20240319 Car Simulator Plan.pptx . Plan for a JavaScript Car Driving Simulator.
20240319 Car Simulator Plan.pptx . Plan for a JavaScript Car Driving Simulator.20240319 Car Simulator Plan.pptx . Plan for a JavaScript Car Driving Simulator.
20240319 Car Simulator Plan.pptx . Plan for a JavaScript Car Driving Simulator.
 
Introduction-to-Software-Development-Outsourcing.pptx
Introduction-to-Software-Development-Outsourcing.pptxIntroduction-to-Software-Development-Outsourcing.pptx
Introduction-to-Software-Development-Outsourcing.pptx
 
AI Embracing Every Shade of Human Beauty
AI Embracing Every Shade of Human BeautyAI Embracing Every Shade of Human Beauty
AI Embracing Every Shade of Human Beauty
 
IA Generativa y Grafos de Neo4j: RAG time
IA Generativa y Grafos de Neo4j: RAG timeIA Generativa y Grafos de Neo4j: RAG time
IA Generativa y Grafos de Neo4j: RAG time
 
Enterprise Document Management System - Qualityze Inc
Enterprise Document Management System - Qualityze IncEnterprise Document Management System - Qualityze Inc
Enterprise Document Management System - Qualityze Inc
 

2015 Velocity SC: Convince your CFO that #perfmatters

  • 1. Convince Your CFO that #PerfMatters @ColinBendell
  • 2. ©2015 AKAMAI | FASTER FORWARDTM
  • 3. ©2015 AKAMAI | FASTER FORWARDTM • Convincing the Business (General) • Save you $100k in 10 minutes • COGS & CapEx Fin. Modeling • Revenue Modeling
  • 4. ©2015 AKAMAI | FASTER FORWARDTM Nobody likes waiting…
  • 5. ©2015 AKAMAI | FASTER FORWARDTM Source: Equation Research on behalf of Compuware
  • 6. ©2015 AKAMAI | FASTER FORWARDTM Source: Equation Research on behalf of Compuware
  • 7. ©2015 AKAMAI | FASTER FORWARDTM Source: Akamai
  • 8. ©2015 AKAMAI | FASTER FORWARDTM But does Perf. Investment Business Sense?
  • 9. ©2015 AKAMAI | FASTER FORWARDTM
  • 10. ©2015 AKAMAI | FASTER FORWARDTM • Use existing business metrics: • Revenue • Conversion Rate • Don't forget in-direct metrics: • Engagement • Successful Logins Lara Hogan http://larahogan.me Run A/B Tests; Compare results
  • 11. ©2015 AKAMAI | FASTER FORWARDTM Create Pride: Show your users' experience
  • 12. ©2015 AKAMAI | FASTER FORWARDTM Compare Yourself with Your Competition Andy Davies http://andydavies.me/
  • 13. ©2015 AKAMAI | FASTER FORWARDTM
  • 14. ©2015 AKAMAI | FASTER FORWARDTM
  • 15. ©2015 AKAMAI | FASTER FORWARDTM Other Business' Success Stories
  • 16. ©2015 AKAMAI | FASTER FORWARDTM Performance Impacts Business Results Source: Walmart, Torbit, Aberdeen Group
  • 17. ©2015 AKAMAI | FASTER FORWARDTM Speed Matters: Conversion Rate vs. Load Time Conversion  Rate  (%) 0-­1 1-­2 2-­3 3-­4 4-­5 5-­6 6-­7 7-­8 8-­9 9-­10 10-­11 11+ Population Conversion  Rate  (%) Page  Load  Time  (Sec)
  • 18. ©2015 AKAMAI | FASTER FORWARDTM Performance Impacts Engagement Josh Fraser @joshfraser
  • 19. ©2015 AKAMAI | FASTER FORWARDTM 2 Seconds Faster Boosted Conversions By 66% http://www.conversionconference.com/blog/case-study-how-making-pages-2-seconds-faster-boosted-conversions-by-66/ Tammy Everts @tameverts
  • 20. ©2015 AKAMAI | FASTER FORWARDTM
  • 21. ©2015 AKAMAI | FASTER FORWARDTM
  • 22. ©2015 AKAMAI | FASTER FORWARDTM Spreadsheets: the Language of the CFO
  • 23. ©2015 AKAMAI | FASTER FORWARDTM
  • 24. ©2015 AKAMAI | FASTER FORWARDTM 10 minute MBA
  • 25. ©2015 AKAMAI | FASTER FORWARDTM Money buys beer
  • 26. ©2015 AKAMAI | FASTER FORWARDTM How do I get [more] Money?
  • 27. ©2015 AKAMAI | FASTER FORWARDTM Long Answer: GDP + ++ GDP = C + I + G + (X-M)
  • 28. ©2015 AKAMAI | FASTER FORWARDTM Short Answer: Your Company Earnes More Money
  • 29. ©2015 AKAMAI | FASTER FORWARDTM Stock Price is What Others Think is the Company Value Company_Value =  Stock_Price *  Shares
  • 30. ©2015 AKAMAI | FASTER FORWARDTM Discount Cash Flow // Common Valuation Technique1 𝐶𝑜𝑚𝑝𝑎𝑛𝑦_𝑉𝑎𝑙𝑢𝑒 = 𝐶𝑎𝑠ℎ_𝐹𝑙𝑜𝑤2 1 + 𝑊𝑒𝑖𝑔ℎ𝑡𝑒𝑑_𝐴𝑣𝑒𝑟𝑎𝑔𝑒_𝐶𝑜𝑠𝑡_𝐶𝑎𝑝𝑖𝑡𝑎𝑙 2 + 𝐶𝐹= 1 + 𝑊𝐴𝐶𝐶 = + ⋯ + 𝐶𝐹? 1 + 𝑊𝐴𝐶𝐶 ? + 1This  is  one  of  many  techniques.  It  is  has  *many*  flaws.
  • 31. ©2015 AKAMAI | FASTER FORWARDTM Discount Cash Flow //a common valuation technique 𝐶𝑜𝑚𝑝𝑎𝑛𝑦_𝑉𝑎𝑙𝑢𝑒 = 𝐶𝑎𝑠ℎ_𝐹𝑙𝑜𝑤2 1 + 𝑊𝑒𝑖𝑔ℎ𝑡𝑒𝑑_𝐴𝑣𝑒𝑟𝑎𝑔𝑒_𝐶𝑜𝑠𝑡_𝐶𝑎𝑝𝑖𝑡𝑎𝑙 2 + 𝐶𝐹= 1 + 𝑊𝐴𝐶𝐶 = + ⋯ + 𝐶𝐹? 1 + 𝑊𝐴𝐶𝐶 ? +
  • 32. ©2015 AKAMAI | FASTER FORWARDTM Profit & Loss Statement (or Income Statement) How  Much  Beer  You  Sold Cost  to  Make  Beer Management Savings  for  more  distilleries Your  Money
  • 33. ©2015 AKAMAI | FASTER FORWARDTM Three Big Leavers to Increase Profits How  Much  Beer  You  Sold Cost  to  Make  Beer Savings  for  more  distilleries Your  Money
  • 34. ©2015 AKAMAI | FASTER FORWARDTM More Money. More Beer.
  • 35. ©2015 AKAMAI | FASTER FORWARDTM We are at a Performance Conference… … how does performance impact your P&L?
  • 36. ©2015 AKAMAI | FASTER FORWARDTM Velocity SC 2015: Financial Impact How LinkedIn used RUM to drive optimizations and make the site faster Thursday @ 13:45pm - Ritesh Maheshwari (LinkedIn) What are third-party components doing to your site? Thursday 14:40 - Andy Davies & Simon Hearne (NCC Group) Reaching everyone Friday 11:50 - Tim Kadlec (Akamai) How Go is making us faster Friday @14:40 - Wilfried Schobeiri (MediaMath PHP 7 Friday @16:10 - Rasmus Lerdorf (Etsy) Speeding up the PlayStation 4 user interface - web performance optimization outside the browser Friday 17:05 - JP CASTRO (Sony) Yesterday's perf best-practices are today's HTTP/2 anti-patterns Friday 17:05 - Ilya Grigorik (Google) Revenue    COGS    CapEx ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔
  • 37. ©2015 AKAMAI | FASTER FORWARDTM Mapping performance to the Income Statement Revenue    COGS    CapEx Reduce bytes ✔ ✔ ✔ Reduce requests ✔ ✔ ✔ Expedite rendering ✔ Reduce time-to-last-byte ✔ ✔ ✔
  • 38. ©2015 AKAMAI | FASTER FORWARDTM Financial Modeling Performance
  • 39. ©2015 AKAMAI | FASTER FORWARDTM What is Financial Modeling? • Excel! Yay! • Technique to build confidence in decision making • Consistent comparison for project planning outcomes • Good: consistent IRR comparison • Bad: ignores Goodwill (Brand) and Job satisfaction (Culture) • Two schools of thought: • Precision: Consider every detail • Directionality: Consider the big inputs
  • 40. ©2015 AKAMAI | FASTER FORWARDTM ProTip 1: Avoid the wormhole of Finance Precision. Use Directionality
  • 41. ©2015 AKAMAI | FASTER FORWARDTM ProTip 2: Use conservative shortcuts
  • 42. ©2015 AKAMAI | FASTER FORWARDTM Model 1: CapEx / OpEx CashFlow
  • 43. ©2015 AKAMAI | FASTER FORWARDTM Build-More or Optimize
  • 44. ©2015 AKAMAI | FASTER FORWARDTM CashFlow Opportunities: Requests, Bytes, TTLB Server Time Handshake
  • 45. ©2015 AKAMAI | FASTER FORWARDTM Pros & Cons of CashFlow Modeling Use this model: • When you are optimizing backend processes • Adding caching layers (back-end, cdn, client) • Optimizing user workflows • Optimizing page layout Don't use this model: • Optimizing client compute (eg: javascript) • Optimizing client rendering (eg: adding webp)
  • 46. ©2015 AKAMAI | FASTER FORWARDTM Cash_Flow =   Capital_Exenses +  Operational_Expenses
  • 47. ©2015 AKAMAI | FASTER FORWARDTM Typical Year-over-Year Spend
  • 48. ©2015 AKAMAI | FASTER FORWARDTM Questions • Will your improvements slow the rate of new hardware purchased? • Will it reduce the operating costs? • Will you be able to turn down servers? • What are the projected costs with and without improved performance?
  • 49. ©2015 AKAMAI | FASTER FORWARDTM Avoid including "Funny Numbers" Don't include: • Productivity gains • Revenue from selling old hardware • Contract [re]negotiation time • Software Licensing savings This will allow you to bypass a 7 week discussion with your procurement about the true cost of your enterprise agreement.
  • 50. ©2015 AKAMAI | FASTER FORWARDTM This company tried computing the true cost of the enterprise agreement.
  • 51. ©2015 AKAMAI | FASTER FORWARDTM Two Models • CoLo / OnPrem • This years Capital Expense contributes to next years' Operating Expense • You can theoretically extend the life of purchased capital • IaaS / PaaS • Everything is Operational Expenses
  • 52. ©2015 AKAMAI | FASTER FORWARDTM Operating_Exepnse =  Number  of  Servers  *  Usage OR Capital_Expense =  Number_of_New_Servers *  Average_Server_Cost Operating_Exepnse =  Number_of_Servers *  Average_Server_KVA *  CoLo_Cost_per_KVA
  • 53. ©2015 AKAMAI | FASTER FORWARDTM Calculating Capital Expenses (CapEx) CapEx =  Number_of_Servers *  Average_Cost • Use "Fully Loaded" cost (incl: cost to procure, security audits, colo service tickets, rack & stack fees, etc) • Use lowest-common-denominator to save inventorying your infrastructure • If in doubt: $5,000 $100,000
  • 54. ©2015 AKAMAI | FASTER FORWARDTM A Note about Virtual Machines • Virtual Servers also require Capital • Don't forget to include (vMotion) buffer • Be consistent in calculation
  • 55. ©2015 AKAMAI | FASTER FORWARDTM Plugging in the Numbers https://goo.gl/6OHj9a
  • 56. ©2015 AKAMAI | FASTER FORWARDTM Calculating Operating Expenses (OpEx) Best Practice: • Include all servers, regardless of role • Assume that increase in user demand = increase in infrastructure use Data-Center OpEx • Most Co-Lo providers bill based on power usage • Assume Co-Lo pricing includes cooling, floor space, and internet • Each datacenter is a special snowflake: be conservative in your calculations
  • 57. ©2015 AKAMAI | FASTER FORWARDTM OpEx OpEx =  Number_of_Servers *  KVA_per_Server *  KVA_Price • Pick one hardware profile that is representative • If in doubt, use the newest hardware’s power consumption • Use 80% of the manufacturers reported fully loaded power draw • Most hardware reports power as Watts and BTU. Assume a power- factor of 0.9 and use: • KVA  =  Watts  /  900 • When in doubt I use these approximation numbers: • 0.5KVA for an average pizza box server • 3.5KVA for a 6U high density compute chassis
  • 58. ©2015 AKAMAI | FASTER FORWARDTM Colo's charge by electricity OpEx =  Number_of_Servers *  KVA_per_Server *  KVA_Price • Most CoLos bundle amenities and charge by electricity • Ranges from $0.70/KVA/mo to $0.20/KVA/mo • Running your own datacenter =~ colo space • If in doubt, a conservative, after tax value: $0.50/KVA/mo
  • 59. ©2015 AKAMAI | FASTER FORWARDTM
  • 60. ©2015 AKAMAI | FASTER FORWARDTM Plugging in the Numbers https://goo.gl/6OHj9a
  • 61. ©2015 AKAMAI | FASTER FORWARDTM Example • 140 “pizza box” type servers ($5k, 0.5KVA) • 400 Page View/s peak • 30% YoY growth rate • Has a problem with personalization tracking
  • 62. ©2015 AKAMAI | FASTER FORWARDTM Before
  • 63. ©2015 AKAMAI | FASTER FORWARDTM After:
  • 64. ©2015 AKAMAI | FASTER FORWARDTM Ignore the size of the OpEx budget Focus on user performance budget
  • 65. ©2015 AKAMAI | FASTER FORWARDTM PageView and Interaction Cost • A Page View requests that return Content-­‐Type:  text/html. • Page View is not perfect: use a metric that represents user activity • Conversions • Checkout-Rate • Logins
  • 66. ©2015 AKAMAI | FASTER FORWARDTM Page View is a proxy for 'user interaction'
  • 67. ©2015 AKAMAI | FASTER FORWARDTM Interaction Cost Interaction_Cost =  (OpEx +  CapEx)  /  Page_View_per_Second • Max page view per second tells us how much money is spent to maintain this peak traffic • You have built implicitly to meet peak traffic • As user demand grows, you will build more hardware in lock step • Interaction Cost to remain constant year over year
  • 68. ©2015 AKAMAI | FASTER FORWARDTM Before (with $/PVs)
  • 69. ©2015 AKAMAI | FASTER FORWARDTM After (with $/PVs)
  • 70. ©2015 AKAMAI | FASTER FORWARDTM What about IaaS / PaaS? • Model still applies with a few tweaks; ignore Capital costs • Each IaaS / PaaS provider has different pricing levers • Focus on cost-per-interaction model
  • 71. ©2015 AKAMAI | FASTER FORWARDTM Please, tell me more about your made-up financial model
  • 72. ©2015 AKAMAI | FASTER FORWARDTM User Performance Improvement (RUM Data)
  • 73. ©2015 AKAMAI | FASTER FORWARDTM Origin Bandwidth Reduced to Below 100mbps
  • 74. ©2015 AKAMAI | FASTER FORWARDTM Reduce Server Utilization $- $10,000 $20,000 $30,000 $40,000 $50,000 $60,000 $70,000 $80,000 $90,000 CapEx… OpEx Assumptions:  5yr  Amortization;;  0.5KVA/server;;  $0.50/VA/mo $200,000 savings  over  5yrs
  • 75. ©2015 AKAMAI | FASTER FORWARDTM Cashflow Summary • Backend performance can impact end-user performance • Performance can have financial impact • Use CapEx and OpEx modeling to show directionality of impact • Avoid the detailed nuances – use conservative generalizations • You can use the same model to project negative financial impact (eg: image responsive images)
  • 76. ©2015 AKAMAI | FASTER FORWARDTM Revenue Probabilistic Modeling
  • 77. ©2015 AKAMAI | FASTER FORWARDTM Revenue Modeling • Your Operations teams can control CapEx and OpEx spending What about Revenue? • How can performance improve total sales? • How performance increases user growth (CAGR) • How performance increases sales per user (ARPU)
  • 78. ©2015 AKAMAI | FASTER FORWARDTM Revenue Modeling is Hard
  • 79. ©2015 AKAMAI | FASTER FORWARDTM Applying Monte-Carlo Simulation • Apply probability based input on an equation • Run many times • Evaluate the results
  • 80. ©2015 AKAMAI | FASTER FORWARDTM Plugging in the Numbers https://goo.gl/6OHj9a
  • 81. ©2015 AKAMAI | FASTER FORWARDTM Notes about Revenue Modeling • Use Revenue Modeling to show directionality • Avoid absolute claims about performance • Consider regression analysis to understand what probability inputs • Know your user performance distribution first. (You are using Real-User-Monitoring, right?)
  • 82. ©2015 AKAMAI | FASTER FORWARDTM So yea, you're going to have to come in to work over the weekend now…
  • 83. ©2015 AKAMAI | FASTER FORWARDTM Summary • Performance correlates to engagement • Make performance part of your culture • Understand the user experience and leverage empathy • Project cash flow savings from performance • Applies to OnPrem and IaaS/PaaS • Project revenue impact from performance
  • 84. ©2015 AKAMAI | FASTER FORWARDTM Performance can make happy users and makes business sense!