Software Usage Analysis

1
Software Usage Analysis

• An Enabler for New Licensing Models

Rick Ingram, PresidentLicense Tracker Inc.
2
Outline

• evolution of licensing models
• usage analysis - technical considerations
• usage analysis - contractual considerations
• other considerations
• summary
• questions

3
Evolution of Licensing Models
Increased value to technology consumers
Enhanced customer relationship for software
vendors
Technology Partnerships
(Unlimited Access)
Usage Analysis Enabled License Models
Remix Models
(Flexible Access)
Pay-Per-Use Models
(Rented Access)
Classification Models
(Variable Access)
Increased risk to technology consumers
Decreased commitment to software vendors
Traditional Client Vendor Relationship
(Fixed Access, named user, floating, duration)
4
Evolution of Licensing Models
Technology Partnerships
(Unlimited Access)
Remix Models
(Flexible Access)
Pay-Per-Use Models
(Rented Access)
Classification Models
(Variable Access)
Traditional Models
Traditional Client Vendor Relationship
(Fixed Access, named user, floating, duration)
5
Traditional Licensing Models

• single user, single license
• named user
• node locked
• multiple users, shared license
• concurrent use license manager
• dongle controlled
• fixed duration
• evaluation
• annual lease
• short term requirements

6
Evolution of Licensing Models
Technology Partnerships
(Unlimited Access)
Remix Models
(Flexible Access)
Business Issues Driving the New Models
Pay-Per-Use Models
(Rented Access)
Classification Models
(Variable Access)
Traditional Client Vendor Relationship
(Fixed Access, named user, floating, duration)
7
Motivation for New Models

• for users
• balancing productivity and efficiency
• long term needs are not always known when making
  initial purchases
  
• mid-year requirements can be satisfied with
  operating budgets
  
• for vendors
• competition requires licensing innovation to
  enhance customer relationships and increase
  market share
  
• users needs are vendors opportunities

8
Evolution of Licensing Models
Technology Partnerships
(Unlimited Access)
Remix Models
(Flexible Access)
Pay-Per-Use Models
Classification Models
(Rented Access)
Classification Models
(Variable Access)
Traditional Client Vendor Relationship
(Fixed Access, named user, floating, duration)
9
Essentials of Classification

• an extension to named user licensing
• users are placed into categories based on usage
  requirements
  
• restrictions on which features can be used
• restrictions on duration of use
• restrictions on number of sessions
• usage analysis for monitoring, compliance and
  true-up

10
Evolution of Licensing Models
Technology Partnerships
(Unlimited Access)
Remix Models
(Flexible Access)
PPU Models
Pay-Per-Use Models
(Rented Access)
Classification Models
(Variable Access)
Traditional Client Vendor Relationship
(Fixed Access, named user, floating, duration)
11
Essentials of Pay-Per-Use

• vendors provide non-owned copies of software to
  users
  
• use of all software (owned and rented) is
  recorded in usage logfiles
  
• once per billing cycle the logfile is sent to the
  vendor for analysis
  
• an invoice is prepared based on actual usage of
  non-owned software
  

License Control File
License Manager
License ManagerLogfile
12
Time Based Pay-Per-Use

• technology consumers are charged for the amount
  of time they used non-owned copies of software
  
• on site availability of rented software ensures
  no lag between detection of need and use of
  software

Standard Concurrent Use Chart
Owned License Limit
Times of Non-owned License Use
13
Transaction Based Pay-Per-Use

• usage charges occur because a software module has
  been used, duration of software use is
  irrelevant
  
• most applicable for consistent duration
  functions
  
• usage of each module or function is logged and
  billed accordingly

14
Evolution of Licensing Models
Technology Partnerships
(Unlimited Access)
Remix Models
Remix Models
(Flexible Access)
Pay-Per-Use Models
(Rented Access)
Classification Models
(Variable Access)
Traditional Client Vendor Relationship
(Fixed Access, named user, floating, duration)
15
Essentials of Remix
Initial Pool

• software purchases are viewed as purchases of a
  technology pool, not just individual licenses
  
• consumers have the right to remix their active
  set of licenses to satisfy current needs
  
• normally used in conjunction with pay-per-use

Needs Change
After a remix
16
Periodic Static Remixing

• this model makes use of existing application
  software and license manager
  
• periodically (every 3 or 6 months), the license
  file is updated to reflect the new state of
  remixing
  
• usage analysis provides information to make
  informed remix decisions

17
Continual Real-time Remixing

• end users purchase licenses of a token-feature
  not the actual software features
  
• when features are checked out the corresponding
  number of licenses of the token-feature are
  checked out
  
• Pay-Per-Use can be easily added

Feature - X (3)Feature - Y (4)
Feature - Y (4)
Feature - X (3)
License Manager
Pool of TokenLicenses

• moving a current product family to this model
  requires either
  
• modification of the application source code
• using a token licensing enabled license manager

18
Evolution of Licensing Models
Unlimited Access
Technology Partnerships
(Unlimited Access)
Remix Models
(Flexible Access)
Pay-Per-Use Models
(Rented Access)
Classification Models
(Variable Access)
Traditional Client Vendor Relationship
(Fixed Access, named user, floating, duration)
19
Essentials of Unlimited Access

• multi-year contract providing unlimited access
  for a fixed annual fee

• the fee is adjusted annually based on usage
  during the year
  
• removes the revenue/cost uncertainty associated
  with pure rental arrangements

20
Enabling New Licensing Models
Technology Partnerships
(Unlimited Access)
Usage Analysis Enabled License Models
Remix Models
(Flexible Access)
Pay-Per-Use Models
(Rented Access)
Classification Models
(Variable Access)
Traditional Client Vendor Relationship
(Fixed Access, named user, floating, duration)
21
Sources of Usage Data

• license manager logfiles
• software vendor proprietary logfile or database

License Manager
License ManagerLogfile
Optional VendorUsage Logger
Vendor ProprietaryUsage Data
22
Logfile Issues - Existence

• is logging turned on?
• ensure proper configuration
• overwriting or deletion of logfiles
• overwrite or append on restart
• encourage proper file management policies

23
Logfile Issues - Integrity

• ASCII versus binary
• ASCII can be edited
• binary is secure,but is it availalable?
• ASCII limits customer base
• completeness of data
• license sharing
• detailed session matching

24
Logfile Issues - Access

• automated or manual
• firewall and phone-home issues
• privacy concerns
• anonymizer may be required
• multiple vendors data in a single file
• extractor may be required

25
Enabling New Licensing Models
Technology Partnerships
(Unlimited Access)
Usage Analysis Enabled License Models
Remix Models
(Flexible Access)
Pay-Per-Use Models
(Rented Access)
Classification Models
(Variable Access)
Traditional Client Vendor Relationship
(Fixed Access, named user, floating, duration)
26
Rental Mode Decision 1

• basic time units for rentals(seconds, minutes,
  hours, days, weeks, months, years)
  

27
Rental Mode Decision 2

• elapsed time or calendar periods
• is a day midnight-to-midnight or is it an
  arbitrary 24 hour period from the start of first
  rental
  

Is this a 1 day rental, or 2?
Tues
Wed
Mon
Time
28
Rental Mode Decision 3

• single period or multiple period rentals
• typically, smaller time units are single period
  and longer time periods are multiple period
  

Is this a 1 unit rental, or 2?
29
Rental Mode Decision 4

• simple or combination rental modes
• rate capping (i.e. hourly_capped_monthly)
• volume discounting

Rental total climbs with usage
Until it hits the cap
It then resets at the next period
30
Rental Mode Decision 5

• minimum session length
• ensures that license checkouts for which no
  business value could be realized are ignored
  

Should these sessions be considered?
Or not
31
Rental Mode Decision 6

• minimum overlap period
• minimum period of time for a rental, shorter
  periods are forgiven

Should there be a charge for this period?
Or not
32
Rental Mode Decision 7

• pricing
• typically defined as a percentage of either
  perpetual license or annual license fees
  
• optional components
• access fee
• billing period minimum
• set pricing to create the desired winwin
  scenario
  
• provide reasonable cost access for minor use
  on-demand
  
• price such that high use encourages perpetual
  license purchase
  

33
Rental Mode Decision 8

• applying rentals to future purchases
• what percentage, if any, of rentals can be
  applied as a discount to perpetual license
  purchases

34
Enabling New Licensing Models
Technology Partnerships
(Unlimited Access)
Usage Analysis Enabled License Models
Remix Models
(Flexible Access)
Pay-Per-Use Models
(Rented Access)
Classification Models
(Variable Access)
Traditional Client Vendor Relationship
(Fixed Access, named user, floating, duration)
35
Pay-Per-Use Cost Monitoring

• vendors need to analyze usage data once per
  billing cycle

• end-users must monitor costs on a continual
  basis
  
• ensuring budgets are not being exceeded
• detecting and correcting improper license use
• making mid-period purchases if rental usage is
  higher than expected
  

36
Usage Analysis in ASP Operations

• in an ASP environment usage analysis of standard
  logs can support
  
• accounting and invoice generation for the vendor
• cost and usage monitoring by customers

37
Summary
2. Usage analysis enables license models that
address these needs in a winwin relationship.
7. Usage data characteristics and the
capabilities of the analysistools must be
consideredwhen drafting contractterms.
6. Technology partnerships providemaximum value
for long-term commitments.
5. Remix models meet changing usersneeds while
protecting overall investment.
8. Monitoring of costs must be done by end users
to maintainthe winwinscenario.
4. Pay-per-use models provide on-demandaccess
for peak periods, or in support of ASPs.
3. Classification models provide the revenue
modelof named users with discounts for
restricted user access.
1. Traditional licensing models do not
adequatelyaddress the needs of end-users and
therefore of software vendors.
38
Questions?
You can only optimize -what you understand
and you can only understand -what you track!
Rick Ingram rick_at_licensetracker.ca (403) 923-733
4
www.licensetracker.ca