Part 1: Optimizing demand management to deliver “Just in Time” cloud service provisioning.
By Bill Irvine
A common phrase overheard during the creation of new cloud based innovation environments for modernized applications is “build it and they will come”.
The surprise for many IT organizations is that they do come and the challenge becomes dealing with that success and the on-going management of their new environments. Many organizations do not operationalize their capabilities or establish the governance processes they need to be successful as a cloud service provider by the time the technology goes live.
Common Complaints about Cloud Services
In my work with customers designing solutions to address the needs of their business via cloud-based innovation, we hear a consistent list of concerns.
“We are always blindsided by requests that we don’t have the capacity to fulfill – it came out of nowhere”
“We never get enough specific information from the business on what they want until it’s too late”
“They always want to over-provision the environments – we’re always in negotiation mode on resource requirements”
“There’s never enough capacity to meet the business and operational demand”
“Nobody ever gives back resources – even when we know they’re not being used”
“There’s never enough budget to buy capacity when we need it”
“We are always getting escalations about the speed of provisioning and spend most of our time reacting to delayed and unfulfilled requests”
“We have to wait for approvals for every piece of the PaaS puzzle”
“Everything we do is custom which makes it impossible to automate”
These and a host of similar issues relate to a common theme – the need for effective demand, capacity and request management to ensure a standardized, streamlined, consistent and automated approach to service provisioning.
In this blog series, we will cover each of the key processes in the lifecycle and their importance in creating an IT service brokerage model that can always support the dynamic business demand. The expectations on IT have never been greater.
Communication is Key to Demand Management
The primary goal of demand management is to understand the pipeline of service requirements from the business and to interpret these needs into a predictable forecast of consumption. This forecast becomes a vital part of ensuring that IT always has the capacity to fulfill the evolving requirements and requests.
Sounds easy enough, but the challenge in predicting demand for most organizations is the difference in language between the Line of Business (LOB), the development teams supporting them and the infrastructure providers charged with hosting the services.
IT capacity planners want technical specifications and details of the individual resource components (CPU, Memory, Storage etc.) required in order to ensure the appropriate configuration of resources in the correct “landing zones”.
The business representatives however, typically present their needs in terms of market growth, marketing initiatives that may drive increases in transactions or potential decreases in business volume based on the seasonality of the service supported. These needs are rarely static by nature and evolve over time from conception to reality.
The conversion of business and service needs into technical resource requirements is often more art than science and relies on effective communication and collaboration between a broad group of stakeholders to continuously interpret, structure and mature demand data into knowledge that can be acted upon.
IT needs to interact proactively with the stakeholders to identify demand as early as possible at its source. This source data should be documented in a system of record so that it can be tracked, aligned by service and updated as more detailed information becomes available. Demand data is progressed through a maturity funnel where requirements are codified, refined, validated, prioritized and compared against historical patterns & trends. This enables the initial business data to be transformed into technical resource requirements and actionable plans.
Demand Maturity Concepts
In order to create a comprehensive and contextual picture of current and future business and service demand, requirements should be subjected to a series of analytical steps to refine the demand.
- Capture data from all available sources. Different sources will have differing levels of specificity from business concept to actual service performance data.
- Understand the sources (e.g. LOB, Service data, etc.) to enable comparisons and correlation with past requirements. Grouping requirements by service should become the overall organizing principle to help make sense of the overall demand.
- Develop, configure and size a logical grouping of resources into a service offering (e.g. Infrastructure or Platform as a Service) to simplify the calculation of future needs and enable IT to better standardize and automate the provisioning processes. Pre-defined service offerings also provide the opportunity to steer the customer towards preferred solutions that are more efficient and cost effective.
- Identify patterns of business activity (PBA) for each service and develop educated assumptions as to future needs through the analysis of past requirements, requests and configurations. It’s OK to make assumptions, the business is often guessing at the early stages. Even placeholder information can be valuable especially early in the funnel. Assumptions can be validated and adjusted over the life of the requirement.
- Develop LOB user profiles and analyze their service usage patterns to further refine the understanding of the needs and requests.
- Understand existing patterns of business activity, prior demand and the technical profile of related platforms consumed by the specific business unit, the applications supporting the service and the volume of transactions to form an evolutionary pipeline or funnel.
Demand requirements managed through these activities will provide IT the confidence to commit to more aggressive service levels and guarantees regarding capacity and associated cloud resource provisioning.
Key Demand Management Roles
As mentioned, there are many parties and stakeholders involved in managing demand effectively. The most obvious and often overlooked stakeholders are the lines of business themselves. IT’s continuous interaction with the business is key to their improved understanding of the customer needs and to break the cycle of being reactive and unresponsive.
Two of the key roles to ensure this ongoing relationship and demand based dialog are the Business Relationship Manager (BRM) and the Service Owner. These roles are critical to understanding the patterns of business and service activity and ensuring appropriate capacity and capability on a service-by-service basis.
The BRM has a primary responsibility to represent all of elements of IT and the associated service provision and performance to the business function. They are responsible for orchestrating the capture of demand from the business and assisting in the conversion of these needs into the technical capacity that meets the expectations. BRM activities in support of demand prediction include:
- Identification of customer needs
- Capture of planned projects and initiatives
- Communicating changes in service profiles or volumes
- “Selling” the improvements in service capabilities and helping to influence customer behavior and optimize the business usage of the services provided.
The Service Owner ensures that there is an understanding and awareness of the service as a whole, who utilizes the service, how it supports the business functions, the service capabilities and the current service performance. The Service Owner will be responsible for:
- Quantification and codification of the overall service needs, resource proportions, configuration and operational dynamics to optimize the performance of the production service
- Key input into decisions regarding resource capacity and configuration changes required
- Creation of the environment profiles and service offerings used in the downstream environments (e.g. Dev, Test, QA) as required by the development and operational functions
Demand Management Benefits
Implemented successfully, demand management will enable improvements across all aspects of service provisioning but especially in the areas of capacity and request fulfillment.
Some of the key benefits include:
- Increased customer satisfaction with services and requests being provisioned without the delays inherent in a reactive environment
- Improved and faster understanding of service and business requirements with demand being objectively quantified
- Capacity based risk is identified and addressed throughout the course of the above activities
- Accurate demand and capacity trending will reduce “over-provisioning” and provide more accurate budgetary planning data to optimize resource / infrastructure costs
- Basis for JIT (Just In Time) purchasing and release of capacity using confidence-driven forecasts
- Improved alignment with business goals giving an accurate “picture” of demand activities required to enable business goal attainment
- Increased confidence in allocation of IT resources and their readiness for service provision
So how do you get started with improving demand management? Some proven initial steps developed with our customers include:
- Start talking to the business customers and associated development teams to open the dialog and establish the process
- Enhance standard requirements capture with each line of business defining their requirements by service
- Capture future needs and updates earlier in the lifecycle to feed the forecasting process
- Update the guidelines for all requirements capture to be consistent regardless of type (e.g. innovation, run, grow etc.) in a common format for input into demand planning
- Establish improved methods for collecting trend based run and growth requirements by service.
- Develop Patterns of Business Activity for each service and monitor key performance and consumption metrics to model current and future operational needs.
- Analyze and redefine the real-time metrics you collect to better track and report against ongoing capacity use, headroom requirements and growth
In my next post in the series, I will discuss the capacity management stage of lifecycle, focusing on the conversion of demand into capacity requirements and optimization of the overall capacity plan.
Bill Irvine is a Principal Strategist with VMware Accelerate Advisory Services and is based in Colorado.