Many organisations are transforming their businesses through modernised tech stacks. But when replacing multiple legacy systems (CRM, ERP, middleware) supporting core business functions, you’re faced with multiple considerations.
These are:
How do you manage risk - both program and technical?
What are the upsides and opportunities of parallel implementation?
What are the key considerations you need to make upfront?
What business and technical strategies can you adopt?
How do you measure success across your program?
A parallel greenfield implementation of multiple systems involves many risks, including
Lack of coordination leading to “desynced” growth,
Project bottlenecks due to a single team being late,
Deadline pressure leading to ad-hoc integrations/solutions,
Low business adoption and/or subpar data quality and data governance.
It is, however, important to take note of the many upsides and opportunities which exist for parallel implementations - if consciously considered diligently implemented. These include:
Common DevOps with shared GIT codebase, shared sandboxes, and shared QA team (for end-2-end integration testing)
Federated Identity & Access Management
Unified security strategy based on SWOT analysis
Global and coherent data architecture
So, what are the key things you need to consider and address in your parallel program?
I. Delivery Methodology
Choose an Agile approach (bottom-up) if you value constant business feedback while growing the platforms at the cost of total build time.
Choose a Waterfall approach (top-down) if you prefer a battle-tested implementation or a visionary direction at the cost of business flexibility.
Alternatively, carefully select a form of hybrid-agile - being extremely mindful of integration planning, design, and implementation.
II. Change Management
Roles (RACI matrix + SME list)
Shared communication plans
Documentation and Training
Localization (e.g. Translations and currencies)
III. Design Architecture Board to decide:
RAID log (Risks, Assumptions, Issues, Dependencies)
Testing, including how to automate it end-to-end across systems
How to mock the pending integration
Audit requirements
Governor limit mitigations
IV. Data Governance Council to draft:
Canonical data model
Migration plan
Data cleaning and enrichment
System of Records with data lineage
Key strategies you may consider to set your multi-system program up for success include:
Putting in place a strong emphasis on ownership for teams, systems, data, and processes
Leveraging phased capability-based (or country-based) rollouts
Visualise Your program charter on the wall!
API-Led Connectivity: use different APIs to follow a “Separation of Concerns” philosophy
Implement a logging framework on the middleware side
And last but not least - how do you measure program success?
Business value realisation across business growth, efficiency gains, customer and employee experience improvement
Clear reporting strategy with KPIs
A low number of errors logged
Data quality has a high score, and all data is trackable and has a purpose
Business stakeholders prefer the new systems - and way of working - to the old ones.
In parallel implementation programs, you’re likely relying upon and leveraging a number of specialised parties. Close alignment and governance are key for program success.
Get in touch to learn more about how Waeg, an IBM Company, can help you manage and deliver multi-stream programs where Salesforce CRM is involved.
Kristian Margaryan Jørgensen, Director of Solution Engineering and Architecture @ Waeg
