Order Management & Fulfillment

Key Go-Live learnings during roll out of IBM Sterling OMS (Part I) – Pilot Modelling

By Jagadesh Hulugundi posted Tue September 29, 2020 12:31 PM


Go-Live of a large IT application is always a major milestone for any business’s transformation journey. Once incepted, the application takes small steps and starts offering enhanced features, functionalities, facilitating and enabling multiple stake holders who are using the application to do their activities more efficiently. 

The objective of this series of blogs is to highlight some key challenges and learnings of going live and the subsequent hyper care window that follows.  In this blog I would focus on Pilot Modelling.

The paramount importance of “Go Live” is demonstrated in the fact that the IT and business teams spend considerable amount of time to get it right the first time itself. Building a large Enterprise Order Management Application (OMS) is like assembling 10,000 parts of a racing car accurately to enable successful outcomes. While exception scenarios are a reality, the applications need to be robustly designed to understand and handle them appropriately. All scenarios for which the application is built need to be executed and exited smoothly. Under all circumstances, the application must have a bail out mechanism to avoid any dead locks that halts businesses. Without design principles, the financial impact on a business can be huge.

During a Go-Live, most teams have checklists, process maps, tools and managerial checkpoints, yet the gaps and loopholes that primarily surface are because of technical limitations, oversight due to missing finer details, manual mistakes and weak infrastructure zones.

Often, production roll outs are not big-bang in the era of Agile, AI and Cognitive. That being said, many business functions and technical teams intersect on a common ground for a pilot strategy. However, the technical solution components are always built for the entire landscape  of the business & IT vision and not just for pilot functions. Now, does this mean that some technical elements needs to be undone, switched off or toned down during a pilot roll-out? If the answer is yes, will this change trigger regression test cycles again, maintenance of multiple run-times, etc? If the answer is no, is the business under risk of propagating everything to production, through pilot roll-out, needs fewer features?

It’s important to explore various avenues for pilot strategy because there is no single silver bullet and every business has unique needs.

Below are few strategies to evaluate and finalize a Pilot Model

a. If the solution encompasses store fulfillment, choose- 

  • A store which does lower than average business transactions on a given day in a quiet neighborhood.
  • If a store with large foot falls is selected, then there is a high possibility that store personnel will be unable to attend to all in store customers which would impact business.
  • Now, that a set of stores are chosen, explore what fulfillment options will the pilot target like BOPIS (Buy Online Pickup In Store) first and then SFS (Ship from store).

    b. Similar to stores, if the solution encompasses warehouse fulfillment, choose a warehouse -

      • Which has additional personnel capacity for daily operations as learning a new system will affect their
      • Which handles less inbound and outbound transactions to avoid delays.
      • A warehouse which has more accurate physical stock mapped to the perpetual

    c. Explore opportunities like onboarding a specific segment of catalog or set of product lines or a department or a category on the new application.
        This works effectively when multiple business entities of an enterprise have individual charters, business goals and decisions get onto a common.
        However, this would need a lot more effort as business personnel need to use multiple applications simultaneously. This in turn would need more  
        effort on training and play book preparation as well. The advantage this model gets is the highest quality of risk mitigation.


      d. For large enterprises with multiple sellers in different countries, seller-based onboarding works very effectively as it eases the regular issues
           around currency conversions, taxation rules, country or region-specific linguistic needs of the targeted business persona’s,


      e. If Call Center is implemented as a solution for more than one region, geography, data center or large number of user profiles, choose a call center
          with low call volumes and only allow few call center personnel to access the application. Ensure that few users are onboarded across multiple user
          profiles though.


      To conclude, there will always be unique opportunities to derive the Pilot model based on the business’s risk appetite, operational constraints and technical maturity of the business organization. If the opportunities are not explored and evaluated, and businesses and IT teams pursue a roll-out with a big-bang approach, the risk of a severe financial impact and damage to brand equity is extremely high and could be serious.


      Stay tuned for my next blog on go-live learnings on Data Migration as a focused technical area.