Do you think the Upgradability is necessary for enterprise application system? My answer is NO!

My previous company is focusing on enterprise application system of  insurance industry, and from 2006 it started proposing product (or baseline) concept to client. But now, three years later, it has been cancelled.

The purpose from company
Proposing upgradability of enterprise application system, we can keep the client in the product line and learn business requirements from client. The cost of product upgrade will be covered by license fee and yearly service fee, the customization cost will be charged on project budget.

Benefits for client
If a client joined the road map of product upgrade, client can get the upgraded and new features free frequently. And some change requests can be pushed into the road map via negotiation. So the budget can be reduced.

It seems everything is OK for both! Is it true?

Issues for company
What’s the solution for the upgradability of enterprise application system? How can we add the customization requirements into the road map of product if we are providing common features for all global clients? How can we manage the expectation if we have one hundred clients on this way already? What is the solution for client if client doesn’t want to upgrade some features? How can we assure the quality of enterprise application system for the frequent upgrade activities?

Risks for client
If you can add customization features into the road map of product easily, how could you trust it is a standard product? The upgrade activity of enterprise application system does not like desktop software, it needs to modify source code of system, do you think it will be a smooth way for adding new features or improving existing functions in current running production? If the upgrade activity includes some additional features which are useless/conflicting for your current business, do you want to stop the frequent upgrade?

Before we kicked off the work, we still believe that all the issues and risks can be resolved and mitigated, all the expectations can be balanced finally. But actually we did not.

By now, no any enterprise application system provider can say: we can provide an upgradability system, and it can be upgraded smoothly and safely. Because this is a change-resource-code work, it means the upgrade activity is a big project, not an simple uninstall and reinstall only.

The biggest issue is that the frequent upgrade activities will impact the quality of system, and many defects will be raised once we do an upgrade, and we need 20 engineers to work one while month for resolving all major defects. Client test team also needs to execute all the test cases again for the upgrade activity, even there is no any new customization feature. You cannot image the suffering for this, because it is frequent!!!!

Finally, the cost of upgrade is over the budget.

As acted the project manager for that I don’t suggest we take this way. Of course we cannot say it is impossible, but it is expensive obviously. And another point what I want to underline is that enterprise application system is based on real business requirements, and which should be considered in advance before we modify anything. But the upgrade is uncontrollable for you, because you will find this comments from provider always: If you want to upgrade the new version system, you have to use those new features/functions, otherwise you cannot upgrade any new version system any more in the future. So actually you are importing the uncontrollable risks into your business if you take frequent upgrade for your system.

My son Xuanxuan is 16+ months old.