Microsoft, Google (Tracking Google’s Acquisitions), IBM and many others software company acquirers have colluded an empirical law that sets the fate of acquirees: their ex-post discount rate runs very high, at 80%.
Leaving out all-talent and all-users acquisitions, we hit upon the dreaded code rewrite, which is always a disaster waiting to happen: since developers, always running away from legacy code and its technical debt in the search of the promised land of unbuggy software, throw themselves into the abyss of the unknown taking advantage of the recently shaken power structure.
Fortunately, a well-conceived Memorandum of Understanding benefitting from the accumulated wisdom of corporate and software archeology should solve this issue including the following rules:
-
- New functionality should be written using the new tools, languages and software platforms of the acquirer (under the assumption that these new tools will improve the scalability, availability, costs or any other metrics)
- A new API encapsulating and exposing functionality between languages and platforms should be preferred (Façade, Front-Controller patterns)
- Legacy code should not be rewritten, except when:
- The benefits, measured by the KPIs of the business, are quantifiably higher than the cost of the rewrite itself, on a module-by-module basis
- In case of doubt, it’s the acquirer the one having to rewrite the acquiree code, not the other way around