Legacy software challenges
Ford vs Tesla: Lessons on software challenges in legacy organisations. Photo Credit: eyeem.

Legacy software challenges

Legacy: Why are legacy organisations' software challenged?

As things start slowing down over the summer, I am looking forward to a summer break, i find myself with more time to catch up on a backlog of reading, shorts, videos, podcasts all stored up.

The genesis of this article is credited to a short on youtube posted by @howtoactuallyinvest where the Ford CEO talks about the benefits that Tesla has over Ford through integration of their software.

Here is the link to the short - Ford CEO on software

The insights from the short applies to many a business organisation out there. A series of all rational decisions made by leaders in the business through the years leads to the current status of software is in the business.

Further, in trying to manage costs, or by having (over-)zealous procurement folks (who rightly as they are asked to do so...) focusing only on price (at the cost of flexibility that maybe needed in the future) the businesses may create strategic inflexibility.

Many legacy businesses have the same set of challenges:

  • Many modules "farmed out" to suppliers because you could "bid them against each other" - be it buying commercial software or use a third party system integrator to write their software
  • Each module has their own software and infrastructure stacks that need managing effectively on a day to day basis (based on the suppliers' preference) - does not matter if it is licensed on-prem or software-as-a-service
  • In the modern world, they also likely have their own data models and trying to get an integrated view to make transactions flow or have a repository to do analytics requires quite some complex data transitions
  • Integrating a variety of "best of breed" (not really sure that applies any more) software to make the information flow effectively requires complex integrations & integration management - each of these integrations are their own software stacks even if they are modern APIs or written in micro-services let alone previous point to point data transfer protocols

Ok that is the situation and we understand the complicatedness that provides. So how do they find a way out of this labyrinthian challenge?

Before we consider that it is important to consider other challenges that one is likely to encounter in reducing the complicatedness. Typical organisations face a few other issues:

  • Some of the legacy software is so ancient (I sometimes say that some of the software is written in "hieroglyphics" - fast to run, impossible to (de-)code), that there are not many folks who know the language the code is written in and at times even those don't understand the logic it codes
  • Businesses cannot pause while software is getting "fixed" - so there is the need to fix it while delivering the software change required by the business, some on a time critical basis (e.g. for competition, regulatory commitments)
  • Any business case for change is challenged - rarely have I seen a (sensible) business case with payback less than five years to make any legacy modernisation work

All of these lead to progressive management after management fixing some of the basics, but delaying the complicated uncoupling and decoupling of legacy and leaving many businesses running on unmodernised software. These businesses therefore expose themselves to risk against competitors who is unencumbered and can therefore respond effectively to the challenge.

So are legacy organisations doomed with ancient software - what is the solution space?

I don't think so. I have met and supported many progressive business and technology leaders who have understood the necessity of progressive modernisation while navigating the complex constraints.

Here are some of the lessons I take away from these interactions:

  1. Engage proactively across the leadership team to face into the risks & tradeoffs between modernising software and delivering incremental change on the old software which progressively is more and more complicated
  2. Most of the tangible progress, & value generation i have observed, is when there is relentless focus on progressive improvement every time any module in legacy software is changed.
  3. Institutionalise the need for change - its not sufficient to have one or two evangelists. To move to a better future - it is critical to make progressive modernisation & eliminating complexity driving legacy a core to any business strategy & investment prioritisation process.

In doing so, technology executives and techno-literate business leaders have a lot to contribute:

  • Avoid jargon and lean into explaining the complicatedness in a simple way and while it maybe great to paint a modern vision for the future to guide what could be - that maybe too high level to get sufficient case for investment
  • It helps to have senior business & technology leaders who understand the challenge and are willing to engage in technical details to understand the constraints (avoiding blame)
  • Keep it simple - a 10-year, 5-year program maybe daunting. Break it into clear way-points that everyone can align on - and make progress - every day.

Many a legacy organisation has the "north star", "future state", "blueprint for the future" architecture map to improve legacy - the ones who are better off are the ones who focus on actions rather than theoretical plans which do not stand the first test of business readiness to change.
J-P (Jo?o-Paulo) Martins

Strategy | Effectiveness | Efficiency | Change

1 年

Yet Tesla had nothing when they started - they built from scratch. So why didn't Ford try to build an EV business from scratch? Trying to shoehorn old systems and value chains into new capabilities, rather than doing blank sheet design, is always going to run that risk. And then you get into some interesting cost of capital and risk arguments - because ground up new ventures may have different 'natural owners' to incumbents. The #EnergyTransition forces this choice on incumbents seeking to grow in new technologies. It's not just software, but supply chain, organisation, governance - Operating Model in fact. Perhaps incumbents need to design their blank sheet operating model, and then ask 'where do current assets/capabilities meet these needs exactly?' (ie where would our incumbent business, systems, assets be the best place to source this element of the Op Model?)

回复

要查看或添加评论,请登录

社区洞察

其他会员也浏览了