The deep digital geology that’s slowing your company down

The world of technology loves quick fixes and silver bullets. It mythologises rapid disruption and change, powered by the latest technology. 20 years ago, these stories of rapid, leap-frogging change were all about Cloud technology. 10 years ago, Blockchain was the hot new thing. Today, it’s Generative AI.

But there’s a contradiction. Walk into your average retailer, bank, airline or insurer and you’ll find a considerable and accumulating pile of so-called ‘legacy technology’. This is the (metaphorically speaking) sticky, stodgy and often smelly technology that sits behind the veneer of modern apps and websites.

In some of the largest, most long-established commercial enterprises, it’s possible to find digital technology stretching back to the 1970s. Yes, these same organisations might be testing out Microsoft co-pilot, or experimenting with open-source large language models, but these new, innovative technologies aren’t replacing the old. Instead, they’re joining them, both forming and rapidly generating the next accumulating layer: a kind of digital parallel of geological strata. 

Likewise, the blossoming of the AI age will likely be visible to future generations as a distinctive layer or transition.

Layers of legacy slow you down

Many organisations have learnt how to work with these layers, and keep their products and services available and secure for customers. But it comes at a cost. The layers slow everything down. Change becomes riskier as the layers deepen. Making fundamental changes, which ripple through the layers, means upgrading and updating fragile IT systems. Those systems can remain in a dormant state for months or years - operating, but in a static, unchanging state. 

When a change is necessary, this dormant IT system becomes a source of potentially cascading failure. This kind of change is a leap in the dark with unknown side effects.

Outages and hacks are the more visible symptom of legacy, but the slowing effect is what kills companies. If, year after year, the pace of change is held back, it’s only a matter of time before faster-moving competitors can impact revenue. These competitors may simply have the advantage of youth - they’ve yet to accumulate the digital strata.

How to manage legacy

Crucially, this must be recognised as a hard problem: Large, established companies will have to chip away at this problem year, after year, after year.

For many companies, the first priority is to slow down the accumulation. The impulse to ship new features and products (over improving, refining, even reducing what you have) doesn’t just “enshitify” - cause the gradual decline in quality of your products - it deepens the layering. Any CTO or CIO offering a quick fix is being falsely certain.

Here some ideas of how to chip away at legacy layers:

• Stop whole lines-of-business. Streamlining your business is an opportunity to instantly drop deep layers of digital legacy - just as you’d stop a line of business that doesn’t make a profit, so you should consider lines of business that have become slow or risky due to legacy.

• Be layer-aware. Modern leaders, including CEOs, CPOs and CFOs, need to know about legacy, and have a feel for the age, health and quality of the layers below the surface of the business.

• Measure and understand your ability to change. Get better at answering questions like: How long does it take for a new idea to become reality? Do you understand low level metrics like DORA as a measure of engineering workflow?

• Find the bottlenecks. Find the parts of your business where legacy has accumulated, the pace of change is slowing, yet you know you need to respond rapidly to changing markets.

• Apply creativity to modernisation and simplification. Test the application of generative AI to explore and simplify legacy code bases and documentation. Consider creative services such as Mechanical Orchard who specialise in cutting edge approaches to the modernisation and simplification of very old technology.

• Don’t reach for big-bang replacement. A common outcome to big bang technology delivery is that your shiny 2.0 sits on top of a decaying 1.0. Instead, take incremental and iterative approaches to changing your technology.

• Avoid the legacy skills trap. Legacy technology often demands outdated skills, practice and culture. Don’t become trapped where you have the skills you need for your past technology but too few of the skills you need for your future technology.