Techbanq

Software · legacy drag

The old system is not broken — it is being worked around

Nobody filed a ticket saying the system is slow; they opened a spreadsheet beside it and moved on. Years later the real process lives in those workarounds — double entry, manual checks, error loops — and that unwritten process, not the ageing software alone, is what drags every order, every close and every report.

Enterprise delivery credentials

  • ISO

    ISO/IEC 27001

    ISO/IEC 27001 certified

Certifications & registrations

The quiet substitution

How the workaround quietly became the real process

We do not have an office in Saudi Arabia

We serve organisations there from our Riyadh office and our engineering hubs in Lahore and Karachi, travelling for the parts of an engagement that need to happen in person.

A legacy system rarely announces its decline. What happens instead is quieter: the day it cannot do something, somebody capable builds a bridge — a spreadsheet that tracks what the system cannot, a re-keying step between two screens, a checking ritual because the numbers stopped being trusted. Each bridge is sensible on the day it is built. Years later, the operation runs on the bridges. New staff are trained on the workarounds before they are trained on the system, and the process documentation — where it exists — describes a workflow nobody has followed in years.

This is why the drag is so hard to see from above. Every individual step has an owner and a reason, so no single report says the operation is slow — it is only visible end to end, in how long an order takes from placement to invoice, or how many hands touch a record between entry and report. And it is why asking people what the system cannot do gets a shrug: they solved that, personally, years ago. The workarounds are not noise around the problem. They are the most accurate specification of it that exists.

Reading the symptoms

The four places legacy drag surfaces before anyone names it

None of these appears in a system log. They appear in how the operation feels, and each is worth describing precisely before deciding what to do about the system underneath.

  1. 01

    Cycle times nobody chose

    The time from order to invoice, or request to approval, is set not by the work but by the handoffs between system and workaround — a record waits for the person who re-keys it, then for the person who checks the re-keying. Walk one transaction end to end with a stopwatch of days, not minutes, and the drag becomes visible without a single metric from the system itself.

  2. 02

    The double-entry habit

    The same fact typed into two places is not just wasted effort; it is a divergence machine. The two copies drift, someone notices at month end, and a reconciliation ritual grows to manage a problem the double entry created. Where you find people comparing two screens, you have found a seam the system could not cross.

  3. 03

    Error loops that everyone budgets for

    A percentage of orders always comes back for correction; a report is always checked before anyone acts on it; a batch always fails on certain records and someone always fixes those by hand. When rework is planned for rather than investigated, the error loop has been promoted from defect to process — and its cost has vanished into the baseline.

  4. 04

    Knowledge that lives in two heads

    The workarounds have owners, and the owners have memories no document shares. The operation is one resignation away from losing not the system — which will keep running — but the unwritten process that makes its output usable. This is the risk that turns a slow system into an urgent one.

Why replacements stumble

A replacement that ignores the workarounds replaces the wrong system

The recurring failure of replacement projects is that they are specified from the official process — the one in the documentation, the one the old system was built for — rather than the real one running in the workarounds. The new system arrives, faithfully automating a workflow nobody follows, and within weeks the spreadsheets are back, now bridging the new system instead of the old. The project delivered software; the operation kept its drag.

The corrective is unglamorous: audit the workarounds first, and treat each one as a requirement wearing overalls. Every spreadsheet is a feature the system lacked; every double entry is an integration that was never built; every checking ritual is a data-quality problem with a human patch. A modernisation scoped from that map fixes what actually slows the operation — and it often turns out smaller than the rewrite everyone assumed, because much of the drag concentrates in a handful of seams.

The options

Relief in place, wrap and extend, or staged replacement

A rewrite is the last option on this list, not the first. Each row is a genuine path, and which one fits depends on where the workaround map says the drag concentrates.

OptionWhat it involvesTime to first reliefDisruption to the operationMain riskBest for
Relief in placeFix the worst seams without touching the core: automate the re-keying, integrate the spreadsheet that matters most, repair the data quality behind the checking rituals.Short — the first seam closed is felt within the quarter it is fixed.Minimal; the system people know keeps running unchanged.The core keeps ageing; this buys time, it does not stop the clock.Drag concentrated in a few seams around a core that still does its job.
Wrap and extendPut an interface boundary around the old system and build new capabilities beside it, so it becomes one component in a modern estate rather than the whole estate.Medium — the boundary comes first, then each new capability lands separately.Low; the old system keeps its job while new work stops going into it.The boundary must be honest — a leaky wrapper couples the new estate to the old system’s quirks.A core that works but blocks everything new the business wants to build.
Staged replacementReplace the system slice by slice, each slice scoped from the workaround map, with the old system carrying everything not yet moved and a way back per slice.Longer to the first slice, but each slice lands as finished relief rather than waiting for a distant go-live.Managed; the operation feels a series of small changes, never one switchover.Discipline — a staged plan that stops being staged becomes a big-bang rewrite with extra steps.A core that is itself the problem: wrong data model, unmaintainable, or unsupported.

Our role

What we examine before recommending you replace anything

  1. 01

    The workaround audit

    We sit with the people who run the workarounds and map them — every spreadsheet, re-keying step and checking ritual — because that map, not the system’s age, is the specification of what a fix must deliver.

  2. 02

    The cheapest honest option first

    We rank relief in place, wrap-and-extend and staged replacement against your map, and we say plainly when the smaller option is enough — a firm that only ever recommends the rewrite is selling the rewrite.

  3. 03

    Execution that keeps the operation moving

    Whichever path is chosen, the constraint is the same: the business runs on this system today, so every change ships in a reversible step and the old system keeps its job until the new one has earned it.

How much change the operation can absorb per quarter is your call; we bring the map that shows where the drag actually lives, and options priced in disruption as well as money.

Common questions

What buyers ask

Our team has built workarounds for years — will fixing the system break them?+

It should retire them, not break them — and the difference is sequencing. Each workaround exists because it does a job, so it can only be removed when the system genuinely does that job instead. We map the workarounds first, fix the underlying gap, run the two in parallel until the people who own the workaround trust the fix, and only then retire it. A project that switches the workarounds off before the system covers them does not remove the drag; it removes the coping mechanism and leaves the drag.

How do we tell whether the old system needs replacing or just relieving?+

Look at where the workarounds cluster. If they gather at the edges — getting data in, getting data out, connecting to other systems — the core is probably sound and relief in place or a wrapper will remove most of the drag for a fraction of a rewrite’s cost and risk. If they compensate for the core itself — the data model cannot represent how you now do business, records need constant correction, the platform is unsupported — then relief is a payment on a debt that keeps growing, and a staged replacement is the honest answer. The workaround map makes this an observation rather than an opinion.

Why did our previous replacement project fail while the old system outlived it?+

Almost always because it replaced the documented process rather than the real one. The old system plus its workarounds encode years of exceptions and unwritten rules; a new system specified from the official workflow reproduces none of them, so at go-live it cannot actually run the operation, confidence collapses, and the organisation retreats to what works. The old system survives not because it is good but because it — with its human scaffolding — is the only complete description of the business. The fix for the next attempt is to treat the workarounds as requirements and replace in slices small enough to retreat from individually.

Map the drag with us

A working session that maps the workarounds around your system and ranks the options — relief in place, wrap and extend, or staged replacement — by disruption and time to first improvement.

Start the conversation