TL;DR (60 seconds):
Your team submits another software request this week. The fifth this quarter, and none of the previous four solved the underlying problem. We see this pattern repeatedly: businesses spending thousands on new tools when the real issue is how work move...
Your team submits another software request this week. The fifth this quarter, and none of the previous four solved the underlying problem.
We see this pattern repeatedly: businesses spending thousands on new tools when the real issue is how work moves between people. The software request is rarely about missing functionality. It is about work that gets stuck, handoffs that fail, or information that lives in someone's head instead of a system others can access.
Process improvement vs new software comes down to this: most workflow problems stem from unclear steps, missing accountability, or poor communication between existing systems. Adding another tool often makes these problems worse by creating another place for work to get stuck.
In the next sections, we will show you how to spot when your team is asking for software to fix a process problem, the real cost of getting this wrong, and a simple method to determine whether you need better processes or genuinely new functionality. We will also cover the few cases where new software actually is the right answer.
The goal is not to avoid all new tools. It is to spend money where it will actually solve the problem.
Your team wants new software every month
The latest request sitting on your desk
Another email landed yesterday. Your operations manager needs a new project tracking tool because "the current system doesn't give us visibility into bottlenecks." The quote: R12,000 per year for fifteen users.
This follows the familiar pattern. Someone identifies a problem, researches solutions, and presents a software purchase as the fix. The logic seems sound: we have a tracking problem, this tool tracks projects, therefore it solves our problem.
But you've heard this before. The previous tracking tool was supposed to eliminate the weekly status meetings that still happen every Tuesday. The CRM was meant to stop deals falling through cracks, yet your sales director still chases quotes manually.
What the previous three purchases actually delivered
Your current software stack tells a story of good intentions and mixed results. The document management system reduced filing time but created version control chaos. The inventory platform eliminated stockouts while generating daily reconciliation headaches. The accounting integration promised real-time reporting but requires manual journal entries each month.
According to research from WBA Consulting, 70% of technology initiatives fail to deliver promised returns because they automate broken processes rather than fixing them first. [Freshworks' 2024 survey](https://ir.
Software requests mask process problems
Your accounts team wants project management software. Your sales team needs a better CRM. Operations is asking for workflow automation tools.
These requests feel urgent and reasonable. But they rarely solve the underlying problem.
Why your accounts team wants project management software
When accounts asks for project management software, they are not really asking for Gantt charts or task boards. They are describing a handoff problem.
Client work gets stuck between departments. Nobody knows who should approve the proposal revision. The brief sits in someone's inbox for three days while the client waits.
The accounts team sees missed deadlines and thinks: we need better project tracking. But the real issue is that handoffs are not defined. Who owns what, by when, and what happens next.
Software will not fix unclear responsibility. It will just give you digital confusion instead of analog confusion.
The real cost of working around broken processes
According to Freshworks research, 20% of software budgets are wasted on unnecessary business complexity. This happens because teams buy tools to work around process problems rather than fix them.
A typical accounts manager spends 90 minutes daily chasing project status, rekeying client details, and hunting down approvals. That is 7.5 hours weekly per person.
New software will not eliminate this waste. If the process is broken, automation just makes it fail faster and more consistently.
Research shows that 70% of technology initiatives fail to deliver promised returns because they automate existing chaos rather than address root causes.
When good people create their own systems
Your best people create workarounds. They build spreadsheets, maintain private email chains, and develop personal reminder systems.
These individual solutions multiply complexity.
The hidden costs of buying instead of fixing
What six software subscriptions actually cost you
Most businesses track monthly subscription fees but miss the compounding costs. A typical mid-sized business running six software tools pays R180,000 annually in licenses alone. Add training time at R2,500 per employee per new system, support costs averaging R15,000 per tool yearly, and integration fees starting at R25,000 per connection.
According to Freshworks research, 20% of software budgets are wasted on unnecessary complexity. For a business spending R500,000 on software annually, that's R100,000 lost to tools that duplicate existing capabilities or create new workflow problems.
The switching costs hurt most. When staff leave, their tool knowledge disappears. New hires need weeks to learn systems that should take days. Data gets trapped in formats that don't export cleanly. The total cost of six subscriptions often exceeds R600,000 annually when you include these hidden elements.
The productivity tax of tool-switching
Context switching between disconnected tools destroys productivity measurably. Research from BIT Services shows that employees lose 23 minutes of focused work time every time they switch between applications that should integrate but don't.
For a team of 15 people switching between five tools daily, that
How to identify process fixes that work
Most software requests mask process problems. The solution is not better tools but better workflows.
We find the real issues by talking to the people doing the work, not the people managing it.
The five questions that reveal process problems
Ask these questions in one-on-one conversations with your team:
Where do you wait? This reveals bottlenecks that no software can fix.
What do you do twice? Double handling usually means unclear handoffs between people or departments.
What information do you chase? If people spend time hunting for approvals, documents, or decisions, the process structure is wrong.
What would break if you were away for two days? This exposes dangerous single points of failure.
When do you work around the system? Shadow processes show where the official workflow does not match reality.
According to WBA Consulting research, 70% of technology initiatives fail because they automate broken processes rather than fixing the underlying workflow problems first.
Mapping your current process honestly
Document what actually happens, not what the procedure manual says should happen.
Follow one typical job from start to finish. Note every handoff, every approval, every piece of information that gets requested twice.
Count the delays. Where does work sit waiting? How long between each step? Who has to chase whom?
Most businesses discover their "simple" processes involve eight people, four systems, and three unnecessary approval loops.
Write down the shadow processes too. The spreadsheets people maintain because the system does not give them the information they need. The WhatsApp groups that bypass formal channels.
BIT Services research shows that new software often just documents existing chaos if underlying processes are not redesigned first.
Calculating what the delay costs you
Convert process problems into rand amounts.
If your sales team spends two hours daily chasing approvals, that costs R2,400 per week per salesperson (assuming R60 per hour). Scale that across ten people and you lose R1.2 million annually.
Count the customer impact too. If quotes take five days instead of two because of approval delays, how many deals do you lose
Three process fixes that beat new software
The pattern repeats across industries: teams request new software when what they need is clearer process. According to WBA Consulting research, 70% of technology initiatives fail to deliver promised returns because they automate broken workflows rather than fix them.
We see this weekly. A manufacturing client wanted project management software to track delays. The real problem: no one knew who approved what, when. A logistics company requested inventory software. The issue: drivers collected different information at each stop, making any system useless.
Here are three process fixes that consistently outperform new software purchases.
The approval process that cut project delays by 60%
A property development firm tracked 18 days average approval time for purchase orders over R50,000. The bottleneck was not system speed but unclear authority levels.
We mapped the actual decision flow. Purchase orders bounced between three managers who each thought someone else should approve. Meanwhile, urgent orders sat in email queues.
The fix took two hours to design: orders under R25,000 go to department heads, R25,000 to R100,000 need operations manager approval, above R100,000 require two signatures including the finance director.
Result: average approval time dropped to 7 days. No software purchased. The cost was one morning of mapping and two weeks of reinforcing the new process.
How standard templates eliminated the rework cycle
A consulting practice wasted 15 hours per week on proposal revisions. Partners repeatedly sent drafts back for "consistency issues" and "missing sections."
The problem was not tracking software but 12 different proposal formats across four partners. Junior staff guessed which template to use, then faced endless revision cycles.
We created three standard templates: diagnostic projects, implementation work, and retainer agreements. Each template included required sections, standard pricing language, and approval checkboxes.
Revision cycles dropped from an average of 4.2 per proposal to 1.1. The practice recovered 12 hours weekly of billable time, worth approximately R36,000 per month. Investment: three days of template design and training.
The handoff checklist that stopped things falling through cracks
A equipment rental company lost customers when maintenance schedules conflicted with bookings. The operations team blamed the booking system. The sales team wanted better CRM integration.
Neither system was the problem. The issue was handoff chaos: maintenance logged work in Excel, sales tracked bookings in their CRM, and no one checked conflicts until angry customers called.
We designed a five-point handoff protocol: maintenance updates a shared calendar, sales checks availability before confirming bookings, both teams
When software actually makes sense
Most software requests fail a basic test: the problem isn't volume, it's chaos. But three situations genuinely justify new tools.
The volume test for software purchases
Software makes sense when your process works but takes too long to execute manually. The calculation is straightforward: multiply the task frequency by the time saved per instance.
A property management company processing 500 rental applications monthly might spend 15 minutes per application on credit checks, reference calls, and document verification. That's 125 hours monthly, or R62,500 in staff time at R500 per hour. An automated screening system costing R15,000 monthly delivers clear payback.
The threshold: if manual execution costs more than R20,000 monthly and the process is stable, automation pays. Below that volume, spreadsheets and templates usually work better. According to Freshworks research, 20% of software budgets are wasted on solving problems that don't require software solutions.
Integration that actually saves time
True integration eliminates rekeying data between systems that already work. The benefit should be measurable in hours saved, not convenience gained.
An accounting practice might spend 6 hours weekly transferring client data from their CRM to billing software. At R400 per hour, that's R124,800 annually. API integration costing R50,000 to build pays back in 5 months.
False integration creates new problems. Connecting systems with different data structures or validation rules often breaks more than it fixes. WBA Consulting reports that 70% of technology initiatives fail because they attempt to connect fundamentally incompatible processes.
Automation that works after process fixes
Software ampl
Redirecting software requests into process improvements
The conversation that changes software requests
When someone asks for new software, we've found one question cuts through the noise: "What specific work is taking too long right now?"
This shifts the conversation from tools to problems. Instead of discussing features, you're mapping where time gets wasted. Most requests reveal process gaps, not software gaps.
Ask for the current step-by-step workflow. How long does each step take? Where do things get stuck? Who has to chase whom? According to WBA Consulting research, 70% of technology initiatives fail because they automate broken processes rather than fix them first.
Document what people actually do, not what the procedure manual says. The gap between official process and reality is where your improvement opportunities live.
Your 30-day process fix trial
Before buying anything, test process changes first. Freshworks research shows 20% of software budgets are wasted on unnecessary complexity that process redesign could have solved.
Pick one broken
Next Steps
Before your team asks for another tool, map where work actually gets stuck.
Start with one department where people complain most about manual work or delays. Spend two hours interviewing the people doing the job, not managing it. Ask what slows them down daily, what they have to chase, and what they rework.
Document three things: the current process (all steps), where it breaks (specific bottlenecks), and what that costs (hours per week, delays in days). Most problems will need process fixes, not software. Fix those first.
If something genuinely needs automation after the process is clean, measure the cost: how many hours does this problem consume monthly, what is that worth in salaries, and what would fixing it save annually.
Only then consider whether you need a tool, integration, or AI solution.
Most businesses have one obvious bottleneck costing £2,000-£5,000 monthly in wasted time. Find yours first.
We help established SMBs identify that one place where AI or automation genuinely makes commercial sense. If you want to find your highest-cost problem without assumptions or sales pressure, book a free 20-minute diagnosis.
About AutoSpark
AutoSpark helps established small and mid-sized businesses find the one place AI or automation is genuinely worth applying, then builds and deploys it. The method is plain: interview the people doing the work, find where work repeatedly gets stuck, rank the problems by what they cost, and only build when the maths shows a clear payback.
AutoSpark is led by Patrick Nesbitt, a CA(SA), CFA and former private-equity investor, so AI is treated as an investment rather than a trend. Not an AI audit. Not a transformation programme. A short, evidence led diagnosis of where the money is leaking and what fixing it returns.
Start here: autospark.ai