Excel is a remarkable piece of software. It has been the backbone of business operations for four decades, and for good reason — it is flexible, familiar, immediately available, and for small organisations doing simple things, entirely adequate. We have no argument with Excel as a tool.
The problem is not Excel. The problem is organisations that have grown significantly, whose operations have become genuinely complex, who are managing multiple people and workflows and data relationships — and who are still treating a spreadsheet as their primary system of record because switching feels difficult and the status quo still more or less works.
More or less. Until it doesn't.
In twelve years of building management systems for South African organisations, we have seen the same pattern many times. An SME or NGO runs on Excel for years. Things are manageable. Then something happens — a key staff member leaves, a funder requests data in a format the current setup cannot produce, a serious error causes a real operational problem — and the organisation realises that what felt like a temporary workaround has quietly become a structural vulnerability.
This article is about recognising that vulnerability before the triggering event occurs, not after.
Eight signs your organisation has outgrown its spreadsheets
None of these individually is a crisis. Together, they form a pattern that most growing organisations will recognise at some point. The question is whether you recognise it while you still have the capacity to address it calmly, or under pressure after something has gone wrong.
The master file problem
There are multiple versions of the same spreadsheet in circulation and nobody is entirely sure which is current. Staff have begun naming files "Final", "Final v2", and "Final ACTUAL" without apparent irony.
Reporting takes days
Producing a management report or a donor report requires someone to spend one to three days consolidating data from multiple sources. That person has other work. Reports are frequently late.
Institutional knowledge is a single person
One person — often someone senior — is the only one who truly understands how the spreadsheet works. When they are on leave, certain things cannot happen. When they resign, there is a crisis.
Integration is a manual job
Data from your accounting system, your operations tracking, and your project management exists in separate places. Combining them requires a human to copy-paste between files on a regular schedule.
Field teams are on paper
Staff working away from the office collect data on paper or WhatsApp and someone back at base re-captures it. This introduces delay, transcription errors, and occasional loss of records entirely.
You cannot answer basic questions quickly
A manager or board member asks a reasonable operational question — how many active clients, what the current job pipeline looks like, what last month's performance was — and you cannot answer it without going to check.
Errors are discovered late
Data entry mistakes surface days or weeks after they happen, often because someone notices an inconsistency while doing something else. By then, decisions may have already been made on incorrect figures.
Growth makes it worse
Every new client, staff member, or project makes the spreadsheet more complicated rather than more useful. The system that worked at ten clients visibly strains at thirty.
The moment your data management system requires a specialist to operate it, it has crossed the line from tool to liability.
Why organisations stay in Excel longer than they should
The question worth asking is not why organisations use Excel — that answer is obvious — but why they continue using it after it has clearly stopped serving them well. In our experience, there are four consistent reasons.
The cost of switching feels higher than the cost of staying. This is the dominant barrier, and it is partly rational. Moving from a known, imperfect system to an unknown new one carries real risk and real expense. What organisations consistently underestimate is the ongoing cost of staying — the staff hours spent on manual work, the risk of errors in client-facing or funder-facing data, and the ceiling the current system places on how much the organisation can grow without adding disproportionate administrative overhead.
The problem is distributed and therefore invisible. When Excel is failing an organisation, it rarely fails dramatically. It fails in hours lost here, errors caught there, decisions made on incomplete data, and reports that go out slightly late. Nobody feels any single failure acutely enough to trigger action. The cumulative cost is real but invisible until someone sits down and calculates it honestly.
There is no obvious champion for change. The person who feels the problem most directly — usually someone in operations or data management — may not have the authority to commission a solution. The person who has the authority — usually a founder or director — may not feel the problem directly at all. The gap between problem and decision-making authority keeps many organisations in systems they have outgrown for years longer than is sensible.
Past experience with technology projects has been poor. Across the South African SME and NGO sector, there is a significant stock of burned organisations — businesses and civil society groups that commissioned a system, found it did not deliver what they expected, and reverted to Excel. That experience creates a rational reluctance to try again. The solution to this is not to dismiss the concern but to understand what went wrong and commission the next project differently.
What the actual cost looks like
One of the more useful exercises we run in our Technology Fit Assessment sessions is helping organisations calculate what their current data management approach is actually costing them. The numbers are almost always surprising.
| Activity | Typical time per month | Cost at R450/hr staff time |
|---|---|---|
| Manual report consolidation | 16–24 hours | R7,200 – R10,800 |
| Data re-entry from field / paper | 8–20 hours | R3,600 – R9,000 |
| Error investigation and correction | 4–12 hours | R1,800 – R5,400 |
| Cross-system data reconciliation | 8–16 hours | R3,600 – R7,200 |
| Ad hoc data requests not in standard reports | 4–8 hours | R1,800 – R3,600 |
| Total monthly overhead | 40–80 hours | R18,000 – R36,000 |
These are conservative estimates, and they account only for time — not for the risk cost of errors in financial or compliance reporting, not for the strategic cost of decisions made on stale or incomplete data, and not for the reputational cost of reporting late to funders or boards. For many organisations we work with, the genuine monthly cost of staying in Excel exceeds the monthly amortised cost of a proper system within the first year of operation.
The economics of switching, in other words, are usually considerably more favourable than they appear from the outside.
What a proper management information system actually does
The term "management information system" sounds abstract and technical. In practice, what it means for an SME or NGO is simpler than the terminology suggests: a system where data is entered once, in the right place, by the right person, at the right time — and where everything that anyone in the organisation needs to know can be extracted from that single source without manual intervention.
In concrete terms, that means several things working together.
A single source of truth. Client records, project data, financial transactions, and operational information all live in one connected system. When something changes in one place, the change propagates where it needs to. There are no parallel spreadsheets to reconcile, no version conflicts, no "which file is the real one" conversations.
Role-appropriate access. Staff see what they need to do their work and nothing more. A field officer enters data through a form designed for that purpose. A manager sees dashboards built for operational oversight. A director sees strategic summaries. The data is the same; the view is tailored to the audience.
Automated reporting. Monthly management reports, funder reports, and board packs are generated from live data on demand, not assembled manually from multiple sources. A report that used to take a day to compile takes twenty minutes to review and dispatch.
Workflow and process support. The system reflects how the organisation actually works — approvals, assignments, stage gates, notifications — rather than requiring staff to manage those workflows outside the system and then record the outcomes in it.
Accessible from the field. For organisations with mobile teams, data can be captured at the point of work — on a tablet or phone, online or offline — and synchronised when connectivity is available. The re-entry bottleneck disappears.
None of this requires a system the size of SAP. For a South African SME or NGO operating at the scale we typically encounter — ten to two hundred staff, a few dozen to a few thousand client records, reporting obligations to management, boards, and funders — a well-specified bespoke system or a properly configured off-the-shelf platform will deliver all of the above. The question is which approach fits the organisation's specific processes and budget, and that question requires honest assessment before it can be answered reliably.
A note on off-the-shelf versus bespoke
Not every organisation that has outgrown Excel needs custom software. The market for business software has matured considerably, and there are credible off-the-shelf options for CRM, project management, NGO M&E, field data collection, and basic operations management that are genuinely good — particularly at the smaller end of the scale.
The case for bespoke development arises when the organisation's processes are genuinely distinctive, when integration between systems that do not natively communicate is essential, when data sovereignty or security requirements exceed what cloud platforms can accommodate, or when the volume and complexity of reporting requirements cannot be met by standard tools without significant compromise.
The honest answer is that many organisations we speak to could be well served by a properly configured off-the-shelf product, and we tell them so. Others genuinely need bespoke development, and the investment is justified by the fit and the longevity of a system built for their specific context. The only way to know which situation you are in is to conduct a proper assessment — one that maps your actual processes, evaluates your real requirements, and tests those requirements honestly against what is available before recommending a build.
Find out what your organisation actually needs — before you commit to anything.
A structured half-day session with your management team. A written assessment of your current technology landscape, your most pressing process gaps, and the realistic options available to address them — including honest counsel on whether bespoke development is warranted or whether an off-the-shelf solution would serve you better. The R6,995 fee is credited in full against your first development phase if you proceed with GEMIS.
Learn about the assessment →Where to start
If you recognised your organisation in the warning signs above, the first step is not to commission anything. The first step is to make the cost visible — to sit down with the people who manage your data day-to-day and calculate, as honestly as you can, what the current approach is actually costing you in time, risk, and operational capacity.
That exercise alone is often enough to create the organisational will to act. And once the will exists, the next step is a proper assessment of what to do — not a sales call with a software company that has already decided what you need, but a structured process of mapping your actual workflows, evaluating your real requirements against what the market offers, and arriving at a recommendation that fits your context and your budget.
The organisations that do this well — that invest in honest assessment before they commit to a solution — almost always end up with systems they use, that staff adopt willingly, and that last. The ones that skip it tend to end up, a few years later, with a system they have quietly stopped trusting and a spreadsheet running alongside it that everyone pretends is temporary.
Excel will still be there if you need it. But it is a considerably better tool when it is a supplement to a proper system than when it is the system itself.