A decade ago, processing loan applications at a Lusaka microfinance office meant a room, a stack of paper files and several pairs of hands. Today one officer with a laptop and a decent system can move the same volume before lunch. Nobody announced that change. It arrived quietly, one tool at a time.
That is the pattern most people miss. Technology rarely deletes a job in a single dramatic moment. It removes tasks. And when enough tasks are gone, the job description quietly becomes something else, often without anyone updating the title on the contract.
Three shifts that got us here
AI stopped being a research project
A few years ago, machine learning was something a large bank or a university lab experimented with. Now a shop owner in Kabwe can draft a supplier email, a school can produce a term newsletter, and a small agency can storyboard a video, all before the kettle boils. The capability did not just improve. It became cheap and available on an ordinary phone.
Software became modular and affordable
Systems that once required a serious capital budget are now assembled from parts: payments here, messaging there, records in the middle. A five-person business can run infrastructure that used to belong to a fifty-person one. That changes who your competitor is.
Work stopped being tied to a building
Remote and contract work opened two doors at once. Zambian professionals can now sell into markets far beyond Lusaka. They can also be compared, on price and output, with someone in Nairobi or Manila. The market got bigger in both directions.
What this actually does to a career
Tasks go before jobs do
Data entry, reconciliation, first drafts, basic reporting, routine customer replies. These are tasks, not jobs, and they are the first to be absorbed. A role made up almost entirely of such tasks is exposed. A role where they take up 30 percent of the week simply gets 30 percent lighter, and that time has to go somewhere else.
The routine middle feels it first
The pressure is not falling evenly. Hands-on trade work is hard to automate, and so is senior judgement, negotiation and relationship management. The squeeze is on the predictable administrative middle: the clerk, the junior analyst, the person whose day is defined by moving information from one form to another.
New roles appear faster than training does
Job titles that barely existed here five years ago now show up in adverts weekly: digital marketing lead, data analyst, cloud systems administrator, livestream producer, AI operations support. The problem is not that these roles are scarce. It is that formal training has not caught up, so the people filling them are usually the ones who taught themselves.
Every job becomes partly a digital job
The pastor schedules an SMS broadcast. The teacher marks with a spreadsheet. The nurse enters records into a system. The mechanic orders parts from a phone. None of these people would call themselves technology workers, but a real part of their competence is now digital. That floor is rising for everyone.
Where this is landing locally
Across the work we do with microfinance institutions, churches, schools, NGOs and SMEs, the same pattern keeps repeating. Loan officers spend less time on files and more on portfolio quality and client relationships. Church administrators move from paper registers to member databases and SMS. School bursars stop chasing receipts and start reading dashboards. Media teams that once needed a full crew now deliver with two people and better tools.
In almost every case, headcount did not collapse. The content of the work changed, and the people who adapted kept moving forward.
If you are building a career
- Learn the tools, but bank on judgement. Software changes every eighteen months. Knowing why a decision is right outlasts knowing which button to press.
- Get fluent with AI rather than defensive about it. The person who uses it well will outcompete both the person who refuses to and the tool on its own.
- Pick a domain and go deep. Technology plus real understanding of lending, agriculture, ministry, health or logistics is worth far more than technology alone.
- Build something visible. A portfolio, a system you set up, a campaign you ran. In this market, evidence beats certificates more often than people expect.
- Keep the human parts sharp. Explaining, persuading, handling a difficult client, holding a room. These stay scarce.
If you are running an organisation
- Budget for training, not just for software. Most failed systems we are called in to rescue were not bad software. They were good software nobody was taught to use.
- Write down your processes before you automate them. Automating a broken process gives you a faster broken process.
- Start with one workflow, prove it, then expand. Whole-organisation rollouts fail more often than they succeed.
- Redeploy before you retrench. Someone who already knows your clients and your rules is worth retraining. That knowledge is much harder to buy than software.
One honest caveat
It is worth being sober about the pace. Headlines run well ahead of reality, and in our market the real constraints are rarely the technology itself. They are unreliable power, connectivity costs, data that was never captured properly, processes that live in one person’s head, and budgets that have to compete with payroll. That is why change here tends to look gradual up close and dramatic in hindsight.
Which is also the good news. Gradual leaves room to prepare.
Where to start
If your organisation is trying to work out which parts of its operation should change and which should be left alone, that is a conversation we have with clients most weeks. Talk to us at [email protected] or call +260 976 351 166.
