OUR DISCIPLINES
Data & Analytics
in South Africa.
Every good decision runs on data someone made usable. This is how data and analytics really works in South Africa: the roles, the market and what it takes to get it right, whether you’re hiring or building your own career.
A LOOK AT THE MARKET
The data and analytics market in South Africa.
Just about every company is sitting on more data than they know what to do with. The real challenge is being able to trust it enough to make big calls on it, and that’s what most businesses are actually struggling with right now.
In places like banking, insurance, and any business that lives under regulation, data has stopped being something you just report on. It’s sitting underneath credit decisions, fraud checks, capital models and the reports that end up on a regulator’s desk. If your data is at all compromised, you’re dealing with a fine, a bad call made at scale, or a number nobody in the boardroom can stand behind. That’s quietly changed who these businesses can actually afford to hire.
The arrival of AI raised the stakes again. Now, everyone wants a model, but a model is only ever as good as the data feeding it. What most companies are learning is that their real problem isn’t the data science at all, it’s the less glamorous stuff underneath: pipelines that actually hold up, everyone agreeing on what the numbers mean, and governance that doesn’t fall over the moment someone looks at it properly.
So the market’s really split into two. At one end you’ve got the people who can work with data that’s already clean and turn it into a chart or a report, and there are plenty of them about. At the other end you’ve got the people who can be handed raw, sensitive, regulated data and trusted to sort out the mess and get it right, and those people are far harder to come by, and paid accordingly.
THE TITLE
What 'data' in a job title actually means.
A handful of titles get thrown around fairly loosely from “data analyst” and “data engineer” to “data scientist” and “BI developer”. The reality is that the same one can mean very different jobs from one company to the next.
A “data analyst” at a retailer might spend the day building dashboards in Power BI, while a “data analyst” at a bank is knee deep in risk models. Same title, but very different worlds, and often… very different pay.
The more useful way to read a data role isn’t the title at all. It’s how close the person sits to the raw data, and how much the business is trusting them to get right. An analyst usually works with data that’s already been cleaned up and answers questions with it. While an engineer is trusted with the data itself, before anyone else can use it, so if their work is off, everything downstream is too. A data scientist is trusted to build something the business will act on, often without anyone being able to check every step of the reasoning.
If you’re hiring or working out your own next move, don’t ask “is this a data role,” rather understand which part of the chain this particular job actually owns, and how much rests on it being right.
THE ROLES
The roles, and what each one involves.
The easiest way to tell these roles apart is to look at where they sit in the data journey, and how much of it they’re responsible for. The tools will change, but that usually won’t. Here are the roles you’ll come across most often.
Analytics engineer
Sits between data engineering and analysis, turning raw pipelines into clean, well-modelled datasets the whole team can trust. A newer role, and an increasingly valued one.
Common tools: dbt, SQL, a cloud data warehouse, version control, and data testing and documentation.
Business Intelligence (BI) Developer
Builds the reporting and dashboards the wider business relies on, and the data models beneath them. Sits between raw data and decision-makers, turning messy sources into something consistent and trusted.
Common tools: SQL, Power BI or Tableau, data modelling, DAX, and ETL or ELT tooling.
Data analyst
Turns data into insight people can act on, working close to the business. Explores datasets, builds reports and dashboards, and helps teams understand what the numbers are actually saying. Judged on the clarity of the answer, not the complexity of the method.
Common tools: SQL, Excel, Power BI, Tableau or Looker, and a working grasp of statistics and data storytelling.
Data architect
Designs how data is structured, stored, governed and moved across the organisation. A senior role focused on the bigger picture, and one that matters most in large or regulated environments.
Common tools: data modelling, warehouse and lakehouse design, governance and security, and cloud data platforms.
Data engineer
Builds and maintains the pipelines and platforms that move data from source to usable. The foundation the rest of the team depends on, and usually the scarcest role to fill well. Owns reliability, quality and scale.
Common tools: SQL and Python, Spark, Airflow or dbt, cloud platforms (AWS, Azure or GCP), and data warehouses such as Snowflake, BigQuery or Redshift.
Data governance specialist
Sets the standards that keep data accurate, secure and usable across the business. Defines ownership, access and quality requirements, and helps teams meet regulatory obligations without slowing down the work.
Common tools: Microsoft Purview, Collibra or Alation, data catalogues, lineage tools, data quality platforms, and governance frameworks.
Data scientist
Builds models that predict, classify or optimise, and turns complex data into something a business can act on. Most effective when the data foundations beneath them are already solid.
Common tools: Python or R, SQL, machine learning libraries such as scikit-learn, pandas and TensorFlow, and statistical modelling.
Machine learning engineer
Takes models out of the notebook and into production, and keeps them running reliably at scale. Part data scientist, part software engineer, and in short supply.
Common tools: Python, machine learning frameworks, MLOps tooling, cloud platforms, and CI/CD for models.
Other disciplines we recruit in
Software engineering · DevOps and cloud engineering · Project and programme management and more
IF YOU'RE HIRING
Hiring someone you can trust with your data.
While it’s important to test whether a data hire can code or query, what matters more is whether you can hand them sensitive, incomplete or confidential data and trust what they do with it.
Now that’s the right place to start when looking to hire. Be honest about what the role is actually being trusted with, because a reporting analyst and a person owning your regulatory data pipeline are not the same risk, and should not be the same search.
Then test for judgement, not recall. The revealing questions are about the messy middle: what they did when the data contradicted itself, when a stakeholder wanted a number the data would not honestly support, when a model looked good but they were not sure. We assess through real situations people have handled, which shows how someone behaves when the answer is not clean, the exact moment that matters most in data.
The strongest data people are careful about what they claim, quick to flag what they are unsure of, and able to explain a limitation to someone senior without hiding behind it. In a regulated business, that honesty is worth more than another framework on the CV. Over-indexing on tools, and under-indexing on whether you can trust their judgement, is the mistake we see most.
WHY LEVELS MATTER
Understanding seniority.
Here is a top line view of what each level looks like in practice, whether you are writing a brief or working out where you sit.
-
A junior professional learns quickly and delivers well-defined pieces of work with support: writing queries, building reports, cleaning and preparing data. The good ones are curious about the numbers, ask why before they answer, and become more reliable month on month. You are hiring for potential and attitude more than finished skill.
-
A mid-level professional works independently and owns a defined area, whether that is a set of pipelines, a reporting layer or a model. They can be trusted to deliver without close supervision, and they understand why an approach works, not just that it produces an answer.
-
A senior professional designs the approach, anticipates where the data or the model will let you down, and improves how the whole team works. Seniors are measured by the problems they prevent and the trust they build in the data, as much as by the analysis they produce, and by the people they make better around them.
-
A lead or principal professional sets technical direction, holds the standard, and translates between the data team and the business. At this level the work is as much about judgement, governance and knowing which questions are worth answering as it is about the hands-on build.
Data skills that command a premium.
In data, the premium follows the work the business most needs to trust. While analysts are relatively plentiful in South Africa, rates for pure analysis are under pressure.
The money has moved deeper: from data engineers who make the whole chain reliable, machine learning engineers who can put a model into production and keep it honest, and to the governance and quality specialists who keep regulated data defensible. Strong SQL still pays everywhere, because it underpins all of it. While in banking, insurance and fintech, anyone who combines real technical skill with an understanding of compliance and risk commands a premium of their own, because that pairing is rare and the cost of getting it wrong is high.
Across all of it, the value sits with people trusted to own a problem end to end, not those doing the repeatable work that automation is already absorbing.
For businesses, that means the scarce, expensive skills are usually the foundational ones, not the flashy ones.
For professionals, earning power comes from depth in the disciplines businesses struggle to hire for and need to trust, not simply another year of broader experience.
Understanding the whole salary package
In South African data roles, the base salary is only ever half the story. Just as much of what a job’s worth sits in the rest of the package:
- Medical aid and retirement: employer contributions vary widely, and a generous one is worth real money.
- Bonuses and equity: structured bonuses in bigger firms, share options in smaller ones, which can outweigh base pay over time.
- Remote and hybrid working: now weighed as heavily as pay by many, and a genuine deciding factor.
- Room to grow: a real learning budget, exposure to harder problems, and access to better data to actually work with.
Where the balance sits depends a lot on the sector. Banks and insurers tend to lead on the stable stuff, strong retirement and medical contributions, structured bonuses, and the security that comes with a big regulated employer, but they’re often the least flexible on remote work. Consultancies pay well and put you in front of a wider range of problems and clients, though the trade-off is usually the hours. Fintechs and startups are where equity comes into play, and where remote and hybrid are almost a given, but more of the reward is tied up in the promise of what the business might become rather than guaranteed today.
None of these is better or worse, they’re just worth considering whether you’re hiring or looking to grow your career.
THE COST
Contract vs Permanent.
With data, this decision has a wrinkle most other fields don’t. Bringing someone in on contract usually means giving a person you’ve known for a week deep access to your most sensitive systems, the customer records, the financial data, the things a regulator cares about. That’s the real question with data contracting, and it’s worth sitting with before you get to day rates.
A lot of data work has a natural shape to it, a migration, a platform build, a reporting overhaul, a backlog that needs clearing, work with a real end date that you’d be mad to hire a permanent team around. For that, a contractor is the precise option: exactly the skill you need, for exactly as long as the work runs, and gone once it’s done. Yes, the day rate looks steep next to a salary, but you’re not paying for benefits, notice or loyalty, you’re paying for capability, now.
The catch is that “gone once it’s done” cuts both ways. Contractors take their knowledge of your data with them when they leave, so the businesses that use them well are deliberate about two things, who they trust with access, and making sure what the contractor learns doesn’t walk out the door with them.
For the professional, contract is a genuine choice, not a fallback. It pays better by the day, throws you at a wider variety of problems, and suits people who like getting in, fixing something real, and moving on. What you give up is security and the slow-built trust that comes from staying.
Data & Analytics salaries in South Africa.
FOR CONTRACTOR ROLES
Seniority
Indicative hourly rate
> Junior data professional
R450 – R550
> Mid-level data professional
R650 – R800
> Senior data professional
R950 – R1200
> Lead or Principal data professional
R1200 – R1600
FOR PERMANENT ROLES
Seniority
Annual cost to company
> Junior data professional
R600K – R750K
> Mid-level data professional
R900K – R1.2M
> Senior data professional
R1.4M – R1.6M
> Lead or Principal data professional
R1.7M – R2.3M
Figures are drawn from Acuity’s own placements across the South African market and are indicative ranges, not quotes. Actual pay and rates vary with specific skills, sector, location and how in-demand a role is at the time. Contractor rates are excluding VAT. Last reviewed: July 2026
IF YOU'RE BUILDING A CAREER
Making your next move a good one.
The strongest data careers are built on trust, not titles. The people who rise are the ones colleagues believe when they say a number is right, or that it is not. Read a role for how much it is genuinely trusted to own, not just what it is called, because a modest title with real responsibility for data that matters will take you further than a senior title rubber-stamping someone else’s.
When you put yourself forward, lead with judgement. Anyone can list tools, but those who stand out talk about the time they caught a problem others missed, pushed back on a misleading conclusion, or made a messy dataset defensible. That is what a good employer is actually listening for, and what we help you show.
At Acuity, we start with where you want to be trusted next, not just what you have done. We only put you forward for roles that fit, we never charge you, and nothing moves without your permission. When the right role is not live yet, we keep you in mind rather than push you at the wrong one.
COMMON QUESTIONS
Data & Analytics questions, answered.
FOR BUSINESSES LOOKING TO LEAD, BUILD & SCALE
Our analysts aren't delivering what we hoped. Do we need data scientists?
Usually not, or not yet. When a data team under delivers, the cause is more often unreliable data or unclear ownership than a lack of advanced skills. Fix what the team is working with first. We help you diagnose whether you have a talent gap or a foundations gap before you hire against the wrong one.
How do we hire someone we can trust with sensitive or regulated data?
By assessing judgement and temperament as deliberately as technical skill, and by being clear on the compliance context from the outset. We look for people who are careful about what they claim, honest about limitations, and experienced in environments where getting it wrong carries real consequences.
Should we build a permanent data team or bring in specialists for a project?
It depends whether the need is ongoing or has an end. Migrations, platform builds and one-off overhauls suit contract specialists; a growing, ongoing data capability is worth building permanently. We help you avoid sizing a permanent team around a temporary spike.
How long does it take to hire a strong data professional in South Africa?
Typically two to six weeks, longer for the scarce roles, data and ML engineers, and governance specialists, where genuinely trustworthy people are in short supply and worth waiting for.
FOR PROFESSIONALS EXPLORING
Should I specialise or stay a generalist?
Early on, breadth helps you find where you are strong. After that, the market rewards depth, particularly in the roles it struggles to staff and trust, like data or ML engineering and governance. We can help you read where your scarcity, and your earning power, really lies.
I don't have a computer science degree. Does that count against me?
Not with us, and increasingly not with good employers. Data draws strong people from statistics, finance, science and elsewhere. What matters is how you think with data and whether your judgement can be trusted, which we assess directly rather than by qualification.
What should I show to stand out?
The moments that prove judgement: a problem you caught, a misleading conclusion you pushed back on, a messy dataset you made dependable. That says more than any list of tools.
Do you place contract and permanent roles, and can you help with remote work?
Both, and yes. Contract is common in data for migrations and defined projects; remote roles, including for companies abroad, are increasingly open to strong data engineers and scientists. Tell us what suits where you are.