Every other week someone asks us which digital skill they should learn. The honest answer is that the question is slightly wrong — nobody pays for a skill, they pay for a problem going away. But some skills sit much closer to an expensive problem than others, and those are the ones worth your evenings.
Here is what we actually see clients paying for, what it pays here, and what we would not start today.
What the market is actually buying
We quote work every week. The pattern is consistent: the money is in things that either bring customers in or stop money leaking out. Anything decorative gets negotiated down; anything attached to revenue does not.
- Paid ads management. A business spending 60,000 shillings a month on Meta ads will happily pay 20,000 to someone who makes that spend work harder. The maths sells itself.
- Web development. Not templates — the ability to take a business requirement and ship something that works on a phone in Kisumu on a bad connection.
- Short-form video editing. Demand has risen every year we have been counting, and genuinely good editors are scarce.
- Copy that understands search. Writing is crowded. Writing by someone who knows what a person typed to get there is not.
- Bookkeeping and systems automation. Unglamorous, deeply sticky, and almost nobody under thirty is learning it.
What it pays, honestly
These are ranges we see in practice, not aspirational numbers from a YouTube thumbnail. Year one, working locally, starting part time:
- Short-form video editing: 15,000 – 45,000 KSh a month, rising quickly once you have a showreel.
- Paid ads: 15,000 – 30,000 KSh per client per month. This scales better than anything else here, because three clients is nowhere near three times the work.
- Web development: 40,000 – 120,000 KSh per project once you can finish one unsupervised. Getting to unsupervised takes most people eight to twelve months.
- Content writing: 2 – 8 KSh per word locally, and considerably more for foreign clients.
The gap between the bottom and the top of each range is almost never talent. It is whether you can show finished work and talk about what it changed.
The highest-return thing you can do in your first year is finish three projects properly and write up what each one changed. Not ten half-projects. Three.
What we would not start cold today
Generic graphic design. Not because design is dead — we run a design studio — but because the entry level is saturated and generative tools have taken the bottom rung off the ladder. Enter through brand strategy or motion instead.
"Social media manager" as a title. The work is real; the title is not sellable on its own. Attach it to ads or to video and it becomes a service someone actually buys.
Dropshipping. The version sold in the courses does not survive contact with Kenyan logistics or customs.
The part nobody mentions
Whichever you pick, two things decide whether it works, and neither of them is the skill.
The first is a connection you can rely on. A client call that drops three times does not get rescheduled — it gets given to someone else. If you are somewhere fibre has not reached, solve that before you solve anything else. It is the cheapest problem on this list to fix and the most expensive to ignore.
The second is being findable. A WhatsApp number is not a portfolio. Two finished case studies on a page you control will out-earn a year of posting.
Pick one thing from the first list. Give it six months of real evenings. Then charge properly.