What this is, and how to use it
Read it once end-to-end. Then live in it: score real candidates, tick the checklist, argue about the weights with your partner.
This atlas compresses the research needed to buy your first farm in South Africa with open eyes — the regional differences, farm-type economics, water law, off-grid realities, and the due diligence that catches problems before you sign. It is written for your specific situation: first-timers, fast learners, two partners, one property, self-sufficiency as the goal, income as a requirement rather than a bonus.
Your progress in the Decision Tool and Due Diligence checklist is saved in this browser (localStorage) — same file, same browser, it remembers. Use the Export button to back it up as a JSON file.
The recommended path: don't buy for at least a year
You said you're not ready to buy yet — that is your single biggest advantage. Buyers under time pressure make expensive mistakes. Use the time deliberately:
Build the knowledge base
Work through this atlas. Take short courses (agricultural colleges such as Elsenburg, Cedara, Grootfontein, Outeniqua, or reputable online production courses). Visit at least 6–8 farms of the type you're drawn to — as a visitor, ideally as free/seasonal labour for a week. Agree the criteria with your partner before looking at a single listing.
Narrow regions, build advisers
Pick two (max three) regions. Talk to local agri-chambers, extension officers, and farm agents — learn their language. Get pre-qualification for an agricultural bond. Line up your team: agri-conveyancer, agronomist, vet, agri-accountant.
Candidates through the Decision Tool
Start scoring real properties. Full due diligence only on your top 1–2. Visit in wet and dry season. Get the 3-year records. Pump-test the borehole. Model the drought year before the bank does.
Negotiate, close, land softly
Partnership agreement signed. Transition plan with a mentor farmer for year one. Budget 12–24 months of living costs plus capital works (water, power, fencing, equipment) separate from the purchase price.
Eight golden rules for first-time farm buyers
Buy water, not land
In South Africa water belongs to the state and you get a use right, not the water. No verified, legal, sufficient water for your intended use = walk away. Everything else is decoration.
A running concern needs evidence
Three years of production, sales, and cost records — or it's a hobby farm wearing a business costume. "The records are at the accountant" with nothing produced is a red flag, not a phase.
Don't pay a lifestyle price for a business asset
Sellers and agents market "the dream". You're buying productive capacity. Price per productive hectare, per megalitre of licensed water, per bearing tree — not per hectare of veld.
Learn before you own
You're capable of learning quickly — so learn on someone else's risk. Seasonal work, share-farming, mentoring. The education you get in year zero is worth more than the farm you overpay for.
Plan for the bad year, not the average
South Africa's rainfall is erratic by design. Stress-test every candidate: one-in-eight drought, commodity price down 20%, tractor broken for a month. If the numbers survive that, they'll survive reality.
See it in wet AND dry season
Dry-season dam mud flats and green-season stocking tell completely different stories. Talk to neighbours, the local co-op, the vet, and the extension officer — they'll tell you what the agent won't.
Budget the transition separately
Purchase price is often only ~half the real cost of years 1–3: borehole work, solar, fencing, irrigation repair, equipment, restocking, and living costs while the farm learns to pay you.
Buy the neighbourhood too
Security, community, schools, a hospital within an hour, and decent neighbours matter more than 10% on price. A cheap farm in a difficult area is neither cheap nor safe.
Regions & Interactive Map
Click a province on the map (or a card below) to load its full profile. Change what the map is colouring by — climate regime, relative land cost, water stress, or off-grid fit.
Select a province on the map or below…
Cross-cutting realities (all provinces)
🦟 Livestock disease zones
Can restrict movement, sales and exports of your animals.
- Foot-and-Mouth Disease management areas (parts of Limpopo, North West, Free State, KwaZulu-Natal and Eastern Cape): cattle/beef movement out of the zones requires tests and permits; property-level controls apply. Confirm the current boundary map with DALRRD — it changes with outbreaks.
- African Horse Sickness vaccination zones cover the coastal belt — relevant if you want horses.
- Bluetongue / Rift Valley / tick-borne diseases vary by region — the local vet is your fastest source of truth.
⛏ Title & land-history checks
The kind of problem that only appears after you've spent money.
- Mining & mineral rights: the state can grant new-order mineral rights even over privately owned surface land. Search the title for mining leases/servitudes.
- Land claims: lodged or settled claims can affect transfer and future use — your conveyancer must search the deeds office and the Regional Land Claims Commission.
- Communal / trust / state land is a different legal world with different (and for first-timers, worse) risk — stick to registered freehold.
- Servitudes: pipelines, power lines, roads, rights of way — they follow the title forever.
🌵 Drought is a feature, not a bug
Assume every second decade has a brutal stretch.
- SA rainfall is among the most variable in the world; two consecutive bad seasons are normal planning inputs, not tail risks.
- Veld recovers slowly after overgrazing — carry capacity in a wet viewing week is not carry capacity.
- Groundwater levels fall during droughts exactly when you need them — ask for pump logs across years, not days.
📉 Markets & logistics
Distance to money is distance to profit.
- Distance to a co-op/depot/abattoir/packhouse and to the N-road network affects input and output costs every single week.
- Commodity prices (maize, wheat, wool, beef) are global — you can't influence them, only your cost per unit.
- Check municipal service reliability (water, road grading, refuse, electricity) — poor municipalities are a real running cost.
Farm Types: what you'd actually be running
Every type is a different business with a different skill curve. "Running concern" quality matters more than type — but some types are far more forgiving of beginners than others.
Grain & oilseed crops
Horticulture — orchards & deciduous
Citrus & subtropical fruit
Viticulture
Beef cattle
Dairy
Sheep & goats
Poultry & eggs
Game & wildlife
Forestry
Agri-tourism & guest farm
Mixed / diversified farm
Evaluating a "running concern"
What separates a business from a lifestyle property with a lawnmower.
Documents to demand (3 years minimum)
- Production records: hectares planted/harvested, yields per crop, herd/sales numbers, wool clip, milk litres
- Financial statements or at minimum income + expense summaries (co-op statements are gold)
- Input invoices & agronomist/adviser reports (they show what actually happened, incl. problems)
- Water: licence/ELU documents, pump hours, irrigation schedules, electricity bills for pumping
- Animal health: veterinary treatments, vaccination/dosing programme, mortality records
- Equipment list with age & service history; buildings condition
- Staff: contracts, wages paid, UIF, housing arrangements
- Insurance claims history (hail/drought/fire tell you the risk profile)
- Existing contracts: grain off-take, milk, packhouse, grazing leases, tourism bookings
Red flags — walk-away signals
- Records exist "with the accountant" but never materialise
- Only one good year offered as proof of performance
- Water licence not in seller's name, or licensed volume ≫ what they actually use (future enforcement problem)
- Owner says "we don't spray / don't vaccinate / don't keep records — we do it by feel"
- Equipment listed at "book value" with no service records
- Sudden seller story: emigrating, divorcing, "mining company buying nearby" — verify independently
- Heavily improved house, tired farm: you're paying for the lifestyle, funding the business
- Staff who've been paid in arrears, or fences/water points in disrepair (symptom of cash flow)
Decision Tool: gate → weights → score
Three steps, in order. The gate kills bad candidates before you score them — scoring a farm that fails a deal-breaker just makes you rationalise it.
Step 2 · Weights — what matters to you
Adjust freely; weights auto-normalize. Presets change the whole set — then tune by hand and save (auto-persists).
| Criterion | What it measures | Weight |
|---|
Step 3 · Score candidates
Score every candidate against the same yardstick. 0 = deal-breaker-poor · 3 = acceptable · 5 = genuinely exceptional. Unscored criteria count as 0, so score honestly and completely.
| Criterion |
|---|
| Weighted score (0–100) |
| Rank |
Gate status per candidate is set in its column header (PASS / not yet checked / FAIL). FAIL = rejected and excluded from ranking. Everything persists in this browser; export regularly.
Water: the variable that decides everything
In most of South Africa, this is not a farming question — it is the whole question.
Water law essentials (SA)
- Use rights, not ownership. Water uses above set thresholds require a Water Use Licence (WUL); smaller domestic, livestock and irrigation uses may be exempt but still must comply with requirements and, where applicable, be registered. Thresholds change — confirm current ones with DWS.
- Existing Lawful Use (ELU). Pre-1998 uses recorded by DWS can carry rights — but you must obtain the actual determination documents and check they match the property, the use, and the volumes claimed.
- The Reserve comes first. Basic human needs and ecological needs are prioritised above all licences — meaning a licence is a ceiling with weather attached, not a guarantee.
- Transfers take time. Assigning/transferring a water use right is an application to DWS (regional office) that can take many months — align your offer suspensive conditions with your conveyancer.
- Check title AND registers. Water servitudes and rights can be registered against the title deed, and DWS keeps its own registers — search both, plus ask about pipelines/canals crossing the land.
- Boreholes: should be sited/drilled by a registered driller and registered with DWS — an undocumented hole in the ground is not an asset you can bank on.
- Dams: larger dams fall under construction/licensing and dam-safety requirements — verify status, especially if you plan to enlarge or enlarge a wall.
Verify BEFORE any offer
- Copy of the WUL/ELU/registration documents — in the seller's name, correct property, correct volumes and conditions
- Written confirmation of transferability from DWS (or your conveyancer's water lawyer)
- Independent 24–48 hour pump test with recovery — yield in m³/hour, third-party report, in the dry season if possible
- Laboratory water quality: electrical conductivity/TDS (salinity), pH, nitrates, iron/manganese, hardness, and E. coli if it's drinking water
- Depth to water, historical levels, pump age & condition, power source & cost to pump (electricity vs solar)
- Dam wall condition, siltation, seepage, registration status
- Actual irrigated hectares vs licensed hectares vs system condition (walk every mainline)
- Municipal supply (if any): agreement, quality, interruption history
Water arithmetic you'll actually use
1 mm of water on 1 hectare = 10 m³ = 0.01 ML. That one conversion lets you size everything.
| Use | Rough water need | Per hectare per season | What it means |
|---|---|---|---|
| Dryland maize | Rainfall only (500–800 mm) | 5–8 ML effective rain | Need reliable summer rain; no licence needed, but no safety net either |
| Irrigated wheat/maize | 400–700 mm supplemental | 4–7 ML | Supplemental irrigation ≈ one solid extra rain season you must pump |
| Deciduous orchard | 600–900 mm | 6–9 ML | Frost-free + licensed water + pollination planning |
| Citrus / subtropical | 700–1,100 mm | 7–11 ML | Year-round evaporation in hot zones; never plan without margin |
| Intensive vegetables | 600–1,000 mm | 6–10 ML | Highest value per drop — but every drop must be delivered |
| Home food garden (100 m²) | ≈ 400–900 L/m²/yr with mulch+drip | 40–90 m³/yr | Modest — but a year-round supply for two people, grown well |
| Dairy (water + herd + parlour) | Extra on top of pasture | +2–4 ML | Animals drink daily — the one demand that never pauses |
Figures are broad planning ranges for discussion, not prescriptions — your agronomist refines them per crop, soil and climate zone.
Off-grid blueprint (and why to phase it)
South Africa is one of the better countries on earth for off-grid living — abundant sun, space, and growable land. The failures come from trying to do it all before you know what you're doing.
⚡ Energy — size it from loads, not brochures
- List real loads and split them: house (fridge, lights, Wi-Fi, washing), geyser (biggest single load — displace it with solar geyser or heat pump), cooking (gas), borehole/irrigation pump (its own system), workshop, fencing energiser.
- Daily energy target: a lean farm household lands around 8–15 kWh/day if heating/cooking/geyser are handled separately.
- Solar array: in SA roughly 4.5–5.5 kWh per kWp per day (better in NC/LP/north, weaker in winter/west/mountains). 12 kWh/day → plan 3–4 kWp, not 2.5.
- Battery: LiFePO4, usable capacity for 1.5–2 days of critical loads, with a generator/grid as the third day. Temperature, warranty cycles and inverter compatibility matter more than headline kWh.
- Inverter: pure sine, continuous rating above your peak simultaneous load (a typical homestead needs 5–7 kVA).
- Pumps: dedicated solar pump system (MPPT or DC) sized to well yield, not to sun hours alone.
- Do it properly: COC-compliant wiring (SANS 10142), accredited installer (PVUSA/SAQCC Solar registers), and a written single-line diagram you keep forever.
SA has excellent solar resource; the real variables are winter/cloudy-day margin and battery replacement cost (budget year 8–12).
💧 Water & sanitation off-grid
- Source diversity: borehole/spring/river (registered, licensed) + rainwater harvesting from every roof + storage. Target ≥3 days of reserve for house + animals.
- Rainwater maths: 1 mm on 1 m² of roof = 1 L. A 150 m² roof in 600 mm rainfall ≈ 90 m³/yr gross — minus first-flush losses. Useful, never sufficient alone.
- Treatment: sediment → carbon → UV/chlorine for drinking; separate garden/drink lines. Test borehole quality before you trust any filter.
- Sanitation: septic + soakaway (percolation test), conservancy tanks on clay, biodigesters, reed beds for greywater. Keep septic/greywater downhill and well away from boreholes and drinking sources (common setback guidance is ≥30 m — confirm local requirements).
- Pressure & distribution: pressure pumps, break tanks, freeze-proofing in cold zones (Free State/WC interior nights).
🥕 Food systems — sequence them
- Month 1: soil first — compost, mulch, beds. Veggies while trees grow.
- Day 1: plant the orchard/barren trees — they take 3–5 years to pay you in fruit.
- Month 3+: laying hens (eggs are the fastest self-sufficiency win); broiler cycles if you have housing.
- Year 1–2: small stock or a few cattle matched to veld — one main livestock system, not three.
- Always: preserve the surplus (drying, bottling, root cellar/cool room), and budget staples you'll still buy (maize meal/rice/oil).
- Tunnels/greenhouse only where water and skills allow — they multiply management as much as yield.
Realistic target for two people in 3–5 years: 60–80% of fresh food (veg, eggs, meat, fruit) with purchased staples and services — full food independence is a hobby farm fantasy that breaks people.
📡 Connectivity & security
- Internet: Starlink (available in SA — good rural answer, needs power + open sky), plus LTE as backup with an external antenna. Test actual signal at the house and at the far camps.
- Farm comms: VHF/UHF radios (dead zones kill coordination), plus SMS fallback.
- Security layered: electric fence (maintained energiser + solar), beams/cameras with 4G backup and local recording, lighting, dogs, farm/neighbourhood watch — and neighbours you trust.
- Power for security: everything on the fence/camera line must survive load shedding — its own small solar kit.
- Insurance: check requirements (many insurers mandate fence/alarm standards — often a useful forcing function).
Risks & Costs: what the brochure leaves out
The purchase is the small invoice. Everything below is the real one — recurring, unforgiving, and mostly invisible until you own it.
Top traps for first-time buyers
🏠 You bought a house
A charming homestead on unproductive land. Value the farm business separately from the lifestyle — and never let the house carry the price.
💧 You trusted the words "abundant water"
Agent-speak for "there's a hole in the ground". Only a test report and licence documents constitute evidence.
📋 You assumed the licence transfers
Water transfers take months and conditions attach. Make the offer subject to it.
🐑 You stocked to the green-season number
Carry capacity must be set from the dry season record — overstocking in year one is how veld and finances die together.
🔧 You bought everything at auction
Contract out (baling, harvesting, spraying) until you know what you'll use — machinery is a capital sink with depreciation.
📅 No drought plan
Feed budget, destocking triggers, and cash reserve decided before the first dry spell — not during it.
🤝 No written partner agreement
Two owners, no exit clauses. Sort it before the deed, not after the first conflict.
🔬 You skip the specialists
Agronomist, vet, hydrogeologist and surveyor cost thousands; the mistakes they prevent cost millions. Call them early.
🧾 You forget the municipality
Rates, service charges, road maintenance, inspections — and unreliable services you end up providing yourself.
🔒 Security last
The farm's risk plan belongs in month one, with the capital to match (fence, cameras, lighting, response).
🎪 Three enterprises at once
Sequence: water → power → soil/stock → income layers. Simultaneous launches eat capital and attention.
📈 You believe the "good year" story
Records across years, incl. bad ones — plus insurance claim history — reveal the true farm, not the highlight reel.
Running-cost map (budget line by line)
| Category | Shape of the cost | Notes |
|---|---|---|
| Labour | Largest controllable cost, monthly | Farm sector minimum wage is set annually (Sectoral Determination 11) — check the current rate with the Dept of Employment & Labour/CCMA. Include UIF, overtime rules, transport, protective equipment, and housing arrangements. |
| Fuel & energy | Monthly, spiky | Pumping hours × tariff, tractor diesel, vehicles, generators. Solar changes this curve — a reason to prioritise it. |
| Animal health & feed | Predictable + drought spike | Vet, vaccines, dosing, minerals, lick; feed and hay in dry years is where savings destroy herds. |
| Crop inputs | Seasonal, pre-harvest cash crunch | Seed, fertiliser, crop protection, contractor fees, agronomist — often borrowed against harvest. |
| Maintenance | Constant, ignored at peril | Fences, irrigation, buildings, vehicles, workshops. Under-maintained infrastructure is a deferred bankruptcy. |
| Insurance | Annual | Hail/crop, livestock (mortality/theft), assets, liability (tourism guests!), household, vehicles. Read exclusions — drought cover is limited. |
| Professional services | Irregular, essential | Conveyancer, accountant, surveyor, agronomist, vet, water/hydro consultants. |
| Finance | Monthly, rate-sensitive | Stress your bond at +2–3% on approval day, not at the current prime rate. |
| Security & connectivity | Monthly | Systems, servicing, internet (Starlink/LTE), radios, community levies. |
| Household | Monthly | Your living costs — including the town trips that get more expensive when you live 60 km out. |
| Contingency | 15–20% of all the above | Non-negotiable. The dry year, the failed pump, and the price dip arrive together. |
⚖ Labour & compliance basics
- Minimum wage, overtime and public holiday rules for the farm sector are set annually — verify the current sectoral determination before you forecast payroll.
- Written employment contracts, UIF registration, and records are required — "the way we've always done it" is not a compliance strategy.
- Employees living on the farm have tenure-related protections — understand this before counting on that cottage.
- OHSA basics: machine guarding, chemical store (locked, ventilated, with SDS), water hygiene, first aid, incident records.
🧾 Tax & structure (get real advice)
- Going-concern sales can be VAT zero-rated when the legal conditions are met — this can be worth hundreds of thousands; only your conveyancer + accountant can confirm it applies.
- Transfer duty vs VAT, and personal/company/trust ownership structures each have financing, liability and succession consequences — decide with an agri-accountant before offer stage.
- CGT: keep records of every improvement from day one; farms are usually sold decades later with tangled histories.
- Start proper books immediately — they also make your own farm decisions faster.
Due-diligence checklist
Everything above, compressed into work you can tick off. Saved in this browser as you go — works for one candidate at a time (start fresh per property).
0 of 0 checks complete
Project plan: from atlas to live map machine
You said big project, Cloudflare, database, scraper. Here's the sane way to grow this file into that — each phase usable on its own, no rewrites.
| Phase | Deliverable | Details |
|---|---|---|
| 1 | This atlas — single-file HTML, research + map + scoring + checklist | Done. Deploy anywhere (it's static). Content edits stay easy while it's one file. |
| 2 | Repo + Cloudflare Pages | Git repo, split content into content/*.json (provinces, criteria, checklist) so research grows without touching markup/JS. Deploy on push; preview URLs per branch. Map stays embedded or loads geojson from /assets (same-origin — no CORS pain). |
| 3 | Data pipeline + D1 database | Cloudflare Worker on a Cron Trigger fetches/normalises listing data → D1 (SQLite at the edge, free tier generous). Start with manual CSV import (you verifying by hand) — build the scraper only where allowed. |
| 4 | Live ranked map | Nightly job scores every listing with your exported weights; markers on the map by score/price/region; shortlists and notes per candidate; later: accounts (Workers + D1 + Cloudflare Access/Auth). Swap Leaflet → MapLibre GL when you want styled choropleths and performance. |
Architecture (target)
┌── listing sources ──────────────┐
│ farm portals · agents · RSS · │ respect robots.txt / terms of use
│ your own CSV shortlist │
└────────────┬────────────────────┘
▼ Cron Trigger (nightly)
Cloudflare Worker ── parse / normalise / dedupe / geocode ──┐
│ ▼
│ D1 (SQLite): listings,
│ sources, scores, notes
▼
Cloudflare Pages (this atlas + /api/* Worker routes)
▲
└── fetch: /api/listings?province=&sort=score
🗄 D1 schema (starting point)
create table listings (
id integer primary key,
source_id integer,
external_id text,
url text, title text,
province text, district text, town text,
lat real, lng real,
price_zar integer,
hectares_total real, hectares_arable real,
farm_type text,
water_source text, water_licence text,
status text default 'new', -- new|shortlisted|rejected|archived
verified_at text, -- human verification stamp
first_seen text, last_seen text,
raw json,
unique(source_id, external_id)
);
create table criteria (id text primary key, name text,
description text, default_weight real);
create table candidates (id integer primary key, listing_id integer,
name text, gate_status text, notes text);
create table scores (candidate_id integer, criterion_id text,
score real,
primary key (candidate_id, criterion_id));🕷 Scraper: build it honestly
- Check terms first. Many SA property portals prohibit automated scraping in their ToS; some block it technically. Where scraping isn't allowed, alternatives: RSS/newsletters, agent feeds, your own manual shortlist CSV (best v1 data anyway), or a paid/partner API.
- Do the boring parts properly:
robots.txtrespected, polite rate limits, identifiable user agent, caching, last-modified checks, and a dead-letter log when a page shape changes. - Normalise hard: R1 250 000 / 1.25m / "price on request" → one integer; "±45 ha", "erf 812", "plot" → separate fields; dedupe on source+external_id and near-match on title+price.
- Geocode once (town → lat/lng) and store — listings rarely carry coordinates.
- Human verification stamp: every listing shows "unverified" until one of you has seen the actual listing/agent — stale and dressed-up ads are common.
📅 Scheduled updates & area matching (the full loop)
- Cron: a Cloudflare Worker on a Cron Trigger (e.g. nightly, 02:00 SAST) fetches each source, parses, normalises and upserts into D1 on
(source_id, external_id). - Area matching: geocode each listing's town once → store
lat/lng→ assignprovince+districtby point-in-polygon using the same district geometry this atlas already embeds (done server-side at ingest). - Change detection: compare
first_seen / last_seen / price_zaracross runs → statusesnew,price-changed,sold(absent for 2+ runs). - Notify: new listings matching your saved filters arrive by email — you only review genuinely new matches.
- Serve: a Pages Function
/api/listings?district=…&sort=scorefeeds a ranked marker layer directly on this map.
⚠ Honest constraints
- Terms of service first: many SA portals prohibit automated collection and some block it technically. Build the pipeline so any source can be a manual CSV drop-in — automate only where permitted (or use partner/RSS/API feeds).
- Listings are dirty: "±45 ha", "price on request", duplicates, stale adverts. Budget most effort for parsing rules + human verification, not the cron job.
- Be a good citizen: low request rates, caching, identifiable user agent. The polite scraper survives; the aggressive one gets banned.
- Geocoding quality decides map quality: one wrong town→coordinate and a farm appears in the wrong district — keep a manual review queue for new towns.
Immediate next steps (when you're ready)
- Read the atlas fully; both of you independently set draft weights; compare and agree.
- Deploy this file to Cloudflare Pages under a name you like (10-minute job: Pages → upload folder).
- Git repo + content split (Phase 2) — I can scaffold this when you want it.
- Create a shared "candidates" sheet (CSV) with your first 10 listings from any portal — that's already data.
- Only then: decide if automation is justified — by then you'll know exactly which fields you actually use.
Sources & where to verify
Use primary sources for anything that touches money: state departments, regulators, and records — not forums and not agents' descriptions.
💧 Water
- Dept. of Water & Sanitation — dws.gov.za: licences, ELU, registers, dam safety, restrictions
- Regional DWS office for the catchment — actual transfer process and timelines
🌧 Climate & rain
- SA Weather Service — saws.co.za: station data, forecasts, climate summaries
- Provincial agrometeorological services for growing-season detail
🌱 Soil, crops & production
- ARC — arc.agric.za: soil/land type surveys, production research
- Crop Estimates Council (via Dept. of Agriculture) — yield benchmarks
- Grain SA — grainsa.co.za: crop cycles, market info
📊 Stats & markets
- Stats SA — statssa.gov.za: agricultural census & structural data
- SAGIS — sagis.org.za: sugar sector balances
- Farmers Weekly — farmersweekly.co.za: market news + classifieds
🏡 Listings
- Property24 — property24.com (agricultural section)
- Private Property — privateproperty.co.za
- Provincial/regional farm agents & auctioneers — often the real stock sits here
☀ Energy
- CSIR / GreenCape — csir.co.za · greencape.co.za: renewable resource maps (solar/wind)
- Accredited installer registers (PVUSA / SAQCC Solar) before any installation
⚖ Law, labour & tax
- DALRRD (agriculture) & provincial depts — extension offices, disease control areas, land admin: gov.za
- Employment & Labour / CCMA — labour.gov.za: farm sector minimum wage & conditions
- SARS — sars.gov.za: VAT going concerns, CGT, records
🧑🌾 On-the-ground knowledge
- Provincial extension officers — free, local, honest about soils and seasons
- Local agri chamber & co-op staff — they know who's selling and why
- Neighbours — the single most undervalued source in every farm deal