Quick answer
African public price monitoring compares equivalent offers rather than isolated headline numbers. As of 15 October 2026, AfricaProxies residential traffic starts at $0.50/GB, with no KYC and crypto funding. For authorised Jumia, Takealot or Konga observations, match variant, seller, currency, stock and delivery settings. Record permissions and timestamps before interpreting differences between markets.
Key facts
| Nigeria examples | Jumia Nigeria; Konga |
|---|---|
| South Africa example | Takealot |
| Kenya examples | Jumia Kenya; Kilimall |
| Offer keys | SKU, seller, variant, condition |
| Cost context | Currency, shipping and stock state |
| Traffic | $1.23 for 1 GB; from $0.50/GB |
What makes two offers comparable?
Two offers are comparable only when the product and fulfilment conditions match. Match model, capacity, colour, condition and seller. Record a merchant’s SKU when available; do not join products solely by a similar title. Keep an out-of-stock result separate from a purchasable offer. A listed delivery fee or destination restriction may reverse a headline price comparison.
What should a public-price record contain?
Keep source URL, observation time, original currency, headline price, shipping information, seller and availability. Preserve the local currency before calculating any converted amount. If permission limits the fields that can be collected, reflect those limits in the schema. Exclude account details, addresses and checkout data from a public-offer dataset.
How should a collection react to denial or throttling?
Stop and inspect authorization when a target returns an access denial or throttling response. Follow permitted access routes and documented limits. Do not respond by increasing concurrency or changing identities. Prefer a merchant feed, partner API or explicit collection agreement when a project needs repeated observation at scale.
Where does the proxy fit in this measurement?
The proxy controls a network-location input. A delivery postcode, user session or merchant locale may override that input. Document those settings before concluding that a price difference is geographic. Test a small authorized sample from each market and keep a control observation under the same product conditions.
Frequently asked questions
Can I compare NGN directly with ZAR?
Keep NGN and ZAR in separate original-currency columns and date any conversion used for comparison. The exchange rate does not make seller, tax, shipping or delivery conditions equivalent. For a Nigeria–South Africa retail study, retain the product variant and offer context before drawing a conclusion from converted totals.
Do all public pages permit repeated collection?
No. Public visibility and permission for repeated collection are separate questions. Review the target's terms, access rules and any authorised feed or API before defining the schedule. A Jumia, Takealot or Konga market study should pause on refusal rather than distribute the same prohibited collection across different exits.
Can a browser response include more than product HTML?
Yes. A rendered page can request images, scripts and other resources beyond the product HTML, increasing transferred traffic. Measure the authorised sample under the actual client setup before budgeting an African study. Keep the browser settings and requested resources consistent when comparing usage across two retailer pages.
Is a missing price the same as zero?
No. A missing price, inaccessible response and explicit out-of-stock state should remain separate values in the dataset. Substituting zero creates a false price signal. Retain the time and observed state for an African offer, so a reviewer can distinguish an unavailable observation from a genuine priced listing.
Where is the connection example?
The retail guide demonstrates a bounded request to an exact target you own or are authorised to test. It does not choose a retailer or collect an entire catalogue automatically. Use its connection handling alongside the merchant's permitted access method and your study's product-matching brief.
Should I enter real checkout details?
No real customer or payment details are needed for a public-offer study. Use permitted public delivery information and record the selected destination without entering another person's checkout data. If your own application needs transactional QA, use its approved test environment and test records separately from the retailer research dataset.