MLM Software Development Company

What the engagement actually involves

Weeks one and two — plan configuration and validation

We read the compensation plan document and return a list of cases it does not define. Every plan document has them. Common examples:

• What happens to volume in a leg that breaks away part-way through a period.
• Whether a rank lost this period is regained at the original threshold or a maintenance threshold.
• Whether personal volume from a distributor’s own consumption counts toward qualification, and up to what limit.
• How a refund three periods later affects a rank that has since advanced twice.
• Whether commission on a cancelled order is clawed back below a de minimis amount.

You decide each one; we encode the decision as a rule. Then the configured plan runs against your volume assumptions to produce a payout ratio, which is the moment an unfundable plan gets caught. If you have not modelled that yet, the plan calculator will show you the shape of the problem in a few minutes.

Weeks three and four — brand, catalogue and replicated sites

Your branding through the back office and the distributor app, the product catalogue with commissionable volume values per item, and the replicated-site templates distributors will use — including the income-claim controls that stop a distributor publishing an earnings screenshot.

Week five — payments, tax and KYC

Checkout providers for retail orders, payout providers for commission, tax calculation for each market you sell into, and KYC verification thresholds. In South Africa this is where VAT at 15% on distributor commission invoices and EFT payout routing through local banks are configured.

Weeks six onward — UAT and the parallel run

User acceptance testing against a scripted set of cases, then the step that protects you. For a migration, we recalculate your last two or three closed commission periods on the new engine and reconcile them line by line against what you actually paid. You approve the switch only once those numbers match.

For a new launch, the same rigour applied to sample volume: a spillover placement, a rank qualifying mid-period, a refund triggering a clawback, and a payout blocked by a KYC hold.

Migration is a data problem, not a software problem

The software import is the easy part. The work is in the data, and it is predictable:

• Sponsor and placement trees that disagree. Most legacy systems store one properly and the other by convention. Both have to be reconstructed explicitly.
• Rank history that was never stored. Ranks are often derived on read in older systems, so historical rank state has to be rebuilt from volume or accepted as unknown before a cutoff date.
• Wallet balances that do not reconcile. Almost every migration finds a discrepancy between the wallet ledger and the sum of commission and payout records. It has to be resolved, not carried across.
• Open commission liabilities. These come across against the original rule set and pay out on it.

We scope this in an assessment before quoting, because the range is genuinely wide and a fixed price set before looking at the data is a price that will be revised.