What We Learned Running 15,000 Stakeholders on One Platform

By Recenvy Team ·

Lessons from operating a South African business platform at scale: data quality, access control, turnaround discipline and why operations beat features every time.

What We Learned Running 15,000 Stakeholders on One Platform

Recenvy now supports organisations managing more than 15,000 stakeholders between them, from SETAs and public entities to brokerages and services SMEs. Scale teaches things that product demos do not. Here is what has held true.

1. Data quality is the whole game

Every failure we have investigated at volume traced back to duplicate or incomplete records. Two records for one employer means two histories, two invoices and one confused conversation. The controls that matter are unglamorous: duplicate detection on entry, required fields that are genuinely required, and one owner for the master register.

2. Nobody adopts a system that makes their day longer

Staff adopt tools that remove work. The features that get used every hour are the boring ones: a search that finds a person by any phone-number format, a queue that shows only what you own, a template that fills the repetitive reply. Elaborate features with a learning curve get abandoned in week three.

3. Access control is an operational feature, not a compliance tax

Teams work faster when they see less. A department that sees only its own queue makes decisions quickly. Give everyone everything and you get hesitation, duplicated effort and a POPIA exposure at the same time.

4. Turnaround time is the metric that changes behaviour

Volume dashboards are interesting. Turnaround dashboards are corrective. The moment a team can see that its average response is 3.4 days against a 2-day commitment, the conversation moves from opinion to process. Publishing that number internally does more than any escalation policy.

5. The channel does not matter, the record does

Work arrives by email, phone, WhatsApp, portal and walk-in, and it will keep doing so. Trying to force one channel fails. Converting every channel into the same tracked record with the same clock is what works.

6. Reporting must come from the transactions

Any number assembled by hand for a board pack is a number someone will eventually dispute. When reports are generated from the same records the work happens in, the debate shifts from whether the number is right to what to do about it. That shift is worth more than any single feature.

7. Phased rollouts beat perfect plans

The successful implementations started narrow: the stakeholder register, then inbound enquiries, then billing, then internal requests. The struggling ones tried to launch nine modules on one Monday. Sequence is a bigger predictor of success than scope.

8. Uptime is a feature nobody thanks you for

At volume, reliability is the product. A system holding the enquiry trail for a statutory deadline cannot be unavailable during a grant window, so capacity planning and daily backups stop being technical concerns and become operational ones.

What it adds up to

Organisations do not struggle because they lack software. They struggle because work enters through too many doors, sits with nobody in particular and gets reported by memory. Fix those three things and most other problems get smaller.

More articles