Price for the shop's month, not the software's seats

Most business software sold in Southeast Asia is still priced the way it was priced in the offices where it was first designed: per user, per month, by card, often in US dollars converted at the checkout. For a company of two hundred people that model works. Headcount is a fair proxy for the value the software provides, and someone in the finance team holds a corporate card.

For the businesses Rimbunara serves, almost every part of that model is wrong. Our median paying customer in Indonesia has three people working in it, at least one of them a relative, and takes its money in cash and e-wallet transfers that arrive unevenly through the month. Its owner may not hold a credit card. She does not think of her staff as users. She thinks about whether this month's sales will cover this month's costs.

A seat is not how a warung grows

When we priced per user, our own data taught us three things. The first was that many owners shared a single login with staff and family to avoid paying for a second seat. That saved them money and cost them control: nobody could see who had issued an invoice or changed a price. The second was that a business's use of the product had little to do with how many people it employed. A two-person textile trader issuing four hundred invoices a month got far more from us than a ten-person restaurant issuing thirty. The third was that price, not missing features, was the reason Indonesian businesses most often gave for leaving after trying a paid plan.

In 2023 we changed the unit. Plans in Indonesia and the Philippines are now priced per business, in bands based on the number of invoices issued, and staff logins are free on every plan. A business pays for the work the product does for it, and an owner can give every employee a login without a second thought. The change also improved what we know. Once each person had a login of their own, we could see how work was divided inside a small business, and that research has shaped much of what we have built since.

Prepaid, small and in the local currency

The second change was the amount and how it is paid. Our Mikro plan costs Rp 49,000 a month in Indonesia and ₱199 in the Philippines. Both are round local amounts, not conversions from dollars, and they do not move when the exchange rate does. Owners can pay weekly or monthly, from an e-wallet or in cash at a convenience-store counter, in the way many already pay for mobile data and electricity.

Weekly prepayment was the most debated part of the design. Our finance team worried, correctly, that it would raise processing costs and failed renewals. It did both. In the first year about one weekly renewal in six failed because the owner had not topped up, and we had to design a grace period and a gentle reminder in the chat app owners already use. But weekly payment matched how cash moves through these businesses, and it brought in owners who would never commit to a month in advance.

Small details of presentation mattered more than we expected. Owners in our research compared our price with things they already bought, a sack of rice, a month of mobile data, a day's pay for a helper, so we show the price as a weekly as well as a monthly figure, tax included, with no conversion and no annual commitment hidden in the small print. When we tested a discounted annual plan, fewer than 3 per cent of Mikro customers took it; owners told us that paying a year ahead felt like lending money to a software company. We kept the annual option for larger businesses, which take it readily, and stopped promoting it to the smallest.

The cost of cheaper pricing was real and should be stated plainly. About 9 per cent of our existing Indonesian paying customers moved down to the new plan within six months, and average revenue per paying business in Indonesia fell by about a quarter. A pricing change that only brings in new customers is rare, and I distrust any plan that promises one.

What we gained was a much larger base of businesses that stay. Paying businesses in Indonesia grew from 58,000 in mid-2023 to 131,000 in June 2026, and subscription revenue there rose from US$6.4 million in 2022 to US$10.1 million in 2025. Businesses that start on Mikro and grow move up the bands as they issue more invoices, which is the only form of upselling I have seen small owners accept without resentment.

The free plan as a floor, not a funnel

Free plans in software are usually designed as funnels: give away enough to create a habit, then withhold enough to force an upgrade. For small businesses I think the free plan should instead be a floor that covers what the law requires. In Malaysia our free plan includes e-invoice submission for up to twenty invoices a month. The smallest businesses can meet their tax authority's requirements without paying us anything, and they move to a paid plan when they need more, not because we have taken compliance hostage.

That position cost us a paid compliance add-on that our finance team had forecast at several million dollars a year. It has been repaid in lower churn and higher conversion, but I would defend it even if it had not been. A product that charges a micro business for the right to obey the law is teaching that business to resent the product.

The approach has limits. Priced per business, a firm with a few very large customers earns less from them than a seat model would. Our Singapore plans, sold mostly to professional firms, still include a per-user element above a certain size, because that is how those firms buy. Pricing should follow the customer's economics, and across four countries those economics differ.

For boards and investors judging small-business software, the useful question is not whether average revenue per customer is rising. It is whether the price is set in the unit, the currency and the rhythm in which the customer earns money. When it is, average revenue may fall while total revenue grows. When it is not, churn will eventually make the point for you.