Ledger defeats software in Gulf stores: what does paper's persistence impose on tech development teams?

A tour of small stores in Oman, parts of the UAE, Qatar, Kuwait and Bahrain reveals a recurring scene: the shopkeeper stands behind his counter, collects the customer’s payment, records the sale immediately in a paper ledger, and when new goods arrive, notes their quantities and dates by hand. At the end of the day and the month, someone reviews the pages and marks them to ensure the accounts are sound. The paradox observed by Matthew Troth, co-founder and product lead at O’Kabbiz, is that many of these establishments already purchased inventory and point-of-sale software, yet the programs remained parked on devices without real use, while the ledger stayed open and performed the full function.

Paper wins in the field for three solid advantages: absolute simplicity, direct trust, and high reliability.The pen and page require no prior training or instruction manuals, and what is written by hand remains visible to the owner and fully under their control. The goal is clear, with no distraction; a few pages can be flipped to reach any previous financial entry. In contrast, most tech companies rely on the traditional commercial path: build the product, launch marketing campaigns, close the deal, ignoring the fact that selling a program is one thing and building a daily habit of using it is entirely another, leaving a wide gap between paying for a license and relying on it in day-to-day operations.

This gap imposed an exceptional work style on co-founder Umid Birdhi, who spent long hours, eight to ten per day, inside a client’s store as if he were a regular employee. He watched sales activity and recorded data on paper, noted WhatsApp messages exchanged by staff to coordinate inventory, and personally entered products and trained workers. This field experience revealed to development teams real needs that do not appear in surveys or feature requests, and made testing every software function hinge on a single criterion: can the worker complete the step with the least possible cognitive load and no complexity? If the answer is no, the feature is postponed entirely.

Building AI tools for the business sector succeeds only after earning trust in the data entry and report matching stages.When the AI assistant in the system was asked for the most suitable date to issue a purchase check for the store, the assistant examined the historical sales record and suggested the sixteenth, because the fifteenth is the peak of monthly cash flows, ensuring sufficient balance in the bank. This swift financial decision was made without an accountant drafting a report or analysing numbers, but it would not have worked at all if sales had not been accurately recorded in the system from the start, proving that calling on AI to make decisions is conditioned on making data entry easy and available via phone or tablet, and matching precisely what the user entered.

What changes in practice for software engineers and product managers in the Gulf and the region is a redefinition of the required technical skill. Excellence is no longer confined to writing complex code or launching dashboards full of charts, but to understanding the simple worker’s behaviour at points of sale and stepping into the field to reduce inventory-logging steps. Startups betting on AI in retail must treat the paper ledger as the real benchmark of their software quality and stop measuring success by the number of closed deals. The decisive criterion that you, as a founder or investor in the regional market, should track is: how many stores have actually stopped opening the paper ledger and relied on your screen?