The Corner Shop Became the Checkout

The neighborhood shop has a peculiar relationship with technology. A distributor can arrive with a tablet, a sales portal and a perfectly good PowerPoint explaining both. The shopkeeper still needs to know whether the cola will arrive before Friday. A message is quicker. It is also already on the phone.

That small preference became Yalo’s large business. Founded by Javier Mata in 2016, first known as Yalochat, the company makes software that lets brands sell, market and serve customers inside familiar channels such as WhatsApp. Its platform ties those conversations to catalogs, promotions, customer records and order systems. A buyer can ask for stock, receive a suggestion and submit an order without learning a new storefront.

The shop was already open. The portal was elsewhere.

Coca-Cola FEMSA supplies a vast network of small Latin American shops. Its sales representatives traditionally visited stores, took orders and offered advice. That relationship had value, but a person has only so many hours and a route has only so many stops. In 2019 the bottler began working with Yalo on a second route: the retailer could order through WhatsApp.

Yalo’s 2022 account of the program says more than a million shops had been digitized, more than 60% of them frequently sent orders through the channel, and over 15% of FEMSA’s orders passed through Yalo. On one day, it says, the channel handled more than 40,000 orders. These are company-reported figures, but the operational point is plain. The sales representative could spend less time collecting the routine order and more time advising the shop.

The chat is just the front door. Behind it sits the less photogenic work: identifying the right store, presenting its catalog and prices, checking offers, recording the order and passing it to fulfillment. In this market, those details separate a cheerful bot from a useful sales channel. A wrong price is not an awkward conversation; it is a wrong invoice.

The sequence is a simplified reading of Yalo’s published product and customer workflows.

The first clever guess was not the winner

Nestlé Mexico offers a useful reality check. It was already selling to store owners through WhatsApp when it worked with Yalo on a product recommendation model. The team used historical buying behavior and ran a month-long A/B test. One approach scored how closely a customer matched a product; another explored ways to raise the basket with an upsell. The upsell approach performed best, with average ticket size rising about 5.3%, according to Yalo’s case study.

There is an unusually copyable lesson here. Start with a buyer who already orders digitally, use the purchase record to make a specific suggestion, and measure whether the suggestion changes the order. An attractive recommendation screen proves nothing. A controlled change in the basket is a stronger argument. The same test could also tell a business to stop sending a suggestion that nobody wants.

“The most natural way to do this was through chat.”Bruno Juanes, Coca-Cola FEMSA, translated from Yalo’s case study

From answering questions to finishing the sale

The company now describes its offering as an intelligent sales platform. Its modules span marketing, commerce, sales execution, loyalty and payments. Oris, its sales agent, is designed to understand a request, recommend products, manage promotions and take the order. Iris handles personalized marketing outreach. Agent Builder, introduced in 2025, lets a business describe a custom agent in ordinary language and connect it to its own data and processes. Earlier Yalo Studio tools include Flow Builder, Analytics and a Sales Desk for human handoffs.

This is where Yalo differs from a general chatbot vendor. It has built around a stubborn B2B problem: many small shops, changing prices and stock, recurring orders and a human sales force that still matters. A website can display a catalog. A support bot can answer a question. Yalo’s aim is to carry the transaction through the messy middle between those two jobs.

The customers are consequently large organizations rather than individual shopkeepers. Coca-Cola FEMSA, Nestlé Mexico, CBC, Grupo Rica and Coca-Cola HBC appear in Yalo’s published case studies; the small retailer is generally the end user of the brand’s channel. Yalo sells enterprise subscription services through negotiated order forms. The contract sets fees; integration requirements and messaging volume also shape the economics of a deployment.

A new continent, the same small counter

A more recent Coca-Cola HBC project took the model to Nigeria. Yalo says the bottler deployed Oris in WhatsApp for thousands of stores, with suggested orders and recommendations suited to a low-data, fragmented retail market. Its January 2026 case study reports a 20% increase in average order value and a 90% conversion rate from chat session to completed order. The figures describe that deployment, not a universal promise.

Yalo says it has digitized more than 4.4 million small businesses and processes over 100 million customer interactions each month. It raised a reported $50 million Series C in 2021 and announced a $20 million extension from Glisco Partners in December 2023. Its scale and funding make the company a serious contender in conversational commerce, alongside other messaging and AI platforms. The distinguishing bet is that the shopkeeper’s ordinary message can become a dependable sales record.

That bet has limits. If a distributor’s inventory, price rules or customer records are wrong, a fluent agent can deliver the wrong answer faster. If shopkeepers prefer a different channel, the convenience disappears. And if recommendations do not beat a control group, they are merely extra words before checkout. Yalo’s better case studies understand this: they count completed orders and basket value, not the number of times a bot sounded personable.

Mata told WIRED that the name Yalo plays on ya lo quiero and ya lo tengo - “I want it now” and “I have it now.” It is a neat promise, provided there is a working system between those two sentences. The corner shop is not asking to admire the AI. It is asking for the cola by Friday.