Zest is testing a different approach to local discovery and booking, raising questions about how consumers will interact with AI-powered services, with agentic commerce on the rise.
Agentic commerce continues to enjoy a buzzy ascent – at least in tech world chatter and investment… if not consumer usage. As it grows in shopping for hard goods to save time or find time-sensitive deals, a question looms if it’s applicable to other areas of commerce, especially local.
Restaurants could be one usual suspect for an agentic commerce takeover. Though it’s not landing as well in food delivery, as we recently examined, dining out has a different set of variables where agents could be welcome. These variables include a more nuanced set of criteria, such as location and ambiance.
This is what Zest is betting on. The AI-fueled restaurant discovery and booking engine wants to reinvent how users find a place to eat. It does this by tapping into consumers’ past restaurant transaction data and blending that with other signals and sentiments that it synthesizes from a variety of online sources.
From these inputs, Zest makes suggestions that are intelligent and personalized. It even creates custom restaurant maps for individual users and adds a social component so users can share those maps or follow others’. Altogether, it mixes elements of Venmo’s social transaction sharing and Snap Map.
Side Door
Zest recently went further down the agentic commerce path with its new Restaurant Concierge feature. This offers conversational interface for users to state their dining preferences. It then makes reservations through integrations with OpenTable and Resy. This brings Zest from discovery assistant to booking agent.
But perhaps most notable about this move is its user interface. The whole thing is done over text messaging. This strips the experience down to its conversational essence, putting aside the past decade+ of mobile UX best practices. All it needs to do its thing is your text or voice instructions.
This approach makes sense after quick consideration, and could be the next UX standard to take over. It’s all based on a simple principle: if you’re no longer spending time in a given app to tap and type your way to specific outcomes – an agent does that for you – does it really matter how slick the interface is?
Beyond typical app interfaces being overkill in agentic contexts, they raise adoption friction. The app universe is famously saturated, with a high bar to colonize the sacred space on users’ home screens. Could products that live solely in messaging interfaces represent a sneaky side door to adoption?
The Agent is the Product
Stepping back, Zest’s approach represents a broader mobile product philosophy that the agent is the product. This includes the oft-discussed “app killer” rhetoric that users will vibe code their way to mobile apps. Much of that is locked up in practical snags, but the nearer-term reality is text-based agentic apps.
Beyond Zest – and a few others such as Poke – this seemingly retroactive trend towards a text-based agentic UX was recently validated on a massive scale by a tech heavyweight: Meta. As we recently examined, Muse has reached 10 million+ active users, largely due to its simplified text-based UI.
This trend could gain momentum. And it makes sense, given that emerging tech often fails to gain mainstream traction because it’s built by engineers. Moreover, AI is handicapped by trust issues and other brand baggage, so its techy aspects accelerate friction. Could agentic texting be the antidote?
Back to Zest’s restaurant use case, the agentic commerce play gains additional appeal in reducing pain points. For example, you can drop it into dreaded group texts about dinner outings. It synthesizes preferences, availability and allergies… then come back with something simple and valuable: a reservation.
