Building an on-demand storage marketplace in 12 weeks

Twistag built Keepwhat, an on-demand storage marketplace that replaces traditional self-storage for individuals and businesses alike, from concept to launch in twelve weeks and to €5,000 in monthly recurring revenue by month four. Traditional self-storage asks you to drive across town, rent a fixed unit larger than you need, sign a long contract and move everything yourself. Keepwhat's thesis was that idle European warehouse capacity and people who simply needed somewhere convenient to put things had never been introduced. Twistag built the two-sided marketplace in React and Node.js on PostgreSQL and AWS, mobile-first, and deliberately deferred ERP integration in favour of a simple API boundary.

Self-storage has barely changed in thirty years. Someone moving flat, renovating, travelling for a season or simply short of space rents a fixed unit at a facility across town, signs a contract longer than they need, and moves everything themselves — twice. The inconvenience is the product. Businesses hit the same wall from a different angle: hotels accumulate off-season furniture, restaurants store equipment between renovations, retailers need overflow space before holiday peaks, and each rents from the same handful of generalist warehousing companies at premium rates. Meanwhile thousands of small and mid-sized European logistics providers had spare capacity sitting idle between contracts. The marketplace that would connect the two sides did not exist, and building one on a traditional software timeline meant months of planning before a single customer could test whether the thesis held.

Twistag compressed what is normally a six-to-nine-month marketplace build into a twelve-week sprint by scoping the MVP to two questions only: does supply match demand, and do the unit economics work.

Twistag built a two-sided marketplace in React and Node.js on PostgreSQL and AWS, with user authentication, real-time availability updates, booking flows and payment collection. The design was mobile-first for a concrete reason rather than a fashionable one — logistics providers needed to list inventory from a phone in a warehouse, and renters needed to search and book on the move.

The architectural decision that mattered most was decoupling the booking system from the logistics back end. Rather than integrating with each partner's ERP in month one, which would have doubled the timeline, Twistag built a simple API boundary and coordinated fulfilment manually. Once the marketplace proved demand existed, integration became a backlog item rather than a launch dependency. The manual work stayed contained, and it stayed informative. Twistag scoped hard rather than building broadly. Everything not required to answer those two questions was cut: no ERP integrations, no partner self-service portal, no automated dispatch. That is a decision about what to leave out, and it is the reason twelve weeks was achievable without shipping something that could not take a payment. The marketplace launched with real supply, real demand and real money moving through it, which is the only test that settles whether a two-sided thesis holds.

Stack: React, Node.js, PostgreSQL, AWS. Services: Product engineering, marketplace platform engineering, product design.

Twistag shipped the marketplace in twelve weeks. Keepwhat onboarded its first three logistics partners as beta hosts and had thirty paying customers within four weeks of launch. The marketplace generated €5,000 in monthly recurring revenue by month four — ahead of the company's own projection for month eight. Conversion from inquiry to booking ran at 20%. By month four the platform had onboarded five major European logistics providers and 100 active customers. By month six Keepwhat had shifted from proving unit economics to scaling acquisition. Handling the first hundred transactions by hand surfaced the edge cases and friction points a premature ERP integration would have designed around wrongly.

The manual fulfilment work we'd deferred became an asset, not a bottleneck.