Fikäfi

There is a particular kind of loss that hotel accounting is not built to see. A guest finds the property, likes the room, clicks through to pay - and then the card declines, or the funds are short, or the checkout page simply times out on a hotel Wi-Fi connection somewhere over the Indian Ocean. The guest shrugs and books elsewhere. Nothing appears in the property management system. There is no cancellation to log, no no-show to flag, no line item anywhere. The booking never existed, so nobody mourns it.

Fikäfi calls these shadow guests, and the company's claim about their number is the most provocative thing it says in public. When it examined the payment layer across its client properties, it found two to three of them hiding behind every completed booking.

That figure deserves scrutiny - it is a vendor's number, drawn from a vendor's client base, and any hotel's ratio will depend heavily on its markets, its card mix and how long it holds an authorisation. But the underlying observation is not really in dispute. Payment failure in travel is common, and the failure happens in a gap between the booking engine and the acquirer where no single department owns the reporting. Revenue managers watch conversion. Finance watches settlement. Nobody watches the seam.

That framing - not a new feature, a renamed problem - is what Fikäfi is really selling. The software follows from it.

Strip away the framing and Fikäfi is a payments layer that sits between a hotel's reservation systems and its bank. It supports credit cards, debit cards and digital wallets, and says it will add local payment methods where a market demands them - a practical concession in regions where a domestic scheme, not Visa, is how people actually pay.

The entry product is Paylinks. These are secure, time-bound payment requests sent by email, SMS or WhatsApp, used for deposits, balances, advance payments and card capture. The WhatsApp channel is not decoration. In the Maldives, Sri Lanka, Egypt and much of the Gulf - Fikäfi's earliest markets - WhatsApp is where the reservations desk and the guest already talk. Putting the payment request in that thread removes a step rather than adding one.

From there the platform extends in two directions. Backwards into revenue assurance: verify a payment method at booking, charge the full stay ahead of arrival, automate the reminder sequence in between, and expose pending, expired and failed payments on a dashboard where somebody has to look at them. Forwards into loyalty: a guest-facing digital wallet holding redeemable credit, a member club for properties with no chain programme behind them, a CRM for pre-stay and post-stay messaging, and a mobile app tying booking and rewards together.

A quieter module handles partner payment automation - calculating and transferring commissions to agents and intermediaries. It is unglamorous, and it is the sort of thing that consumes a finance team's month in spreadsheets at a 200-room independent.

Fikäfi's direct customer base is independent hotels, resorts and small-to-mid groups - properties with real rate complexity but no in-house payments team. Its site lists more than seventeen properties, among them Abbey Hotel, Ayala Hotels, Bandos Maldives and Jetwing Hotels, with a geographic centre of gravity across the Indian Ocean, the Middle East, Egypt and the UK.

These are exactly the operators squeezed hardest by online travel agency economics. Commission on an OTA booking commonly runs 15 to 30 percent. A chain can absorb that because it has a points programme, an app and a brand that pulls guests back direct. An independent resort in the North Malé Atoll has none of those things, so it rents the customer relationship in perpetuity.

Fikäfi's answer is deliberately unsubtle. Rather than issue points - cheap to grant, hard to redeem, and the reason most independent loyalty schemes are theatre - it issues wallet credit. Real money, spendable on the next direct stay. The proposition to the operator is a straight arbitrage: pay the guest something in advance to come back direct, or pay the OTA more after the fact.

The company's headline client result comes from a 15-property group in Saudi Arabia, which it says recorded a 58 percent lift in direct revenue within two weeks of deployment. Two weeks is fast enough to be worth interrogating - a change of that speed reflects timing and collection mechanics rather than demand creation. Which is consistent with what the product does. It does not find new guests. It collects from the ones already there.

The visible model is transactional. Fikäfi earns on card processing, quoted from 1.95 percent, with no setup fee and no monthly minimum. Collected funds settle to the hotel's chosen bank account by SWIFT - monthly, bi-monthly, or whenever the property asks. That last detail is a small but genuine concession: most processors settle on their own cycle, and a seasonal resort's cash flow does not follow a processor's calendar.

The software - paylinks, wallet, member club, CRM, dashboards, partner automation - is bundled around that rail rather than licensed separately. It is a familiar fintech shape: give away the application layer, monetise the flow.

The second door is more interesting, and it opened in March 2026.

Elektraweb, a cloud hospitality platform combining a PMS, booking engine and channel manager, embedded Fikäfi's payment orchestration layer across its product. Elektraweb's reported footprint runs to thousands of hotels across more than 40 countries and four continents. In one integration, Fikäfi's addressable base went from a direct sales list to a platform's installed base.

Elektraweb's decision is the part worth noting. Payments look, from a platform team's whiteboard, like an adjacent feature - an API and some paperwork. They stop looking that way once you meet local acquiring rules in 42 countries. Choosing to embed rather than build is a scope judgement, and it is the judgement that gives a company incorporated in September 2025 a route to thousands of properties by the following spring.

Hotel payments is not an empty room. The closest functional comparison is Sertifi, now part of Flywire, which pairs payment capture with e-signature for reservation contracts and group agreements and is well established in pre-stay collection. Selfbook attacks a narrower slice - checkout conversion on the hotel's own website. Mews, Cloudbeds and Oracle OPERA all offer payment modules inside their own property management systems. Underneath all of them sit the processors: Adyen, Stripe, Shift4, FreedomPay, and regional acquirers such as Red Dot Payment.

Fikäfi's position is defined by three choices rather than a technical claim.

First, packaging. It sells the recovery of failed and pending payments as its own revenue line, not as a feature buried in a payments tab. The shadow guest framing is a marketing device, but it is a marketing device that reorganises how an operator reads their own numbers - and that is a real thing to sell.

Second, price transparency. Publishing 1.95 percent with no setup fee and no monthly minimum is unusual in a category where rates are typically quoted only after a discovery call. It also anchors the conversation on cost, which favours a challenger.

Third, distribution posture. By building an orchestration layer designed to be embedded, Fikäfi positions itself as infrastructure for PMS vendors rather than a competitor to them. That is a smaller headline than owning the guest relationship, and a much shorter path to scale.

Fikäfi was co-founded by Andre Privateer, who serves as chief executive, alongside co-founder Imran Saeed. Privateer's background is in hotel distribution rather than payments - more than fifteen years across digital transformation, commercial strategy and development work with owners and hospitality groups in Europe, Asia and the Caribbean. He also holds a role as venture partner for travel technology at Machinelab Ventures, which gives him an unusual double perspective on the sector: the operator's problem and the investor's pattern library at the same time.

That distribution background shows in the product. Fikäfi's entire thesis is a distribution person's complaint - you spent the money to get the guest, so why did nobody check whether the guest paid?

Then there is the name. Fika is the Swedish institution of stopping work for coffee, deliberately and without apology. It is a peculiar flag to plant in fintech, where the vocabulary tends toward velocity. But the company treats it as a product specification rather than a mood board. Its stated values are blunt to the point of being unfashionable: no sales pressure, value-driven solutions, adaptable technology, security first.

The argument behind them is that hotel software accumulated complexity nobody asked for. Walk into a back office and count the systems required to answer one question - did we get paid? PMS, channel manager, booking engine, terminal, acquirer portal, reconciliation spreadsheet. Six places for one number. Selling less, in that context, is a harder promise than selling more.

Most of the noise in hospitality technology is currently about guest experience - AI concierges, dynamic personalisation, contactless everything. Fikäfi is working in a duller register. Whether the card on file still works. Whether the deposit arrived. Whether the commission split was calculated without somebody rebuilding it in Excel.

None of that demos well at a conference. All of it appears in a bank account, which is a useful property for a vendor to have when hotel technology budgets tighten.

The company's exposure is equally clear-eyed. It is young, it is small, its published client numbers are its own, and its scale case currently rests heavily on a single platform partnership. Payment orchestration is a category with well-funded incumbents who can match a price. And the shadow guest framing, powerful as it is, is a claim the industry has not yet independently tested at scale.

What Fikäfi has done is pick an unattended seam and name it. In a market where most vendors promise to bring more guests through the door, a company insisting you first collect from the ones already standing there is at least asking a different question.