Most software promises to make a queue move faster. Dixa started with a stranger idea: that the queue was the problem. When four Danes founded the company in Copenhagen in 2015, they refused to call what they were building a helpdesk. They called it a "customer friendship" platform - an unusually warm phrase for enterprise software, and one that told you exactly where they thought the industry had gone wrong.
A decade later, that framing has hardened into a strategy. Dixa unifies every channel a shopper might use to reach a brand - phone, email, live chat, WhatsApp, Instagram - into a single screen, then layers AI on top to resolve the routine questions before a human ever sees them. The pitch to its 850+ customers across 42 countries is blunt: let machines handle "where's my order?", and let people handle the conversations worth having.
01 / THE PRODUCTOne inbox, and a machine that reads it
The everyday reality of ecommerce support is repetition. A large share of tickets are variations on the same handful of questions: where is my parcel, can I return this, change my address, refund me. Dixa's answer is a workspace it calls Team Hub, where an agent sees the customer's order history, loyalty status, past conversations and an AI-suggested reply on one screen - the whole product, in a sense, is the absence of tab-switching.
On top of that sits Mim, the AI agent Dixa launched in December 2023. Built on OpenAI's GPT models, Mim resolves inquiries end-to-end - refunds, order tracking, order changes, FAQs - and hands off to a human when it hits something it shouldn't guess at. Its defining constraint is also its selling point: Mim answers only from the company's own knowledge base.
"We engineered Mim to seamlessly integrate with a company's knowledge base, reducing the likelihood of AI misinterpretations."Jakob Nederby, Co-founder & CTPO
Alongside Mim runs an AI Co-Pilot for the humans - one of the first OpenAI-powered co-pilots in the category when it shipped in 2023. It drafts replies, improves tone, translates across dozens of languages, and summarizes long threads so an agent picking up a conversation isn't reading from the top.
02 / THE CUSTOMERA roster that reads like a shopping trip
Dixa made a decision most CX vendors avoid: it narrowed. Rather than sell to every industry, it aimed at mid-market ecommerce and consumer brands - the businesses where the shopping cart is the whole business and a bad return experience costs a repeat customer. The customer list looks less like a software case study than a high-street window.
The results it publishes stay in that world. Cycling brand Rapha reports a 14% CSAT lift and a 17% increase in customer retention. Homeware and gifting retailer Oliver Bonas cites 90% CSAT and roughly double the productivity per agent hour. Across the base, Dixa points to containment rates - the share of conversations closed without a human - reaching well past half on email and into the eighties on chat for some brands.
03 / THE PROBLEMThe economics of the boring 80%
Strip away the language and the problem Dixa attacks is a cost structure. Support teams are expensive, high-churn, and buried in repetitive work; at seasonal peaks, brands hire temporary agents just to keep the queue from collapsing. If AI can reliably close the routine tickets - Dixa says up to 80% of routine issues - the math on headcount, response time and holiday hiring changes.
"Our Human + AI strategy is all about harnessing the power of AI to enhance customer service teams and deliver exceptional customer experiences."Mads Fosselius, Co-founder
The framing matters. Dixa sells automation not as a headcount cut but as a way to protect the jobs worth keeping - let the bot take the parcel-tracking questions so the human can handle the upset customer, the complicated return, the moment that decides whether someone shops again. Whether that framing holds under a CFO's spreadsheet is the industry's open question. Dixa has staked its brand on the optimistic reading.
04 / THE DIFFERENCEA decade of buying before the boom
The generative-AI wave of 2023 caught most software companies flat-footed, bolting a chatbot onto an old product. Dixa had a head start it had paid for. Across 2021 and 2022 it went on an acquisition run - knowledge-management platform Elevio, then a $43M double purchase of Berlin's Solvemate (AI automation) and Lyon's Miuros (CX analytics). By the time ChatGPT made AI a boardroom topic, Dixa had the automation and analytics pieces already in the building.
That distinguishes Dixa from the giants it competes against. Zendesk and Freshworks built customer service for everyone; Intercom leans product-led SaaS; Gorgias, its closest lookalike, chases the same ecommerce niche. Dixa's argument is that it combined omnichannel, a narrow vertical focus, and years of AI groundwork into one platform - not a chatbot stapled to a ticketing system, but a stack designed around the machine from the acquisitions up.
05 / THE HANDOFFFounder to chairman, cleanly
Dixa did the rare thing in 2025: a clean leadership transition. Co-founder Mads Fosselius, who had run the company as CEO, moved to chairman; operator Christian Lohmann took the CEO seat to scale it. Co-founder Jakob Nederby stayed on as Chief Product Technology Officer, keeping a founder's hand on the product. It is the kind of handoff most founder-led startups fumble - the ones that manage it usually do so because someone decided the next chapter needed a different job description.
06 / THE MODELWhere it sits in the market
Dixa runs a straightforward B2B SaaS model - subscription pricing that scales with agents, conversation volume and AI usage, plus native integrations into Shopify, Magento and WooCommerce so the order data an agent needs is already on the screen. It sits in the mid-market gap between the sprawling enterprise suites and the lightweight tools, betting that consumer brands want something built specifically for how they sell.
The larger bet underneath all of it is a claim about where value lives. Customer service is one of the most outsourced, offshored and quietly undervalued functions in commerce. Dixa's whole thesis is that it is actually where loyalty gets won or lost - and that a platform which handles the repetitive at machine speed while keeping humans on the human moments is worth building a company around. Ten years in, 850 brands and $164M later, the argument is still being tested in real inboxes every day.