A clothing label is a tiny confidence trick. It offers one calm letter or number - M, 8, 42 - as if bodies, fabrics and brands had signed the same treaty. They have not. The medium that hangs neatly in one closet can pinch, billow or trail on the floor in another. Online, where the fitting room is replaced by a product page, that mismatch becomes a business expense. The shopper hesitates, orders two sizes, or leaves. If the box arrives wrong, the retailer buys the return journey.
True Fit built a company in that gap between label and lived result. The Boston software firm collects signals about garments and shoppers, then learns from an unusually decisive event: whether a purchase was kept or sent back. Its recommendations appear inside retail sites as quick fit guidance or a personalized size. Increasingly, the same intelligence can surface in search, chat and AI shopping agents.
This sounds like a digital tape measure. It is closer to a translation system. A shopper who keeps one brand's slim-cut jeans in a 30 has supplied evidence that may help interpret an unfamiliar brand's 31. Add garment construction, material, category, body context and a preference for loose or tailored, then repeat across a large network. True Fit's premise is that the answer to “Will this fit me?” lives in the relationship between person and product, not in either one alone.
01 / Origin storyThe jeans outside the dressing room
The founding story begins with a scene ordinary enough to be overlooked. While studying at Babson, Romney Evans waited outside a dressing room as his wife struggled to find jeans that fit. He saw a data problem hiding inside the hop, tug and sigh of trying on denim. Evans emailed classmates looking for somebody who could turn the observation into a business. Jessica Murphy, a former apparel buyer who understood how clothing is designed and merchandised, responded.
True Fit was founded in 2009. Early on, it experimented with a consumer-facing shopping destination - a “test kitchen,” Evans called it in a Babson alumni note - while developing recommendation software for online, mobile and store channels. The business eventually pivoted toward retailers. That move was important. A single shop sees only its own catalog and customer history; a network can learn how fit travels across brands.
“There is technical fit and there is how a consumer chooses to wear a garment or footwear.”Jessica Murphy, co-founder and CEO
Murphy's distinction is the key to the product. Technical fit asks whether a garment's measurements accommodate a body. Personal fit asks whether this shopper wants the shoulder sharp, the waist forgiving or the trouser deliberately oversized. Fashion makes both questions unstable: two styles from the same shoe brand can warrant different sizes. A system that merely normalizes charts misses the preference hiding inside the word “fit.”
02 / The machineA genome made of decisions
True Fit calls its data foundation the Fashion Genome. The company says it reflects nearly 20 years of purchase-and-return outcomes and now spans more than 100 million shoppers, 60 million active products, 91,000 apparel and footwear brands and more than $616 billion in analyzed transaction value. Those are company-reported network figures, and definitions have evolved as the platform has grown. The useful idea beneath the big numbers is simpler: clicks show curiosity; kept products show what worked.
At full strength, the system takes in a retailer's catalog, product attributes, available sizes and historical orders and returns. A shopper may add height, weight, age, familiar brands and preferences, or connect a history through a True Fit profile. The platform maps the new item against comparable products and outcomes, then returns a recommendation and a degree of confidence. For a visitor who does not create a profile, it can still offer aggregate, product-level guidance such as whether an item tends to run small.
Retailers receive more than a widget. True Fit offers analytics about shopper preferences, category affinities, product fit and return behavior. It can be integrated through standard product-page code, a Shopify app, APIs and data services. In 2026, the company added two routes aimed at agentic commerce: its own conversational shopping agent and access to fit intelligence through Model Context Protocol, or MCP, for retailers and technology platforms building agents of their own.
03 / The customerSelling confidence to the retailer
True Fit is a B2B software company. Its paying customers are apparel and footwear retailers, brands and marketplaces; shoppers use the service inside those stores. Pricing is not public. The economic pitch is visible, though: answer sizing doubt before checkout, lift conversion, reduce multi-size ordering and returns, and collect better first-party information about what customers prefer.
Its customer pages name ASICS, APL, Lands' End, The Very Group, Silver Jeans, PacSun, Forever New, JCPenney, Moosejaw and Lucy & Yak, among others. The range matters. Sneakers, denim, children's clothing, plus-size fashion and uniforms pose different fit questions, yet the underlying workflow is the same: translate an unknown item using outcomes from known ones.
These case-study numbers are useful but need adult supervision. They come from the company and its customers, implementations differ, and registered users are often more engaged shoppers to begin with. True Fit says controlled tests across retailers typically show a 1 to 2 percent sitewide conversion lift, while fit-related return reductions can reach 40 percent when guidance is followed. The honest promise is not that every return disappears. A dress can arrive late, look different or simply lose its charm. Fit software addresses one costly reason among many.
04 / The moatThe interface is not the advantage
Fit technology is crowded with alternatives: static charts, questionnaires, review summaries, retailer-built models, body scans, virtual try-on and vendors such as Bold Metrics, Fit Analytics, WAIR, Prime AI and TrueToForm. Some estimate from body measurements. Some model the garment. Some render an avatar. A retailer can also train a recommendation system on its own transactions.
True Fit's differentiation is the longitudinal, cross-market record. It learns not only that a shopper clicked a jacket, but that people with comparable histories kept or returned it; not only how one brand labels a waist, but how that brand relates to many others. Competitors can recreate a conversational box faster than they can recreate years of network outcomes. True Fit captures the point neatly in its own evaluation guide: “The data is the model.”
That network also creates the familiar platform challenge. Retailers must contribute catalog and outcome data and trust the privacy architecture. True Fit says recommendations do not require exposing raw personally identifiable information to outside AI agents, and its MCP design uses scoped access, data minimization and revocable credentials. Those controls will matter more if autonomous agents progress from suggesting clothes to buying them.
05 / The next fitting roomWhen the product page starts talking
Generative AI gave retail a new interface but not necessarily new judgment. A general shopping bot can paraphrase reviews and inspect a chart. It may still have no memory of which size a similar shopper kept, how this fabric behaves or whether “relaxed” means pleasantly loose or comically large to the person asking. True Fit analyzed live shopping conversations and says as many as 70 percent of fashion questions concerned fit and sizing. The company saw an opening: let the language model handle conversation while its domain system supplies the answer.
Its AI Shopping Agent, announced in February 2026, is designed to intervene at moments of hesitation. A shopper can ask about waist placement, stretch, structure, inseam or hem rather than reducing the exchange to a size selector. The agent can sit on a product page, listing page, homepage or search experience. For companies that already have an agent, True Fit's MCP integration offers fit recommendations, confidence signals and product context without forcing them into True Fit's interface.
A general AI can read the label. True Fit's bet is that commerce will pay for a memory of what happened after the box was opened.The fit-intelligence thesis
This is where True Fit fits in the market: below the conversation and above the raw retail data, a specialist decision layer for apparel and footwear. It is enterprise SaaS, retail analytics and fashion AI in the same garment bag. The company is not trying to design the clothes or deliver them. It wants to own the moment when a shopper asks whether one particular thing will work.
06 / What lastsThe quiet verdict in the closet
True Fit's culture mixes merchants, data scientists and software teams across Boston, London and Mumbai. Murphy now serves as CEO; Evans leads marketing. Company interviews describe a close, experimental workplace, and Built In has repeatedly included it in Boston workplace lists. The team has raised more than $100 million in disclosed venture funding, including a $55 million Series C in 2018 and $30 million in 2021, to turn a dressing-room annoyance into infrastructure.
The lasting test is mundane. The parcel arrives. The shopper tries on the item, looks in the mirror and decides whether it belongs in the closet. No chatbot eloquence can negotiate with that moment. True Fit has spent its life collecting the verdict. Its newest opportunity is to make sure the next generation of shopping software listens.