{"slug": "contracted-arr-considered-harmful", "title": "Contracted ARR considered harmful", "summary": "The article argues that \"Contracted Annual Recurring Revenue\" (CARR) is a misleading metric used by startups to inflate their financial figures, often reporting revenue from multi-year contracts with exit clauses as if it were immediate cash flow. It highlights cases like AI startup 11x, which reported nearly $10M in CARR but had only about $3M in actual recurring revenue, and warns that the media amplifies these inflated numbers, harming reporters, employees, customers, and founders who rely on them. The author concludes that CARR is simply a dressed-up version of the already weak ARR metric, optimized for fundraising rather than reflecting true business health.", "body_md": "There’s a new acronym floating around AI fundraising decks: CARR. Contracted Annual Recurring Revenue. It sounds more rigorous than ARR, almost like someone in finance finally got involved – but it is, in practice, a way to make the number bigger while sounding more disciplined about it.\nIt means that if you signed a contract in March, the start date is June, you raise funds in March or April and you get credit for revenue that’s committed but not live yet. That sounds somewhat OK. but it’s used far far more insidiously than that.\nIf a company signs a three-year enterprise deal, and year one is discounted at $1M, year two steps up to $2M and year three is full price at $3M – CARR reports $3M as the recurring part. It is not in any meaningful sense a three-year contract, because it very likely also has exit clauses.\nScott Stevenson modeled this out: by quarter five, a company running this playbook is trumpeting it’s ~$100M in “ARR” to the media press while actual cash-generating revenue sits around $35M – which is a 3x inflation baked into the metric before anyone questions the definition.\nDoes someone question the definitions? Maybe the better investors, but the media really doesn’t. The contracts behind those numbers are often not real commitments.\nLast year TechCrunch took out a hit piece on 11x, an AI sales automation startup backed by Benchmark and a16z that reported approaching $10M in ARR. TechCrunch’s reporting found that only about $3M came from contracts that survived past a three-month trial period and 70-80% of customers were churning. ZoomInfo (listed as a customer on 11x’s website) had run a one-month pilot, concluded the product performed worse than their human SDRs, and left.\n11x said that they used “contracted ARR (CARR)” and their investors knew, so when 11x raised at a $350M valuation – questions were brought up as to how this happened.\nI’ve also heard recently about startups using “pARR” in pitches. While I think this was a joke, it could actually make it’s way into real pitches if we don’t put a stop to using absurd metrics like this. It’s barely an exaggeration. “If I make $300 in three hours, I can call it $100/hour, then annualize it to $800K+ ARR!”.\nMy big problem is the press amplifies all of it. TechCrunch (apart for some hit pieces as mentioned) regularly runs headlines about startups hitting $10M ARR in three months, the number travels through LinkedIn, through pitch decks, through board presentations. It picks up credibility at each stop and loses so much of it’s nuance at each stop, quickly becoming slop revenue.\nAnnounce inflated CARR, get the press pickup, use the headline to close the next round and before anyone figures it out, the lie has already made it around the world and you’ve raised 3 more years of runway. The metric is optimized for fundraising velocity and you’re being duped in the process.\nWhy do I care?\nThe VCs obviously don’t care, because they do their own due diligence. They see a $20M CARR number, they open the books, and they find it’s around $2M so they adjust. The $20M-in-CARR-$2M-in-ARR company is a known species in venture now, there’s quite a few of them around you.\nThe people who get used are the reporters who parrot the numbers without asking what “contracted” means or how the ARR is defined. The SaaS CFO recently put out a post about how many variants of ARR he sees.\n- If you’re a reporter and you put out these numbers as-is, you’re hurting others by lying.\n- If you’re an employee who join based on a growth narrative that isn’t real, you’re not getting the full picture.\n- If you’re a customer who buy based on market position that’s inflated, you’re being duped.\n- If you’re a founder and you see others lie, you’ll do it too, and you’re going to hurt others.\nCARR isn’t a better version of ARR. It’s ARR with a suit on. And ARR was a weak metric to begin with.", "url": "https://wpnews.pro/news/contracted-arr-considered-harmful", "canonical_source": "https://arnon.dk/contracted-arr-considered-harmful/", "published_at": "2026-04-19 14:20:07+00:00", "updated_at": "2026-05-22 16:11:42.440351+00:00", "lang": "en", "topics": ["startups", "venture-capital", "artificial-intelligence"], "entities": ["Scott Stevenson", "TechCrunch", "11x", "Benchmark", "a16z"], "alternates": {"html": "https://wpnews.pro/news/contracted-arr-considered-harmful", "markdown": "https://wpnews.pro/news/contracted-arr-considered-harmful.md", "text": "https://wpnews.pro/news/contracted-arr-considered-harmful.txt", "jsonld": "https://wpnews.pro/news/contracted-arr-considered-harmful.jsonld"}}