{"id":587,"date":"2025-11-30T13:59:40","date_gmt":"2025-11-30T13:59:40","guid":{"rendered":"http:\/\/www.andersson-recycling.com\/index.php\/2025\/11\/30\/durability-not-growth-will-define-ai-winners\/"},"modified":"2025-11-30T13:59:40","modified_gmt":"2025-11-30T13:59:40","slug":"durability-not-growth-will-define-ai-winners","status":"publish","type":"post","link":"http:\/\/www.andersson-recycling.com\/index.php\/2025\/11\/30\/durability-not-growth-will-define-ai-winners\/","title":{"rendered":"Durability (Not Growth) Will Define AI Winners"},"content":{"rendered":"
Happy Thanksgiving to all who are celebrating! We\u2019re grateful for you all being part of GTMfund through our media platform, GTMnow
<\/p>\n
There are many independent perspectives shared through content. We read a ton of them, and there\u2019s connective tissue between some that reveal a broader macro and message. This week, we\u2019re highlighting a few pieces from different leaders because, together, they reveal a deeper pattern emerging across AI.<\/p>\n
Across the ecosystem, a single truth is coming into focus:<\/p>\n
Most AI traction today is experimental, not durable. Three independent perspectives were written, which are highlighted here: <\/p>\n Jamin Ball\u2019s ERR vs. ARR analysis<\/p>\n<\/li>\n Des Traynor\u2019s Four Horsemen<\/p>\n<\/li>\n Cassie Young\u2019s customer success renaissance <\/p>\n<\/li>\n<\/ul>\n But together, they describe the same structural shift:<\/p>\n We\u2019re entering an AI cycle defined not by adoption, but by retention. <\/a><\/figure>\n<\/p><\/div>\n <\/a><\/figure>\n<\/p><\/div>\n We just wrapped up our last event of the year, so are looking at events for 2026 now.<\/p>\n Whether it\u2019s an offsite, conference, or any other kind of event, BoomPop<\/a> makes event planning happen with end-to-end planning all in one place.<\/p>\n They handle everything from venues to experiences, so you can focus on the delights of meeting in person, not the logistics.<\/p>\n Being part of GTMnow, you\u2019re eligible for full-service event planning for just $99 per person (terms apply). Head to boompop.com\/gtmfund<\/a> to explore seamless support for your events.<\/p>\n In his March 2024 Clouded Judgement<\/em>, Jamin Ball surfaces one of the most important (and most uncomfortable) truths about AI revenue today: a significant portion of what startups call ARR is really ERR: Experimental Run-Rate Revenue. Many have since added to the narrative, including ourselves in this edition<\/a>.<\/p>\n Classic SaaS ARR works because retention is predictable:<\/p>\n 90\u201395% gross retention keeps customers for a decade.<\/p>\n<\/li>\n 120%+ NRR compounds revenue.<\/p>\n<\/li>\n Investors can underwrite future cash flows with confidence.<\/p>\n<\/li>\n<\/ul>\n In AI, three structural breaks are happening at once: <\/p>\n 1. Annualizing without commitment. 2. Buyers are experimenting aggressively. 3. Early AI categories are commoditizing. Often, AI companies are reporting growth that is neither predictable nor durable. If ERR explains why early traction is unstable, the Four Horsemen<\/strong> explain how to assess what\u2019s actually durable. In a Cheeky Pint podcast episode, Intercom\u2019s co-founder expressed that today\u2019s AI boom has created an illusion of traction. ARR is rising, logos look impressive, and rounds are getting bigger, but a large share of this revenue is experimental, not durable. Much of what\u2019s being counted as ARR is actually ERR (experimental run-rate): pilots, POCs, and opt-out trials funded by innovation budgets and AI FOMO.<\/p>\n He introduces the Four Horsemen of AI – the filters that reveal whether traction is real:<\/p>\n Revenue backed by usage<\/p>\n<\/li>\n Usage tied to business impact<\/p>\n<\/li>\n Deep, differentiated AI<\/p>\n<\/li>\n Positive unit margins<\/p>\n<\/li>\n<\/ol>\n The idea is that startups should have all four of these. But most AI companies meet only the illusion of the first horseman.<\/p>\n
<\/strong>And the market is beginning to price that reality in.<\/p>\n\n
Not by growth, but by durability.<\/p>\n
<\/picture>\n
<\/picture>\n
<\/div>\nJamin Ball (Clouded Judgement): ERR vs. ARR<\/a><\/strong><\/h2>\n
\n
<\/strong>Teams are multiplying monthly usage by 12 and calling it ARR, even when the revenue doesn\u2019t behave like a contract.<\/p>\n
<\/strong>They\u2019re testing multiple vendors in parallel. In many cases, the first tool used in a prototype is not the one running in production.<\/p>\n
<\/strong>Competition is exploding, prices are falling, and switching happens before companies can even measure churn.<\/p>\n
Gross retention will be the truth serum of this cycle.<\/p>\nDes Traynor (Intercom): The Four Horsemen of AI<\/a><\/strong><\/h2>\n
\n