{"id":14445,"date":"2026-07-01T10:00:43","date_gmt":"2026-07-01T04:30:43","guid":{"rendered":"https:\/\/gtm360.com\/blog\/?p=14445"},"modified":"2026-06-28T20:38:30","modified_gmt":"2026-06-28T15:08:30","slug":"technology-matters-until-cost-matters-more","status":"publish","type":"post","link":"https:\/\/gtm360.com\/blog\/2026\/07\/01\/technology-matters-until-cost-matters-more\/","title":{"rendered":"Technology Matters &#8211; Until Cost Matters More!"},"content":{"rendered":"<p><em>Anthropic<\/em> and other Artificial Intelligence labs have recently started charging customers for token usage in addition to per-seat subscription. This is like <em>Prime Video<\/em> charging for a subscription and asking extra money for some movies!<\/p>\n<p>This move has had a knock-on effects on Cloud Service Providers and SAAS vendors who consume tokens. <em>Microsoft<\/em>, <em>ServiceNow<\/em>, and <em>Workday<\/em> have responded by either tacking on usage costs to seat-based pricing or by switching totally to usage-based pricing.<\/p>\n<p>This has in turn escalated token costs for AI end customers. In this oft-repeated news story, Uber&#8217;s CTO is quoted saying that the rideshare giant blew through its entire 2026 token budget in the first four months itself.<\/p>\n<p><a href=\"https:\/\/gtm360.com\/blog\/wp-content\/uploads\/2026\/06\/ai-token-price-rising-fi.jpg\" target=\"_blank\" rel=\"noopener\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone wp-image-14448 size-full\" src=\"https:\/\/gtm360.com\/blog\/wp-content\/uploads\/2026\/06\/ai-token-price-rising-fi.jpg\" alt=\"\" width=\"630\" height=\"328\" srcset=\"https:\/\/gtm360.com\/blog\/wp-content\/uploads\/2026\/06\/ai-token-price-rising-fi.jpg 630w, https:\/\/gtm360.com\/blog\/wp-content\/uploads\/2026\/06\/ai-token-price-rising-fi-500x260.jpg 500w\" sizes=\"auto, (max-width: 630px) 100vw, 630px\" \/><\/a><\/p>\n<p>Naturally, customers are finding ways to save money. A recent issue of <em>Applied AI<\/em> newsletter via <em>The Information <\/em>reports the top five cost-saving measures used by end customers of AI:<\/p>\n<ol>\n<li>Using model routers to switch between models<\/li>\n<li>Prompt engineering<\/li>\n<li>Rationing employees&#8217; AI usage<\/li>\n<li>Pushing for discounts<\/li>\n<li>Stop using AI<\/li>\n<\/ol>\n<p>Applied AI cites <em>Novo Nordisk<\/em> as Exhibit A of the last tactic:<\/p>\n<blockquote class=\"skr-bq-noline\">\n<div style=\"background-color: #fafafa; padding: 15px 30px;\">In a dose of operational reality, some companies are saving money by actively restricting AI usage when traditional, structured software is cheaper.\u00a0For instance, when pharmaceutical giant <em>Novo Nordisk<\/em> recently analyzed past clinical trial data using Anthropic&#8217;s Claude, executives realized the standard version of Microsoft Excel &#8211; even without Microsoft&#8217;s Copilot AI features &#8211; was both cheaper and more reliable.<\/div>\n<\/blockquote>\n<p><!----------END----------><\/p>\n<p>Let me call this <strong>AI Repatriation<\/strong>, the process of ditching AI and instead using traditional SAAS \/ COTS software.<\/p>\n<p>This term is inspired by <a href=\"https:\/\/www.uniquesolutionsit.com\/solutions\/cloud-repatriation\/\" target=\"_blank\" rel=\"noopener\"><em>Cloud Repatriation<\/em><\/a> and <em><a href=\"https:\/\/gtm360.com\/blog\/2024\/10\/02\/will-chatgpt-kill-saas\/\" target=\"_blank\" rel=\"noopener\">SaaS Repatriation<\/a>.<\/em><\/p>\n<ul>\n<li><em>Cloud Repatriation<\/em> is ditching cloud and moving IT workloads to onprem \/ private cloud infrastructure like <em>HPE Greenlake<\/em>.<\/li>\n<li><em>SaaS Repatriation<\/em> is canceling SaaS subscriptions and building the equivalent applications inhouse using AI vibe coding tools like <em>Claude Code<\/em>, <em>Cursor<\/em>, and <em>Github Copilot<\/em>.<\/li>\n<\/ul>\n<hr style=\"width: 70%;\" \/>\n<p>Novo Nordisk&#8217;s AI Repatriation corroborates three predictions I&#8217;d made in my blog post entitled <a href=\"https:\/\/gtm360.com\/blog\/2023\/11\/08\/technology-matters-just-not-at-the-top-of-funnel\/\" target=\"_blank\" rel=\"noopener\"><strong>Technology Matters \u2013 Just Not At The Top Of Funnel<\/strong><\/a>.<br \/>\n<!--BLOCKQUOTE, NO LHS LINE, GREEN--><\/p>\n<p>1. Novo Nordisk decided that AI was the right technique to analyze clinical trial data.<\/p>\n<blockquote class=\"skr-bq-noline\">\n<div style=\"background-color: #00c89640; padding: 15px 30px;\">In today\u2019s world, if one vendor goes to market with Excel-based Predictive Analytics and another vendor goes to market with AI-based Predictive Analytics, the AI vendor will appeal more to the prospect.<\/div>\n<\/blockquote>\n<p><!----------END----------><\/p>\n<p>2. After token costs skyrocketed, the pharma major went back to the drawing board, and realized that there&#8217;s more than one way to skin the clinical trial analysis cat.<br \/>\n<!--BLOCKQUOTE, NO LHS LINE, GREEN--><\/p>\n<blockquote class=\"skr-bq-noline\">\n<div style=\"background-color: #00c89640; padding: 15px 30px;\">Excel can do the job.<\/div>\n<\/blockquote>\n<p>3. The company found that it could save truckloads of money by doing the said analysis on Excel. As its Chief Transformation Officer, Stephanie Bova, said, &#8220;If I can do it better and cheaper and more reliably in Excel, I&#8217;m going to tell you to stay in Excel&#8221;.<!--BLOCKQUOTE, NO LHS LINE, GREEN--><\/p>\n<blockquote class=\"skr-bq-noline\">\n<div style=\"background-color: #00c89640; padding: 15px 30px;\">Despite believing in the superiority of AI for analytics, many prospects might still settle for an Excel-based model, so I\u2019d caution vendors to avoid going head-on against Excel!<\/div>\n<\/blockquote>\n<p>On a side note, while I couldn&#8217;t find any mention of this in the aforementioned newsletter, I&#8217;m sure the Excel option involves more manual work than Claude towards cleansing the clinical trial data, setting up pivot tables, selecting the right statistical model, and so forth. Of course, it&#8217;s very much possible that, even after loading the extra labor costs for those activities, Excel came out cheaper than the AI option.<\/p>\n<hr style=\"width: 70%;\" \/>\n<p>Before token costs went up, AI was <em>faster, better and cheaper<\/em> than Excel, but after token costs went up, Excel became <em>better, cheaper and more reliable<\/em> than AI.<\/p>\n<p>Technology mattered &#8211; until cost mattered more!<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Anthropic and other Artificial Intelligence labs have recently started charging customers for token usage in addition to per-seat subscription. This is like Prime Video charging for a subscription and asking extra money for some movies! This move has had a knock-on effects on Cloud Service Providers and SAAS vendors who consume tokens. Microsoft, ServiceNow, and &#8230; <a title=\"Technology Matters &#8211; Until Cost Matters More!\" class=\"read-more\" href=\"https:\/\/gtm360.com\/blog\/2026\/07\/01\/technology-matters-until-cost-matters-more\/\" aria-label=\"Read more about Technology Matters &#8211; Until Cost Matters More!\">Read more<\/a><\/p>\n","protected":false},"author":4,"featured_media":14448,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[17,8,1],"tags":[108],"class_list":["post-14445","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-b1-integrated-marketing","category-it-marketing","category-mandatory-category","tag-ai-repatriation"],"_links":{"self":[{"href":"https:\/\/gtm360.com\/blog\/wp-json\/wp\/v2\/posts\/14445","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/gtm360.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/gtm360.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/gtm360.com\/blog\/wp-json\/wp\/v2\/users\/4"}],"replies":[{"embeddable":true,"href":"https:\/\/gtm360.com\/blog\/wp-json\/wp\/v2\/comments?post=14445"}],"version-history":[{"count":7,"href":"https:\/\/gtm360.com\/blog\/wp-json\/wp\/v2\/posts\/14445\/revisions"}],"predecessor-version":[{"id":14487,"href":"https:\/\/gtm360.com\/blog\/wp-json\/wp\/v2\/posts\/14445\/revisions\/14487"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/gtm360.com\/blog\/wp-json\/wp\/v2\/media\/14448"}],"wp:attachment":[{"href":"https:\/\/gtm360.com\/blog\/wp-json\/wp\/v2\/media?parent=14445"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/gtm360.com\/blog\/wp-json\/wp\/v2\/categories?post=14445"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/gtm360.com\/blog\/wp-json\/wp\/v2\/tags?post=14445"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}