商业支出报告：2026年春季AI与软件支出研究亮点

Introduction3Introduction3 A note from our economist4A note from our economist4 About this reportAbout this report 5 Key findingsKey findings 66 Ramp AI IndexRamp Al Index 7 Ramp RateRamp Rate 1212 Charts from the cutting room floor21Charts from the cutting room floor21 IntroductionIntroduction Everyone's a pundit. That would be fine, except that everyone iswrong, especially on AI and the economy.wrong,especiallyon Al andthe economy. A note from our economistA note from oureconomist I've found myself getting annoyed hearing everyone's definitiveproclamations on this still-nascent market. Why does it feel likepeople are so confident about an industry with no precedent,little available data, and inherently limited history from whichanyone can claim expertise?I've foundmyself getting annoyed hearing everyone's definitiveproclamations on this still-nascent market. Why does it feel likepeople are so confident about an industry with no precedent,little available data, and inherently limited history from whichanyone can claim expertise? Consider these three things I heard about a year ago — manyconfident proclamations from folks in the industry and none ofwhich have turned out to be true.Consider these three things I heard about a year ago - manyconfident proclamations from folks in the industry and none ofwhich have turned out to be true. 1AI will be commoditized. The price would come down,models would converge, and spending would level off as theprice of tokens went to zero. Instead, average monthly AItoken spend on Ramp has grown 13x since January 2025,dwarfing every other business spend category we track.Al will be commoditized. The price would come down,models would converge, and spending would level off as theprice oftokens wentto zero.Instead,average monthly Aitoken spendon Ramphasgrown13xsince January2025dwarfing every other business spend category we track. 2AI will create a hiring boom for software engineers.The Bureau of Labor Statistics still projects a nearly 20%increase in hiring for software engineers over the nextdecade. Instead, employment at software companies fell in2024. The latest data from private sector datasets showsa similar slowdown in tech hiring.2Al will create a hiring boom for software engineers.The Bureau of Labor Statistics still projects a nearly 20%increase in hiring for software engineers over the nextdecade. Instead, employment at software companies fellin2024. The latest data from private sector datasets showsa similar slowdown in tech hiring. 3Only model companies will capture value.Remember the"AI wrapper" discourse? Cursor went from $1 million to $2billion in annual revenue in 26 months, built entirely onfoundational model APIs. It's one of the fastest-growingcompanies, according to spend data from Ramp.Onlymodelcompanieswillcapturevalue.Rememberthe"Al wrapper" discourse? Cursor went fromS1million to $2billion in annual revenue in 26 months, built entirely onfoundational model APls. It's one of the fastest-growingcompanies, according to spend data from Ramp. This is an extraordinarily uncertain market, yet it attracts so manydefinitive claims from actors without access to data on actualspending behavior. The so-called experts often have a profitmotive. Those who don't tend to hedge so much they saynothing useful at all.This is an extraordinarily uncertain market, yet it attracts so manydefinitiveclaimsfromactorswithoutaccesstodataonactualspending behavior.The so-called experts often have a profitmotive. Those who don't tend to hedge so much they saynothing useful at all. The purpose of this report is to give you more certainty. We trackover$100 billion in annual business spend across 50,000companies. That's the dataset behind what follows. We'reconfident about what we know well, and we make nomisrepresentations.Thepurpose of this report is to give youmore certainty.We trackover 100billion in annual business spend across 50,000companies.That's the dataset behind whatfollows.We'reconfident about what we know well, and we make nomisrepresentations. Sincerely,Sincerely, The Ramp Business Spending Report is a quarterly analysis ofcorporate spending trends based on billions of aggregated,anonymized transactions from over 50,000 businesses usingRamp Bill Pay and corporate cards.TheRamp Business Spending Report is a quarterly analysis ofcorporatespending trends basedonbillions of aggregated,anonymized transactions fromover50,o00businesses usingRampBillPayandcorporatecards. About this reportAbout this report Our transaction set is built using company models that extractline-item text from paid receipts and bills uploaded by thepurchaser following a sale. We use internal and external data,alongside proprietary Ramp models, to categorize businessesin size, segment, and sector. Small businesses representcompanies with one to 24 employees. Medium-sizedbusinesses represent companies with 25 to 99 employees.Large businesses include companies with 100 or moreemployee