G2's "Evaluation Maze": Evaluation Is Now the Longest Buying Stage — and the CFO Is Vetoing Approved Deals

G2's "Evaluation Maze": Evaluation Is Now the Longest Buying Stage — and the CFO Is Vetoing Approved Deals

G2's 2026 Buyer Behavior Report: "The Evaluation Maze" (published July 22, 2026; online survey of 1,038 B2B software decision-makers fielded June 2026, plus 55 qualitative interviews) is the companion to its Answer Economy research (G2's Answer Economy: AI Chatbots Are Now the #1 Shortlist Gate, and Review Citations Are the Trust Signal) — and it documents the other half of the AI-shifted journey: what happens after the shortlist. AI made discovery fast; it made winning the deal harder.

The journey's center of gravity moved to evaluation. Evaluation is now the longest stage of the buying journey at 40%, surpassing research (down to 36%). After a vendor is selected, IT security review is the single biggest source of delay (39% of buyers, 50% among enterprises), followed by budget approval (32%) and implementation planning (25%). Meanwhile, internal resistance to AI adoption doubled from 16% to 29% in one year — the toughest resistance to AI purchases now comes from inside the buyer's own walls.

"AI has taken most of the friction out of finding software, but it raised new questions about cost, security, and internal trust… Winning the deal today means going beyond discoverability and actually helping buyers work through that scrutiny and defend the decision internally." — Tim Sanders, Chief Innovation Officer, G2 (PR Newswire, July 22, 2026)

Finance gatekeeping is the sharpest new number: finance involvement in software decisions jumped 31% → 46% in a year, and 49% of buyers say a CFO vetoed an already-approved software purchase in the last 12 months¹ — rising to 54% among organizations with dedicated token/LLM budgets. This substantially hardens the CFO Vetoes and Contract Term Compression: The Rise of Finance Scrutiny in AI Software Sourcing pattern: the veto isn't a corner case, it's near-universal experience.

Shortlisting dynamics (an update to the Answer Economy picture): 82% of buyers sourced software recommendations from an AI chatbot in the last 24 months, and 50% say chatbots had their greatest impact when narrowing and comparing options. But review sites (38%) have now overtaken AI chatbots (37%) as the top source influencing which vendors make the shortlist — the first time that's flipped. Machine visibility gets you considered; review citations get you shortlisted.

Agents are in the buying process — on a leash. 61% of buyers use or plan to use AI agents in software buying, with top use cases: evaluating total cost of ownership (51%), building shortlists (51%), researching solutions (49%), and evaluating shortlisted vendors (46%). But only 47% would let an agent research and recommend while humans keep final authority; just 9% would let an agent execute purchases within guardrails, and 2% without pre-approval.

SaaS economics: 70% say the pace of innovation is pushing them toward shorter contracts; 80% of organizations now give developers/technical teams a dedicated token or LLM usage budget; preference for outcome-based pricing more than doubled (11% → 23%); 49% have already been offered a variable-cost option instead of a seat license.

What a founder should do with this: the deal is won or lost in the evaluation stage — so arm your champion to survive security review (enterprise especially), anticipate the CFO veto with defensible TCO and kill-switch/flex terms, keep review-platform citations current because they now out-rank chatbots at the shortlist moment, and expect the buyer's agents to scrutinize your TCO math before any human reads your deck. For the production-side trust economics of the agents buyers are purchasing, see G2's 2026 AI Agents Report: The "Distrust Tax" — Oversight, Not Capability, Is Now the Adoption Bottleneck; for journey-stage strategy, Enterprise Software Buying Journey: Where AI-Enabled Founders Must Win Each Stage.