Forecasts should weight implementation

Jason Lemkin’s July 9 SaaStr post argued that AI sales teams should judge pipeline by execution signals rather than enthusiasm alone, as enterprise buying cycles become harder to read. The post did not publish a formal forecasting framework, but it described a higher operating standard for AI selling and deployment execution at a time when more buyer research is happening before a prospect speaks to sales. Semrush said in a study published July 8 that 622 valid responses from U.S. B2B professionals showed AI tools now influence vendor discovery, evaluation, shortlist formation and purchase decisions. Deloitte said in its 2026 enterprise AI report that success depends on moving “from ambition to activation,” underscoring a broader market shift toward implementation evidence. ### Why are AI forecasts getting less reliable in the first place? Semrush said buyers are using AI tools earlier in the B2B process, which means more vendor evaluation can happen outside the channels revenue teams have historically measured. The company said respondents use AI to scope categories, compare vendors and build shortlists before direct contact with sales, reducing the visibility of conventional lead-source and stage-based models. (saastr.com) For revenue teams, that creates a forecasting problem. If the visible signal is an executive meeting or a late-stage call, but the real conviction formed earlier through technical research and AI-assisted comparison, pipeline can look stronger than it is. That inference is supported by Semrush’s survey findings and by SaaStr’s emphasis on execution over headline enthusiasm. (semrush.com) ### What counts as implementation evidence in an AI deal? A named technical owner is one of the clearest markers because it shows the buyer has assigned responsibility inside the account. A defined first workload matters for the same reason: it narrows the project from general AI interest to a deployable use case with boundaries, dependencies and a likely start point. Those signals were highlighted in the source briefing drawn from SaaStr’s July 9 post and related enterprise AI reporting. (saastr.com) Security and governance work is another threshold. Deloitte’s recent enterprise AI materials said organizations are advancing AI transformation through governance, new operating models and proven return on investment, while its broader AI reporting has stressed activation over ambition. When a buyer has begun a security review, governance review or rollout planning, the deal has moved from aspiration toward implementation. (saastr.com) ### Why isn’t executive enthusiasm enough anymore? Jason Lemkin wrote in related SaaStr posts that AI systems are raising expectations for speed and output across go-to-market teams, but he also described deployment and management work that continues after the initial decision. In a separate post published this week, SaaStr said it now operates 20-plus AI agents and that managing them consumes a material share of leadership time, illustrating that adoption does not end at purchase. (deloitte.com) That matters in forecasting because senior buyers can be genuinely interested while the organization remains unprepared. Deloitte said organizations stand at the “untapped edge” of AI’s potential and tied progress to activation, not stated ambition. A forecast that treats executive sponsorship as equivalent to deployment readiness risks overstating close probability. (saastr.com) ### Which proof points should move forecast confidence up? A first deployment plan should carry more weight than a polished executive call. In practice, that includes a technical owner, a specific initial workload, security or governance review underway, success criteria for rollout, and an enablement or training plan attached to adoption. Those elements indicate the buyer has started solving for implementation, not just approval. (deloitte.com) Deloitte’s June analysis said deploying a copilot is the easy part, while full transformation requires work redesign, governance and measurement. That framing aligns with a stricter forecast model: confidence should rise when the account shows operational readiness, cross-functional ownership and a path to production. (saastr.com) ### What should teams watch next? Semrush said its survey fieldwork ran in March and April 2026, and Deloitte’s 2026 enterprise AI report tracks adoption and impact through 2025, giving revenue teams two current benchmarks for how buyers are researching and operationalizing AI. The next practical checkpoint is whether open opportunities show named owners, defined workloads and active governance review before quarter-end forecast calls. (semrush.com) (deloitte.com)