The AI landscape doesn't move in one direction — it lurches. Some techniques leap from experiment to table stakes in a single quarter; others stall against regulatory walls, technical ceilings, or organisational inertia that no amount of hype can dislodge. Knowing which is which is the hard part. The State of Play cuts through the noise with a rigorously maintained index of AI techniques across every major business domain — classified by maturity, evidenced by real-world adoption, and updated daily so you always know where you stand relative to the field. Stop guessing. Start knowing.
A daily newsletter distilling the past two weeks of movement in a domain or two — delivered to your inbox while the index updates in the background.
Each dot marks the weighted maturity of practices within a domain — hover for a brief summary, click for more detail
AI across the revenue cycle from lead identification to closed deal. The most consistently mature domain: three-quarters of practices are good practice, including lead scoring, pipeline forecasting, and conversation intelligence. CRM copilots are mainstream. The few leading-edge practices involve autonomous prospecting and deal-coaching agents. Momentum is moderate — most gains are incremental rather than transformative.
The headline: The fully automated AI salesperson lost the argument this fortnight. Teams that keep a person reviewing each message before it sends are booking more than twice the revenue of teams that don't.
Almost every sales organization now runs AI somewhere — roughly 87% of them. Very few get money out of it. Only 7% of business-to-business teams hit 90% forecast accuracy, and 69% of sales operations leaders say forecasting is harder than it was three years ago, despite everything they've bought. The reason is boringly consistent and has not changed in four years: the customer data these tools run on is a mess. Duplicate records, contacts that go stale at 25% a year, and the fact that roughly four-fifths of what actually happens in a deal never gets typed into the CRM (customer relationship management system) at all. A small group of companies with clean data, a defined sales process, and managers who genuinely run a weekly review are getting 14-28% higher win rates and much faster new-hire ramp. Everyone else bought the same software and got alerts nobody reads. If you are in that second group, you are in the majority — but the gap compounds every quarter.
Salesforce's flagship sales AI stalled, and the market noticed. Agentforce, its autonomous agent product (software that acts on its own without being prompted), has reached 23,000 of 150,000 customers and drew an analyst downgrade. The reason given was not price or competition — customers said their own data "is not organized well enough for meaningful AI work." Before you buy an agent, ask your team what percentage of your account records they would stake a forecast on.
AI sales-rep pilots hit 41% of large B2B teams — and 40-60% were switched off within 90 days. Broken integrations and email deliverability failures were the cause. In one head-to-head test, the human-reviewed setup booked a third as many meetings as the fully automated one but produced 2.3 times the revenue. Buy the research automation; keep a human on the send button.
AI got software pricing right only 29% of the time in a 14,000-query study. When it was wrong, it underquoted 62% of the time — meaning the error direction costs you margin, not just credibility. Salesforce shipped an autonomous quoting agent in the same window. Quotes should stay a human-approved step; 89% of professionals surveyed already insist on that.
China fined Trip.com roughly $766M for algorithmic pricing abuse. Regulators explicitly named data and algorithm coordination between competitors — the same theory behind the US RealPage settlement and 89-plus state bills. If your pricing tool ingests any competitor data you did not obtain publicly, that is now a legal review, not a procurement one.
Real-time in-call coaching became a default feature, not an upgrade. Zoom shipped AI Sales Assist as standard in its revenue product, and Gartner puts the win-rate gain at 8-12% within three months. If you are paying a premium line item for this, your next renewal is a renegotiation.
Your data cleanup is now the gating item on every AI purchase in this function. Only 15% of enterprises say their data is ready for autonomous agents. Sequence it the other way round — deduplicate and enrich before you buy the agent — or you will pay twice. Ask your revenue operations lead for a duplicate rate and a completeness percentage this quarter.
Hallucination liability is arriving through the courts, not the regulator. ("Hallucination" is when an AI tool confidently makes things up.) More than 1,490 court decisions have now produced sanctions traceable to AI-fabricated content, and 47% of enterprise AI users report making a major decision on invented output. Name an accountable owner for every customer-facing AI output — quotes, proposals, contract drafts — before legal does it for you.
The signals your sales team buys are going dark. Between 60% and 73% of buyer research now happens inside ChatGPT, Perplexity and Gemini, which leave no trace on your website analytics. Third-party intent data is increasingly blind to most of the buying journey. Shift budget toward first-party signals — product usage, job changes, existing-customer behavior — which still convert at three to four times the rate of cold lists.
The bottleneck is a management routine, not a product. Two separate 60-day field tests found the 14% win-rate gain appeared only in teams whose managers actively used the AI scorecards. United Rentals rolled out to 6,000 reps and hit 87% tool adoption but only 43% coaching adoption. No vendor sells the missing half.
Training, not access, is the adoption gap. Fifty-seven percent of executives use AI for sales content against just 6% of the individual contributors who actually sell; three-quarters of those contributors name lack of training as the top blocker. Seat licenses are the cheap part.
Consolidating your stack creates new single points of failure. A supply-chain breach at one competitive-intelligence vendor exposed CRM data for around 195-200 enterprise customers through old access credentials. Separately, platform mergers have measurably degraded forecast accuracy and deal-qualification rates when merged systems orphan legacy data. Simplification is worth doing — with an exit plan written down.
Go deeper: the full Sales & Revenue briefing — the longer analytical write-up, plus every practice we track in this domain with its maturity rating, the tools to consider, and the evidence behind our assessment.