Humantic AI

Amarpreet Kalkat is the Founder and CEO of Humantic AI, an original thinker and prolific writer on the role of AI in sales who firmly believes in the principles of buyer-first selling. His Sold With AI Substack is read by more than 4,000 sales leaders every month, and he has appeared on dozens of podcasts and spoken at events like Dreamforce and Sandler Summit.

Why are win rates falling even as sales teams become more active, more funded, and more AI-enabled than at any point in history?

Across industries, win rates have dropped to 19%. When it comes to quota achievement, 78% of sellers missed quota last year, up from 69% the year before. The effort has never been higher. The results have never been worse.

More was supposed to be the answer. It turned out to be the problem.

We spent a decade pouring investment into seller productivity. Help reps do more, send more, reach more. The logic seemed sound. What it missed entirely was the buyer absorbing all of that volume on the other end, growing more fatigued with every passing quarter. LinkedIn's State of Sales research confirmed what most sellers already sense. Buyers are more frustrated with vendor outreach than at any point on record, not despite the rise of sales technology but because of it.

Early work in behavioral AI revealed a deeper limitation. Technology had become very good at processing data about companies but had paid almost no attention to truly understanding them. Not what the company is actually trying to solve this year. Not where the real budget pressure is coming from. Not how the people inside it think, what makes them trust someone, or how they evaluate risk before committing to a decision. That gap, between knowing the account and knowing the person, is where most deals are won or lost.

This is where a new category of thinking is emerging. Real buyer intelligence operates on two levels simultaneously. The first is account intelligence, understanding the company's actual priorities for the year, the specific challenges its leadership is trying to solve, and where your solution fits meaningfully within that context. The second is people intelligence, understanding the decision-makers themselves, their communication style, their risk tolerance, and whether they respond to data or narrative when making high-stakes decisions.

In practice, only two things ever really matter in a deal. What the company needs and what the buyer wants. Most sales technology has made some progress on the first. Far fewer have seriously attempted the second. Humantic AI is built on the belief that combining both into a single, coherent view of the buyer is where the real leverage lies. The one thing we drive is helping sellers become experts in buyer intelligence, both the buyer's business and the people behind it. In one enterprise pilot involving 20 sellers, adapting outreach across both dimensions generated $26 million in additional pipeline within six weeks. A similar approach across 12 sellers at another organization produced $60 million in 30 days. Across more than 50 such engagements, the pattern is consistent: a 109% increase in qualified pipeline, a 16.2% lift in closed revenue, and 36.5% faster deal velocity. Not from more activity. From more relevance.

But the promise of buyer intelligence comes with real tension, and it deserves to be acknowledged directly. The more precisely an organization understands its buyers, the greater the responsibility to use that insight carefully. Over-personalization can feel invasive. Misread signals can produce false confidence in a deal that is already dead. As AI becomes more embedded in sales workflows, the risk is not just inefficiency. It is trust erosion on both sides of the table. Sellers who rely too heavily on AI-generated insight without exercising judgment often come across as formulaic rather than genuine, which can be worse than having no insight at all.

The human element does not shrink in a well-designed AI model. It becomes more important. Knowing the company's key priorities for the year and knowing that a specific buyer needs third-party proof before committing are both useful inputs. Deciding what to do with that information, what to send, when to follow up, and how to frame your solution against the buyer's actual business challenges is still entirely a human call. AI sharpens judgment. It does not replace it. The organizations seeing the most compounding results consistently treat buyer intelligence as context for better human decisions, not a substitute for making them.

Where this is going matters as much as where it is now. Buying committees have grown from an average of five people a decade ago to more than 11 today, with some enterprise deals involving 17 or more stakeholders. The complexity of getting a "yes" has multiplied, while the tools most teams rely on were built for a simpler era. That makes understanding not just valuable but necessary. The leaders who define the next decade will not be the ones who deploy the most AI. They will be the ones who use it to understand buyers more precisely across every stakeholder at every stage, without losing the judgment to act on what they learn. For leaders, the shift is simple. Stop optimizing for activity and start optimizing for understanding.

The missing layer in sales AI has never been more data. It has been understanding. That distinction is what will separate teams that scale from those that stall.