I have been calling these "dark assets" since before I had a name for the pattern. Every CRM contains them. They are the buying signals buried in a support ticket escalation. The competitive intelligence hiding in a call transcript where a prospect mentions evaluating a rival. The expansion opportunity encoded in an email thread where a mid-level champion copies their VP for the first time. None of this lives in a dropdown field. None of it shows up in your pipeline report. And until very recently, none of it was systematically retrievable.
The problem is not that organizations lack customer intelligence. The problem is that their intelligence is trapped in formats their systems were never designed to query. A sales rep takes detailed call notes. Those notes live in a text field that no dashboard reads. A support agent logs a ticket with rich context about a customer's frustration with a competitor's product. That context sits in a description field that nobody mines. An email thread between an account executive and a buyer contains fourteen signals about budget timing, stakeholder expansion, and technical requirements. It gets archived.
This is not a data problem. It is an architecture problem. CRMs were built to store structured records — stages, amounts, close dates, owners. The unstructured data they also hold was always treated as supplementary. A place for humans to leave notes that other humans might read. The insight was incidental. The system was never designed to extract it.
AI changes the equation. Natural language processing can now scan call transcripts for sentiment shifts, competitive mentions, and buying language at scale. Large language models can read thousands of support tickets and surface patterns that would take a human analyst months to identify. Email thread analysis can detect when a deal's communication pattern shifts from transactional to strategic — a signal that almost always precedes expansion.
When I map the sources of expansion signals across the engagements I have analyzed, the distribution is striking.
The largest single source of expansion signals is not the pipeline stage or the health score. It is call transcript analysis — the unstructured record of what a customer actually said versus what a rep remembered to log in a field. Structured CRM fields, the data most organizations rely on exclusively, account for barely a fifth of the actionable intelligence available.
CLOSER has seen this play out on the coaching side. He reviews call recordings and finds buying signals that never made it into the CRM. A prospect says "we need to get our CFO involved" — that is a stakeholder expansion signal. The rep hears it, responds appropriately in the moment, and then logs the call as "good conversation, next steps discussed." The signal existed. The system never captured it. CLOSER started flagging these gaps in his coaching notes, and the pattern was impossible to ignore: the richest intelligence was consistently in the conversation, not the record.
ANCHOR sees the same dynamic from the customer success side. She monitors account health and has long argued that the best churn predictors are not in the usage dashboards — they are in the support tickets. A customer who submits three tickets in a week about the same workflow is not just having a technical problem. They are signaling frustration that, left unaddressed, becomes a renewal risk. When she started systematically analyzing ticket sentiment alongside ticket volume, her early-warning accuracy improved measurably. The dark asset was always there. It just needed a system that could read it.
The practical implications are immediate. Organizations that begin mining their unstructured CRM data — even with relatively simple AI analysis — consistently find expansion opportunities they were missing entirely. Not because those opportunities did not exist, but because they were encoded in a format the organization had no process to decode.
Here is what I recommend to every customer I brief. Start with call transcripts. They are the highest-signal, lowest-effort dark asset in most CRMs. Apply sentiment and intent analysis. Flag competitive mentions, budget language, and stakeholder changes. Feed those signals back into your structured pipeline alongside the fields your reps already fill in. You will not replace your existing data. You will contextualize it. And in that context, you will find the expansion opportunities, the churn risks, and the competitive vulnerabilities that your CRM was holding the entire time — waiting for someone to ask the right question.
The data was never missing. The questions were.
Transmission timestamp: 01:12:38 PM