Potrebujem več zaključenih poslov.
En korak, ki ta teden šteje.

Next-Best-Action — demo run

Input: sample-deal-notes.md — Meridian Health Analytics, €62k ACV, Proposal stage, 3 weeks.

The one action

Get a direct conversation with Marcus Reyes (CFO, economic buyer) scheduled before the September 5th board review — ask Dana to broker the intro this week, framed as "I want to answer his pricing question myself so he has accurate numbers for the board," not as going around her.

Which gate fired

Tier A, gate 2 — no economic buyer access at a late stage. The deal has been in Proposal for three weeks. The only person who has ever been on a call is Dana (champion). Marcus (the person who actually owns the budget line and will decide) has never spoken to the rep directly — every data point attributed to him ("thinks it's a bit expensive," "supportive in principle") has arrived filtered through Dana. That satisfies the gate exactly: late stage, single-threaded, no direct access to the buyer.

What else looked tempting, and why it lost

• Address the pricing objection (Tier B, content move). This is the obvious pull — a specific objection is sitting right there in the notes. But it already failed once: the rep sent a comparison sheet on 2026-08-12 specifically to answer this objection, relayed through Dana, and got no reply for two weeks. That's not evidence the objection is unanswerable — it's evidence the channel is broken. Sending a second comparison sheet through the same proxy is repeating the exact move that already stalled. Fix access first; the objection may resolve itself in five minutes once Marcus is actually on the line, or it may turn out to be a different objection than Dana thinks it is.
• Build urgency against Verrant before the board date (Tier B). Also genuinely time-pressured — the board review is 10 days out — but the same logic applies: urgency has to land with the person feeling the deadline. Right now the rep has no way to make that case to Marcus directly, only to Dana, who is already motivated and doesn't need convincing. Creating urgency for the wrong audience isn't zero-value, but it isn't the bottleneck.
• Ask for the close. Not remotely appropriate yet — there's an unresolved, unverified objection from the actual decision-maker and no direct line to him. Closing language this week would read as tone-deaf.

The reasoning that ties it together: every plausible content move this week depends on reaching Marcus, and none of them currently can. Fixing that is the one move that makes every other move possible — which is exactly why it outranks them rather than sitting alongside them on a list.

Honesty check

• The notes never state Dana and Marcus's actual reporting relationship or how much sway she has over his final call — that would sharpen whether "ask Dana to broker" is even the right route in, versus finding a second path to Marcus independently.
• No stated cadence for this account before Discovery, so "how slow is too slow" for Dana's replies was judged against this deal's own history only (she has never gone more than a week without responding) — she hasn't fired the silence gate, but that's a judgment call, not a hard number from the data.
• Deal value (€62k) wasn't a factor in which gate fired — this rubric narrows on structure, not stakes. A cross-deal view of whether this deal is worth this much attention relative to the rest of a pipeline is a different question, answered by DEVENTA / CLOSE — Deal Intelligence, not by this product.

Deal Autopsy — demo run

Input: sample-lost-deal.md — Solvane Systems, €45k ACV, Closed Lost, 19 weeks in pipeline.

1. Deal snapshot

Solvane Systems, industrial IoT platform, ~280 employees. €45,000 ACV endpoint security add-on. Reached Negotiation-adjacent territory (proposal under internal review) before dying. Logged loss reason: "Chose competitor." Time in pipeline: April 2 – August 14 (19 weeks).

2. Timeline reconstruction

3. True point of death

Not the logged reason. "Chose competitor" implies a straightforward loss on merits. The actual point of death is July 3 — the proposal that answered the migration question but never answered the specific switching-cost / remaining-contract-term question Okonkwo, the actual economic buyer, had asked eight days earlier. That question was deferred at the June 26 call and then simply never came back. Priya's departure in late July removed the one person still relationship-invested in pushing the deal forward, but even had she stayed, the buyer's actual stated objection was still sitting unanswered. The champion loss removed the chance of recovery; the unanswered proposal is what created the thing that needed recovering from.

4. Signal audit

5. Root cause classification

Primary: unaddressed displacement cost. The single clearest thread running through this deal, raised twice — once by the champion in April, once directly by the economic buyer in June — is that this was a displacement sale (unseat a budgeted incumbent) that was never actually sold as one. The proposal that should have made the switching-cost case explicit instead made a generic capability case. Secondary, contributing: single-threaded champion loss — the rep never built an independent relationship with Okonkwo; the only non-Priya-mediated contact happened after she'd already left the account. That compounded the primary problem rather than causing it: even a well-answered displacement case would have been harder to land with a freshly-arrived, unfamiliar buyer and no internal advocate.

6. Generalizable vs. one-off verdict

• Generalizable — treat this as a real pattern to watch for: deferring a displacement/switching-cost question to "we'll cover it in the proposal" and then not covering it. This wasn't bad luck; it's the same signal surfacing twice (April, June) and being handled the same deferred way both times. That's a seller-side, controllable behavior, and it would create the same risk in any deal that requires unseating an incumbent with committed budget — not just this one.
• Generalizable, secondary — also worth watching: the economic buyer had no relationship with the seller independent of the champion, so the champion's departure was structurally fatal rather than just inconvenient. This is the same access risk Next-Best-Action's Tier A2 gate exists to catch while a deal is still live — this autopsy is what it looks like when that gate never got triggered in time.
• One-off — do not generalize: the specific fact that Priya was personally promoted to a different internal team in July. That is her career move, not a repeatable seller-caused event, and there's no clear lever in this deal that would have prevented it. What generalizes is not "champions get promoted" — it's "this deal had no second relationship to fall back on when she did."

7. Process change

Recommended, on the strength of the generalizable finding: when a deal requires displacing an incumbent with already-committed budget, the proposal itself must explicitly price and answer the switching-cost / remaining-term question the moment any stakeholder raises it — not defer it forward. Pair this with building a direct line to the economic buyer as soon as they're introduced, rather than letting the champion remain the only channel through Negotiation.

Honesty check

• Confidence in the root-cause call: moderate-high. The displacement question was raised explicitly twice, by two different people, and the proposal is directly verifiable as not having addressed it — that's stronger evidence than most single-deal autopsies get.
• What this autopsy cannot tell you: whether "defers displacement objections to the proposal" is something this rep does on every deal like this, or whether it happened once here for reasons specific to this deal's timing (compliance audit workload, the CISO intro landing awkwardly). One loss is one data point. Confirming this as a real, costly pattern — versus an unlucky one-off — needs either a second Deal Autopsy on a similar lost deal showing the same deferral, or a DEVENTA / CLOSE — Deal Intelligence pass across the live pipeline checking whether other displacement-type deals currently show the same unanswered-switching-cost gap.