DroneBlocks

Key Findings — Fathom Notes

What 17 weekly recaps (Mar 17 – Jun 16) reveal about how the business actually runs. Each finding links to the pillar/principle that addresses it.

prepared 2026-06-22for Henrysource 17 recaps

// A — systemic risks

Fix these or they recur

1

Eric is a single point of failure.Named owner of the large majority of action items across all 17 meetings — procurement, inventory, shipping, finance, vendor comms. If Eric is out, shipping stops.

2

A second bus factor: Dennis + Aziz on the tech side.Dennis is sole owner of product/firmware/sim; Aziz (a contractor, with college finals) is the only hardware builder. Both are single points.

3

China supply chain is the recurring crisis engine.In one quarter: counterfeit SD cards, paint-contaminated light-show drones + a $6k order swap, no shipping lane to Australia, 2-month customs delays, and the CM5 shortage threatening 16 orders incl. a 10-unit June 28 deadline. This drove the US-manufacturing pivot.

4

Work floats across meetings because nothing holds it.The ribbon cable, counterfeit SD cards, light-show customs, CM5, and the Teachable cancellation each reappear for weeks. Clearest case for a tracked board + a human triage gate.

5

Fulfill-on-demand with no inventory visibility.Stock is known by bin-check/memory; the $6k Noreen/Cedar Valley swap is the cost of no order-tracking.

// B — money & data integrity

Margin and trust in the numbers

6

Hardware margin is thin; the strategy is to pivot to software/services.Flagged case: a $51k FPV sale yielded only $6–7k margin. Repeated intent to "land and expand" into high-margin software, custom sims, and pro services (Amazon, BNSF).

7

CRM/finance data is unreliable.HubSpot mid-overhaul; $3.5M in stale deals closed out; deal bloat from QBO estimates; estimate source-of-truth ambiguous (QBO vs HubSpot).

8

Commission calculation is broken.QBO credits the estimate creator, not the account manager who closes — so Lindsey's commissions are hand-calculated.

9

Demo/loaner hardware has no accountability.The Kerrigan unit had to be chased down for a Pitsco deadline; a formal CC-hold program was proposed.

// C — growth & funnel leaks

Where demand leaks out

10

Lead capture is leaking at every seam.Conference QR codes drove traffic but 0 signups (broken landing-page CTA); "no system for capturing attendee contact info"; PPC ineffective at $76–465 cost-per-lead.

11

Email deliverability is at risk.unmannedairlines.com sender domain flagged by spam filters; DMARC/SPF/blacklist issues hurting cold outreach.

12

Trial funnel migration is incomplete.24.6% trial conversion is strong, but 11 enrollments still happened in Teachable — the funnel isn't fully on the new platform, and trial→paid follow-up isn't built.

13

Repair work is a liability.A Tello repair attempt = 1 unit smoked, 3 ripped traces → repairs discontinued.

// D — what's working

Protect these

14

University referral network is the bright spot.The Juan Calderon → Gustavo → Diego chain drives year-round revenue that offsets the K-12 summer slowdown. Structuring deals under the $5k approval threshold is smart.

15

Small targeted events beat big conferences.A ~$3k Texas STEM event paid for itself with 2–3 kit sales and out-engaged ISTE (225 flyers vs 25–50).

16

Part 107 prep is an emergent strategic wedge.Repeatedly the standout value-add; an FAA-listing free tier could be a real lead engine.

17

Financially stable and capitalized for the right move.~$200k cash + ~$160k Plains Capital earmarked for inventory/growth. The money to fund the forecast-inventory shift already exists.

// E — the through-line

What it all adds up to

The exercise was meant to surface who owns what and how it's surfaced. The notes say the deeper problem is single points of failure (Eric, Dennis/Aziz) running an on-demand business with no visibility layer — while a healthy software/services pivot and a stable balance sheet give room to fix it. The operating principles aim squarely at that: make ownership explicit, make state visible, and remove the people-shaped bottlenecks.