8 to 10fSigned letters of intentTier A, from the $500k campaign in one quarter
6eCustomers installed by month 2427 boxes, in 3 states
$7.13MePeak needCumulative, in month 18, with no implementation fee
01
The ask
Two closes, with a gate between them.
About $4.0M carries months 1 to 10: certification, the benchmark, the 90-day letter-of-intent campaign and the delivery platform. About $3.5M, priced once the gate is passed, carries months 11 to 24 and installs the first customers, from month 13. Before then, shadow-mode trials run on buyers' own past cases. Peak cumulative need is $7.13Me in month 18, with no implementation fee, so the figures are conservative.
People and operations 31%Go-to-market and sales 28%Product and certification 15%Delivery and working capital 14%Contingency and reserve 12%
31%People and operations
28%Go-to-market and sales
15%Product and certification
14%Delivery and working capital
12%Contingency and reserve
02
Three levels
Each level buys something the one below cannot.
The model was run at $2M, $6M and $20M. None of the three nominal amounts funds its own plan, so each is right-sized: $2.5M for proof, $7.5M to month 24, $22M for 36 months. The middle one is the plan. Each card's cushion is the right-sized raise less the deepest point of cumulative cash within that level's own horizon.
Implementation fee per customer
Showing an implementation fee of $0 per customer
$A · Proof
$2.5M
Right-sized from a nominal $2M
The box works, is certified, and one customer runs it.
Certification, a published benchmark, one reference site live by about month 12, and the pre-raise test data. A bridge to the next round, with no campaign money.
$B · The plan
The plan
$7.5M
Right-sized from a nominal $6M
Buyers sign, and we can deliver to several at once.
Signed LOIs
8 to 10 Tier A and about 25 Tier B from the $500k sprintf
All four Answer Sheet items and national reach. The only level at which Allodial becomes a grid-flexibility and mesh company rather than an appliance vendor.
Need is the lowest point of cumulative cash. Runway counts the months before cumulative need passes the right-sized raise. A negative cushion is shown in parentheses. The model collects the implementation fee 2 months after each install; the $90k case is computed the same way.
Cushion against the originally modelled $2M / $6M / $20M, over 36 months: $2M: ($1.72M); $6M: ($2.39M); $20M: ($2.22M).
Cumulative cash needed, 36 months
$A Proof
$B, the plan
$C Full build
Right-sized raise
Before implementation fees unless the toggle above says otherwise. Delivery and hardware cash start 8 months after close, once certification is done and sprint letters begin converting.
Research computing, county human services and credit unions lead.
21 US segments were profiled and scored on motive, ability to pay, tolerance for an early product, the chance of a signed letter inside 90 days, and reference value. The market is not the constraint; capacity to sell and deliver is.
The top eight by fit score
Top eight segments by fit score
Rank
Segment
Fit score
Role
Why
1
Research universities (R1 and R2)In the sprint
4.15
Lead
Can run the box, tolerate an early product, and research computing directors can sign fast.
2
Health systems
4.10
Lead (once hardened)
Strongest motive and money; security review makes a 90-day Tier A letter hard, so the sprint aims for Tier B.
3
County human services, 10 county-run SNAP statesIn the sprint
3.70
Lead
The SNAP admin cost shift lands on their budget, and the signer can act inside a quarter.
4
Cities over 50kIn the sprint
3.65
Scale
An IT department and a building with a heat load; council approval is the gate above threshold.
5
Credit unions over $500MIn the sprint
3.65
Scale
Private, fast and funded, with regulators who favour controlled infrastructure.
6
State SNAP and benefits agencies
3.55
Lead (slow)
The biggest prize, but RFP-driven: Tier B in the sprint, Tier A later.
7
Tribal governmentsIn the sprint
3.50
Niche
Sovereignty is literal here; relationship-led, on a long horizon.
8
Large counties over 100kIn the sprint
3.45
Scale
Has a CIO and a data centre, and is the natural host for towns under 50k.
Channel and host, not buyers
Public power utilitiesKnows every building in town; the natural partner for demand-response payback.
Regional MSPs and integratorsResells and runs infrastructure for others; scored in the channel map, not as a buyer.
Utilities, telcos and district heatHosts and channel partners, not end customers. Preformed Line Products (PLP) is a channel candidate; nothing is signed.
Kept off campaigns
Small towns under 50kBuys through a county or council of governments, which buys once for several towns.
Law enforcementAI in policing raises the one objection that cannot be answered; serve it only inside a city or county deal.
School districtsAI plus children; can still come in as part of a city or county deal.
Defense and national securityExcluded: likely blocked by hardware-provenance rules.
Deployed on day one of a 90-day window, 4 Jan 2027 to 4 Apr 2027, with no mid-course changes. Public sector takes 81% of spend, and county human services in the 10 county-run SNAP states takes 47%. Money flows to whichever campaign's next $1,000 buys the most letters, under a founder-capacity limit and diminishing returns.
The ten campaigns of the $500k sprint
Campaign
Segment
Channel
Spend
Weighted LOIs
Tier A LOIs
K05 County HHS: independent reps (former HHS directors)
County human services, 10 county-run SNAP states
Independent reps
$89,000
2.8
1.7
K01 R1/R2 research computing: founder outbound to named directors
Research universities (R1 and R2)
Founder outbound
$82,000
5.9
3.0
K11 Credit unions: founder outbound plus LinkedIn to 748 over $500M
Credit unions over $500M
Founder outbound
$67,000
3.5
1.7
K04 County HHS (10 SNAP states): founder outbound with a SNAP cost-shift brief
County human services, 10 county-run SNAP states
Founder outbound
$62,000
2.9
1.2
K03 County HHS (10 SNAP states): state association sponsorships
County human services, 10 county-run SNAP states
State associations
$50,000
2.2
1.1
K02 R1/R2: R&E network presence plus grant co-writing
Research universities (R1 and R2)
R&E networks and grants
$45,000
2.4
1.4
K06 Counties: NACo Legislative Conference (19-23 Feb 2027)
Counties
National conference
$35,000
1.0
0.4
K10 Credit unions: America's Credit Unions GAC (28 Feb-4 Mar 2027) plus state leagues
Credit unions over $500M
National conference
$30,000
1.1
0.5
K08 Cities 50k+: LinkedIn matched audiences to CIOs and city managers
Cities over 50k
LinkedIn paid
$20,000
0.4
0.2
K18 Tribal governments: relationship programme through tribal technology associations
A Tier A letter names a site, a box count, the budget line and a signer with budget authority. Weighted LOIs count Tier A at 1.0 and Tier B at 0.4, inside the window.
The range: about 10 Tier A letters, 8 at the pessimistic quartile
5,000 simulated runsf. Box: P25 to P75. Whiskers: P10 to P90. Bar: P50. One common market-reception factor means that if the pitch misses, it misses everywhere at once.
Events inside the window
19 to 23 Feb 2027NACo Legislative ConferenceWashington DC · About 2,000 county officials · campaign K06In the window
28 Feb to 4 Mar 2027America's Credit Unions GACWashington DC · Credit union leadership · campaign K10In the window
15 to 17 Mar 2027NLC Congressional City ConferenceWashington DC · City leadersIn the window
5 to 8 Apr 2027HIMSS27Chicago · Health systems; one day after the window closesOutside
Before any close: Q4 2026
01Start reseller onboarding with a cooperative-contract holder, so a county can buy without an RFP.
02Recruit two or three independent reps on retainer-plus-commission terms, ready to start in January.
03Build the named account lists: 700 county human services agencies, 326 research computing directors, 748 credit unions.
04Check the winter meeting dates of the county associations and human services director groups in the 10 county-run SNAP states.
The first customer is expensive; the twentieth is cheap.
24 times the customers costs 6.3 times as much to deliver. Variable delivery cost per customer falls 35% as the custom-app learning curve matures and states are reused, and $570k of every 24-month bill is the fixed platform.
Headcount follows sites, not customers: S3 needs more field staff at its peak than S4 because of one 12-site county.
Peak contractor FTEs by role and scenario
Peak contractor FTEs by role
S1
S2
S3
S4
S5
Legal
0.4
0.4
0.5
0.5
0.6
Solutions / field engineer
0.5
0.6
2.9
2.7
6.3
App developer
0.9
0.9
1.6
1.6
2.1
DevOps / SRE
0.2
0.2
0.5
0.6
1.0
Project manager
0.4
0.4
1.6
1.4
3.6
Electrician
0.3
0.3
1.9
1.3
3.8
Hydronic contractor
0.3
0.3
2.4
1.6
4.8
Security / compliance
0.2
0.3
0.5
0.8
1.1
Proposal / grant writer
0.1
0.2
0.3
0.6
0.9
Customer success
0.1
0.1
0.3
0.6
0.9
Rules of thumb: 1, 5 or 10 new customers a month
Month
Showing month 12
Monthly budget for 1, 5 and 10 new customers a month
Each month
1 new a month
5 new a month
10 new a month
Delivery opex
$165k
$675k
$1.28M
Acquisition (CAC)
$100k
$500k
$1.00M
Total opex
$265k
$1.18M
$2.28M
Hardware paid out
($211k)
($1.05M)
($2.11M)
Hardware collected, net of channel cost
$316k
$1.58M
$3.16M
Service revenue
$34k
$169k
$338k
Net monthly cash
($126k)
($480k)
($891k)
Hardware working capital tied up
$1.05M
$5.27M
$10.5M
Steady intake, 3 boxes and 1 site per customer, full retention. Hardware is working capital, not spend: 5 months pass between ordering boxes and collecting payment, so each new customer a month ties up about $1.05M.
Monthly opex ≈ $23.8k + k × ($24k deal + app on the curve + $24.4k site and installs + $31.6k pipeline + $100k CAC) + $2.2k × customers + $475 × installed boxes + $15.1k × new states
Quote implementation at $90k; hold $75k as the floor.
Once reseller margin and rep commission are counted, a steady-state customer loses about ($88k)e up front at today's pricing, and service only breaks even: $45k a year against $43.6k of cost, a margin of $1,380. Growth then consumes cash in proportion to its speed.
Gross profit
Cost to win and deliver
One-time result
With the recommended fee
The fee options
Implementation fee options
Option
One-time result per customer
Today: no implementation fee, $100k CAC
($88k)
$75k implementation fee
($13k)
$90k implementation fee
$2k
$75k fee, CAC down to $60k after references
$27k
$90k
Quote implementation as a line item, about 26% of a 3-box order. Break-even is $88k.
$75k
The negotiating floor. Plan for CAC to fall toward $60k once reference sites exist.
The rate card and team size move the cash; CAC and hardware cost move the margin.
Each input was pushed to a plausible bad and good value, one at a time, and the model recalculated. The two biggest cash levers are negotiable before the raise. The two biggest margin levers need real-world data.
Funding need for the $7.5M plan, 36 months
Base $7.79M at a $75k implementation fee. Later first deals slightly lower the 36-month need because less hardware is bought inside the window, but they delay every proof point.
One-time result per customer
Bad case
Good case
Test with real money first
01Conversion rates. A $25k to $50k pre-raise test on the three campaigns that carry the sprint: research computing outbound (K01), county human services outbound (K04) and one contracted rep (K05). Settles the LOI forecast within one quarter.
02Emberholt's real cost. Send the RFQ for the platform, enclosure and heat-rejection path. Moves every customer by up to $60k.
03Three contractor quotes. An installation subcontractor in one target state, a managed SRE and 24/7 monitoring firm, and a compliance firm for SOC 2. The rate case alone swings the $7.5M plan by $2.5M.
04The implementation fee. Quote implementation as a line item in every LOI conversation and record the reaction. Break-even is about $88k; the question is whether buyers accept it or CAC must fall instead.
05Channel terms. Get the actual margin from one cooperative-contract reseller and the commission from one rep. Together about 10% of hardware in the model.
Every row is yellow in the model: proposed, or a placeholder a later step replaces. Change one on the Assumptions tab and every tab that depends on it recomputes. Valuation is deliberately not stated; the second close is priced on the letters of intent.
Assumptions still proposed or placeholder
ID
Assumption
Current value
Status
A8
Sprint start date (planning)
4 Jan 2027
Proposed
D2
Deposit at order, public-sector customer
0%
Proposed
D3
Deposit at order, private customer
30%
Proposed
D4
Payment terms after acceptance (days)
45 days
Proposed
D5
Order to delivery lead time (weeks)
12 weeks
Proposed
E1
Tier A weight: signed, names site, box count, budget source, signer with budget authority
1
Proposed
E2
Tier B weight: signed non-binding letter of interest
0.4
Proposed
E3
Tier C weight: qualified meeting or pilot request, unsigned
0
Proposed
E4
Tier A to paying customer conversion
30%
Placeholder
E5
Tier B to paying customer conversion
8%
Placeholder
E7
LOI to purchase order, private (months)
4 months
Placeholder
F1
Maximum share of sprint budget in one campaign
25%
Proposed
F2
Minimum distinct segments across the 10 campaigns
3 segments
Proposed
F3
Efficiency floor: each campaign's LOI/$ as share of portfolio average
35%
Proposed
F4
Outcome optimised
P25 of weighted LOIs
Proposed
F5
Monte Carlo runs
5000 runs
Proposed
F6
Maximum campaigns aimed at private segments (credit unions, manufacturers)
2 campaigns
Proposed
G1
Who sells during the sprint
Founders + 1 contract AE
Proposed
G2
Excluded segments
Defense: opportunistic only
Proposed
G4
Planning horizon for staffing and rules of thumb (months)
24 months
Proposed
G5
Internal sales capacity: qualified conversations founders plus 1 contract AE can run in 13 weeks
250 conversations
Proposed
G6
Rep-led campaigns (reps carry their own conversations)
K05, K15, K18
Proposed
H1
Proposed platform fee per customer per year
$24,000
Proposed
H2
Proposed service fee per box per year
$7,000
Proposed
H3
Use proposed pricing in Steps 5 and 6? (Yes / No)
Yes
Proposed
H4
Implementation fee per new customer (one-time professional services)
$0
Proposed
H5
Channel cost: reseller margin on cooperative-contract sales plus rep commission, as a share of hardware revenue
10%
Proposed
H6
Contract AE loaded cost per month
$15,000
Proposed
H7
Months from raise close to first scenario deal (certification done, sprint-sourced POs arriving)
8 months
Proposed
H8
Sprint pipeline carried: prospects x months (sprint LOIs kept warm before they convert or drop)
Every figure on this page is read from the model workbook when the page is built, so a change on its Assumptions tab re-flows here on the next build.
00Frame and assumptions Fixed the reference customer (3 Emberholt boxes) and defined what counts as a Tier A letter.
01Customer segment atlas Profiled 21 US segments and ranked them by fit and by 90-day Tier A likelihood.
02Channel map Priced 16 channels and 22 campaign candidates, and set the in-window events and pre-sprint work.
03The $500k campaign portfolio Chose ten campaigns under diminishing returns and founder capacity, then ran 5,000 simulations.
04Delivery workload and rate card Built unit costs per prospect, deal, site, box, customer and state at researched US rates.
05Staffing and rules of thumb Ran five staffing benchmarks over 24 months and derived budgets for 1, 5 and 10 new customers a month.
06Investment options ladder Laid out what $2M, $6M and $20M each buy over 36 months, then right-sized them.
07Red team Added sales costs and realistic timing, stress-tested every major input, and listed what to test first.
08The ask Settled $7.5M in two closes, with the gate between them.
Notes
fForecast. Letter-of-intent counts come from 5,000 simulated runs of the campaign portfolio; "8 to 10 signed letters" is the 25th to 50th percentile of those runs. Every conversion rate is an estimate until the pre-raise test measures it.
rRounded. Line items are rounded to $0.01M, so the second close's items show $3.49M against $3.50M unrounded. No item is missing.
eEstimate. A planning figure from the model at mid US market rates, with no implementation fee unless stated.