Earn from successful delivery.
Model your share of contracted delivery revenue. Your opportunity depends on usable capacity, eligible traffic, the agreed delivery price and your revenue share—not simply how many devices you install.
OMNIMESH FOR SERVICE PROVIDERS
Make your modem refresh earn more.
Explore a 10,000-node OMT rollout alongside a planned DOCSIS 4.0 modem refresh: paid storage, local delivery and measurable network savings.
01
Earn storage and delivery revenue
Model paid storage and your share of contracted delivery.
02
Reduce network spend
Estimate the paid traffic costs your deployment could avoid.
03
Build a rollout that fits
Compare investment, operating costs and cash payback.
Illustrative business case. Adjust the starting assumptions to your network and proposed terms; results are not a quote or guaranteed return. Inputs stay on this page until you choose Save analysis. Saved analyses are shared with signed-in site members.
How OMT creates value — and how to build your proposal ↓
Saved analyses · shared cloud workspace
Sign in with your Milestone account to save and reopen analyses.
New analysis. Save captures the current calculator inputs and results.
Shared records retain prior revisions and the saving user's identity. Follow-up dates are tracked here; no reminder emails are sent. Saved results are planning estimates, not approved quotes.
10,000 integrated nodes. 8 TB each.
The starting case assumes 60% of usable storage earns $10/TB/month. It retains your $12 one-time license, $12 monthly support and 5% OMT revenue participation. Delivery revenue starts at zero.
Incremental cost: $50 compute + $400 storage + $12 license + $50 allocated engineering per node = $512. The engineering allowance totals $500,000 at 10,000 nodes and is assumed payable with the rollout. An earlier engineering payment would increase the initial funding requirement. The normal modem and its planned installation belong to the operator’s existing refresh budget; additional OMT installation is assumed zero. All are planning assumptions, not OEM quotes.
8 TB integration, heat, power, modem certification, firmware isolation and storage contracts require validation. A shared enclosure does not make storage hardware free. Ten-year hardware life is a planning assumption with a separate 5% annual failure reserve.
Each deployment can strengthen the whole network.
Your current rollout supplies network A. Add a prospective operator network B to explore the potential benefit of connecting their resources.
100% illustrates the theoretical ceiling, not expected adoption. Try a lower fraction to reflect locality, availability, permissions and demand. This fraction is a scenario assumption, not a measured probability.
Resource capacity adds; potential connections can grow faster. The connection index is dimensionless. It is not dollars, company valuation, throughput or a multiplier applied to revenue. Operator cash flow above covers network A only.
Framework: separate networks n² + m²; fully connected network (n + m)²; potential uplift 2nm. This illustration uses n² + m² + q × 2nm, where q is your assumed realized fraction. The q adjustment is OMT scenario methodology, not a formula attributed to Van Alstyne.
See the monthly financial breakdown
OMT gross is context only; it is not added again to provider benefits. Content-provider origin-egress savings are excluded from the ISP’s savings.
Cash through the rollout
| Month | Active nodes | Monthly net | Cumulative |
|---|
Installation costs occur as nodes arrive. Hardware renewal occurs at the entered life, in addition to the failure reserve. Depreciation is not charged again in cash flow. All projected commercial earnings remain conditional on agreements and settlement acceptance.
Compare changes to your business case
Steady monthly net with one assumption varied: 50%, 100%, 150% of baseline. Cache share uses −20 / baseline / +20 percentage points, bounded at 0–100%.
| Assumption | Lower input | Baseline | Higher input |
|---|
Review network capacity
Presets are planning values. Confirm access capacity and placement with your engineering team.
FROM OPPORTUNITY TO AGREEMENT
One modem refresh. A new storage business.
Integrating OMT into a planned modem shipment could share the enclosure, power supply, distribution and installation workflow. The operator adds paid storage capacity and can serve eligible content closer to subscribers. This is an OEM product proposal; DOCSIS 4.0 itself does not include OMT storage or guarantee locality, revenue or savings.
The starting network case uses 50,000 subscribers across 200 service groups, with 50 enrolled nodes per group, an 8 Mbps node cap and 50% upstream reserve. The assumed 2 Gbps shared upstream is a planning input, not a promised DOCSIS service rate. Confirm routing, congestion, failure domains and repair traffic with engineering. DOCSIS 4.0 background from CableLabs.
Close the proposal around three commitments: an OEM incremental-cost quote, contracted paid storage demand, and a pilot that measures peak invoice reduction without degrading subscriber service. Existing cable construction costs are not counted as savings; no speculative upgrade deferral is included in the starting case.
OMT is designed to turn participating network capacity into a delivery service. Use this tool to agree on the economics first, then prove them in a focused pilot.
Keep more eligible traffic local.
Delivery from participating nodes can displace paid upstream traffic. The calculator limits savings to demand outside existing caches and uses your marginal transit and peering costs, so the business case reflects costs you could actually avoid.
Start with a rollout you can fund.
Choose who buys the hardware and funds installation, household benefits, power and support. Review the phased cash curve before committing to scale. Ownership changes your investment; it does not make the system’s costs disappear.
Two additional opportunities: backhaul and storage
Private backhaul savings: enter only the separate bill you can reduce or retire, the avoidable percentage, any overlap with transit/peering already counted, and the month the saving begins. Fixed circuits do not save money merely because traffic falls. Savings begin at the chosen month without automatic rollout scaling; choose that month only when the required rollout and contract changes are achievable.
Storage rental: allocate a percentage of usable, coding-adjusted disk capacity to rental. Revenue is capped at the smaller of that allocation and contracted TB, multiplied by the rental rate and provider share. Additional storage operating cost is deducted; hardware is already counted once. Rental scales with rollout. Capacity allocated to rental is unavailable to the delivery cache; validate that the remaining cache can support your delivery assumptions. This model does not predict the resulting cache-hit change.
The favorable cable example includes assumed storage rental demand; separate private backhaul savings remain zero. These benefits improve ROI only when additional revenue and avoidable costs exceed additional expense. Keep storage rental separate from any storage already bundled into delivery pricing. Original v1 saved analyses reopen with both opportunities set to zero; later records retain their saved inputs.
Build a resource network that becomes more useful as it grows.
OMT’s opportunity extends beyond selling software or installing storage. Participating operators contribute capacity, locations and access to customers. A shared platform can make these resources available to more workloads and paying customers, creating reasons for additional operators to participate.
More operators → more resources and coverage → more useful services → more paying demand → stronger incentives to join.
Marshall Van Alstyne and Geoffrey Parker describe demand-side economies of scale: participants create value for other participants, while a larger network can improve matching between supply and demand. This provides a framework for OMT’s platform strategy, not a prediction of its financial results.
In a Metcalfe-style illustration, connecting two equally sized networks doubles the combined connection index relative to keeping those same networks separate. Compared with just one starting network, the fully combined index is four times as large. Storage capacity only doubles. The distinction matters when explaining platform potential to an operator or investor.
Our proposed early advantage is embedding OMT into modem refresh programs and establishing operator relationships, interoperability and paid demand. A blue-ocean position and durable leadership are strategic hypotheses to validate; this tool does not establish an absence of competitors.
Turn platform potential into commercial evidence.
Track participating operators, usable capacity, coverage, active paying customers, fulfilled workloads, repeat purchases and contribution margin. Demonstrate that adding an operator increases useful service for existing participants. Cross-operator access, settlement, security and service quality must be agreed and proven.
Van Alstyne & Parker: Platform Business—From Resources to Relationships (2017) · Van Alstyne presentation: network value illustration, slide 55 (2019) · Boston University: two-sided platform feedback and pricing
Why the technology matters to the economics.
Distributed delivery
Nodes supply shares that a serving node reconstructs into an object. This creates a path for local delivery and potential paid-transit savings. Actual locality, delivery demand and upstream capacity must be confirmed on your network.
Recoverable storage
The default 5-of-7 coding needs five shares to reconstruct data, providing tolerance for missing shares when placement and availability support it. The model accounts for 1.4× storage overhead and separately models internal transfer traffic.
Capacity under operator control
Supply caps and reserved upstream help size participation around ordinary subscriber traffic. The calculator checks the proposed footprint before presenting economics. Repair and prefill traffic need additional engineering allowance.
What needs to be proven in your pilot?
Measure delivered traffic, billable-peak displacement, household impact, support effort and actual hardware performance on independent links. Pi 4 profiling has identified a software-cryptography ceiling; the current calculator does not model CPU limits. Hardware selection and measured serving throughput must therefore support the capacity you enter. No production SLA is implied by a lab benchmark.
Agree pricing, revenue share, cost ownership and settlement acceptance before treating modeled benefits as committed revenue. Production payouts require the settlement and witness gates described in the technical notes.
Use the tool in your next customer conversation.
- Set the footprint. Enter access technology, subscriber count and target nodes. Expand network assumptions if the suggested rollout does not fit.
- Agree the commercial inputs. Enter marginal transit cost, delivery price and provider share. Open traffic assumptions and “Who pays & rollout costs” together; these determine whether the opportunity pays.
- Read the whole business case. Annual net operating benefit is the full-deployment run rate after operating costs. Initial investment and hardware renewal are reflected in the phased cash forecast. Payback is shown only when recovered cash remains nonnegative through the modeled horizon.
- Save the proposal. Sign in using your Milestone account, enter the prospect and analysis name, select a sales stage, add notes and a follow-up date, then click Save analysis. Existing viewers can read; editors can save. All site members share the library.
- Compare and refine. Open a saved analysis or compare it with current inputs. Use revision history to review earlier proposals. “Save as new analysis” prepares a separate copy; click Save analysis to store it. Save business case / PDF opens your browser’s print dialog.
Technical evidence, assumptions & household protections
Useful evidence. Limited conditions.
The September 10 handoff reports 0.22s first byte and 1.75 MB/s for a 16 MiB object with 512 KiB chunks on the eight-node pilot. A shared Wi-Fi access point constrained the test. These are not guaranteed production results.
Default coding is 5-of-7. Storage overhead is 1.4×. Internal transfer overhead is separately editable: coding alone does not determine billable delivery.
Scenario revenue is not a payout.
Price, revenue share, duty cycle, support, platform fees and benefit funding require agreement. Two monthly tie-outs within 0.5% and an accepted independent witness are required before payouts. The handoff reports that witness acceptance remains blocked.
No availability or first-byte SLA has been agreed. Measure production coding on independent links.
Ordinary internet comes first.
Billing failure must not interrupt content delivery. Suspending OMT participation must never disable household internet.
Configured supply caps exclude repair traffic in the current implementation. Reserve capacity for repair, prefill and subscriber demand.
Calculation methodology
Network gate: whole nodes per group are limited by adoption and shared upstream after the reserved percentage, at the smaller of configured cap and node connection speed. Independent-market stripe viability is an average check; real placement and failure-domain diversity still need validation.
Successful contracted delivery = aggregate upstream ÷ internal transfer overhead × seconds per month ÷ 8,000 × duty cycle × eligible utilization. Decimal GB and a 30-day month are used. Repair, retries and test traffic earn no modeled revenue.
Billable-peak displacement is independently estimated from peak utilization and capped by OTT demand after existing caches. Paid transit and peering share one displacement pool. Only genuinely avoidable cache costs should be entered. An average-to-peak divisor from the handoff is not used: it cannot determine 95th-percentile savings without a traffic profile or busy-hour assumption.
2,000 × 4 TB at 5-of-7 yields 5.714 PB usable, not 5.7 PB of monthly earnings. Disk sizing remains a workload decision; low volume alone does not prove that a disk can never earn revenue. The current favorable default is 10,000 nodes with 8 TB each, 5-of-7 coding and 60% paid usable storage. Rental demand is capped by the storage allocation; the remaining 40% is available for other workloads. Occupancy is assumed as each node enters service, not a verified sales pipeline. A $75–150 install does not exceed a $300 device. Credit cost defaults to its full face value; a lower economic cost is not assumed.
Monthly cash includes provider-owned hardware and installation, support, funded benefits, funded power, expected replacement reserve and platform fees. Hardware owned by OMT or homeowners is excluded from provider capex, not from the wider system’s costs. Subscriber churn and content popularity are not simulated; lower eligible utilization or node targets to stress those risks. Avoided-cache cost scales with rollout and is a user assumption, not a claim that existing caches can be retired.
Source: omt-roi-handoff-bundle.md, exported September 10, 2026. Measurement labels describe the handoff’s reported evidence. Commercial inputs are assumptions until validated with partners.