HOME FRONT2026 measurement methodology
Use the same cohort, eligible footprint and observation window to compare orders, installations and acquisition cost.
Published October 25, 2026. Updated October 27, 2026. This is a methodology and scenario guide. Home Front has no cleared campaign dataset or independently verified industry averages to publish here.
Four planning measures
Orders per rep-day = knocks per rep-day × contacts per knock × orders per contact
Use a contact-to-order close rate. A pitch-to-order rate also needs a pitches-per-contact factor. For observed productivity, divide reconciled orders by all staffed selling days, including zero-order days, and state how partial days are counted.
Expected installs = submitted orders × (1 − assumed pre-install fallout)
This two-outcome model assumes every non-fallout order completes within the selected window. Pending orders require separate treatment. The output can be fractional because it is an expectation, not a counted customer or a forecast validated against records.
Installed footprint share (%) = unique installed addresses ÷ eligible serviceable addresses × 100
Freeze the eligible address set and dates. Homes passed, homes actually orderable and addresses contacted are different denominators. A share in one window is not an incremental lift; lift needs a comparable starting point, and causal lift needs an appropriate comparison design.
Cost per completed install = allocated acquisition cost ÷ completed installs
State which costs are included: fees, labor, travel, media, systems, credits and internal support where applicable. Keep network construction cost separate unless the comparison explicitly allocates it. With zero installs, cost per install has no defined value.
For actual outcomes, keep completed installations and installed accounts later reversed separate. Net Installed Pull-Through uses (completed installs − reversed installs) ÷ submitted orders from one frozen cohort. It is an observed-cohort measure when the input counts are reconciled; the scenario model above is not a substitute for those records.
Readiness, staffed time and retention
This proposed measurement specification extends the pilot scorecard. It supplies definitions, not observed results, industry benchmarks or performance targets. Agree the fields and counting rules with the provider before collection.
Record the provider's ready-for-service (RFS) milestone and the address's actual orderability at the time of each field activity. A construction milestone alone does not establish that an address can order service. Keep pre-RFS, non-orderable activity separate from activity at an RFS-confirmed, orderable address. Retain other combinations and unknown statuses as separate groups.
Freeze the readiness evidence, approved address release, offer, selling dates and observation cutoff for each comparison. Preserve an activity's original classification when an address later becomes orderable. An expression of interest or preregistration is not a submitted or provider-accepted service order; report it separately under its approved definition.
Verified contact events per staffed hour = verified contact events ÷ all staffed program hours
Use every recorded hour assigned to the measured field program, including travel, briefings, administration and zero-contact periods. Sum representative-hours rather than elapsed team hours. Freeze rules for breaks, partial shifts and shared time; report missing time records. A selling-time-only rate may be a separate measure with its own denominator.
For this specification, a verified contact event is a recorded two-way conversation with an authorized household decision maker. Count one qualifying event per visit ID and retain repeat visits. A knock, unanswered attempt, pitch or conversation with another person is a different event. Also report unique verified-contact addresses, counted once per cohort, divided by the same staffed hours. That second rate measures distinct address reach rather than repeat conversations. If staffed hours are zero, neither rate is defined.
Observed day-90 active retention (%) = accounts confirmed active at day 90 ÷ all matured installed accounts in the frozen cohort × 100
Define day 90 as 90 elapsed days after the original provider-confirmed installation timestamp. An account matures only when that point is on or before the report cutoff. Query the provider's historical status at that point; today's active status cannot establish day-90 status. This is point-in-time active retention, not proof of uninterrupted service throughout the period.
| Group | Treatment |
|---|---|
| Matured installed accounts | Retain the original installed cohort. Classify each account as active at day 90, inactive at day 90, service installation reversed by day 90, or unresolved at day 90. Use one mutually exclusive outcome per account, with service-installation reversal taking precedence. |
| Unresolved day-90 status | Keep these accounts in the matured denominator. Until all outcomes are resolved, label active accounts divided by the full matured cohort as the confirmed-active share, a lower bound, rather than a complete retention rate. Show the unresolved count. |
| Installed but not yet mature | Show separately. Do not classify these accounts as retained or lost at day 90. |
| Pending or canceled before installation | Keep visible in the original order funnel. These orders have no completed-installation timestamp and do not belong in the installed retention denominator. |
Show service-installation reversals by their effective date and keep later corrections visible. A commission or billing adjustment that leaves service active is a commercial adjustment, not automatic customer loss. Replacements must link to their original record under one frozen acquisition-counting rule. With no matured installed accounts, retention has no defined value.
For a hypothetical cohort of 100 submitted orders, assuming all non-fallout orders finish by the same cutoff, 28% fallout implies 72 installs and 11% implies 89. That is 17 more modeled installs, a 17 percentage-point increase in installation conversion, and a 23.61% relative increase over 72.
These percentages are editable example assumptions supplied for illustrating the arithmetic. They are not industry averages, Home Front results, competitor results or evidence that a particular CRM caused an improvement. Changing order volume or footprint as well would answer a different question.
Load or edit a scenario in the simulatorNo campaign is presented as a Home Front case study here. Before publishing one, reconcile the following records and obtain permission to disclose the market, provider and results.
| Record | Definition to retain | Review before publication |
|---|---|---|
| Cohort and window | Unique order IDs, selling dates, staffed rep-days, eligible address file and observation cutoff. | Include zero-output days and all stated exclusions; avoid choosing only the best dates. |
| Provider outcomes | Accepted orders, completed installs, unresolved orders, cancellations and post-install reversals. | Reconcile the original order set to the provider's actual status records. |
| Economics | Agreed payable event, allocated costs, adjustments and costs excluded from the calculation. | Do not present submitted orders, commission eligibility or billed fees as cash received. |
| Permission and context | Authorized client naming, source owner, method, comparison basis and known limitations. | Separate a company-reported result from independent verification and a causal experiment. |
Velocity One's Glo Fiber case study reports customer orders as a share of a stated serviceable footprint. Its rep-selling-day definition counts days on which a rostered representative recorded an order. Those definitions differ from completed-install penetration and all-staffed-day productivity. The distinction is a comparison question, not evidence that the company concealed installations.