Begin with the customer and task
A useful deployment record identifies the customer or customer type, the task, the site or setting, and the observation period. If those details are missing, state the gap rather than assume the result applies elsewhere.
Read each claim separately
| Question | Evidence to look for | What remains separate |
|---|---|---|
| Has it been delivered? | A dated delivery or acceptance record | Payment and productive use |
| Is someone paying? | A disclosed contract, revenue, or customer confirmation | Autonomous operation and profitability |
| Does it complete the work? | Task definition, attempts, acceptable completions, duration | Performance on other tasks or sites |
| How much help does it need? | Intervention definition, frequency, and human time | Whether assistance is remote or on site |
| Does the deployment repeat? | Repeat orders or comparable results at additional sites | Announced capacity or planned rollout |
Demonstrations, trials, paid deployments, and repeat purchases describe different evidence. Do not infer payment from a named partnership, or autonomy from a revenue figure.
Compare cost over the same period
A useful comparison divides the cost attributable to the deployment by acceptable tasks completed in that period. Include equipment depreciation or lease charges, integration, maintenance, energy, supervision, recovery work, and downtime where measurable.
State the accounting period and assumptions. Keep actual costs separate from modeled costs. If completed-task counts or human support costs are unknown, leave the comparison incomplete.
Read company evidence in context
Company pages provide a structured commercial-evidence panel. An empty field means that the structured assessment has not been completed; it does not mean the company has no customers or revenue. Product descriptions and existing research remain available separately.
Every populated assessment needs a source, evidence class, date, task scope, and limitation. Read Sources & Research Standards for the labels. Commercial evidence can inform the CE-N review, but does not automatically change a bottleneck score.