# Quality of Revenue ## The revenue exists, but most of it comes from services, and the reports have gaps The company describes three revenue channels. The first channel is **platform deployment**. It has an initial fee and an annual maintenance fee. The second channel is the **integration of domain packs**. It is an annual subscription for customers that have AI infrastructure. The third channel is **deep template deployment**. It has an initial fee and a maintenance fee for pre-built industrial templates. The first channel is the largest share of the current revenue. About half of the recognized revenue does not occur again. > [!warning] Signals of low confidence in reported numbers > - One fiscal year has **more than one version** of revenue in the documents. The projections show a third and higher "actual" figure. > - Signed agreements do not back most of the revenue that the company plans for the current year. A financial model projects revenue that is almost two times the revenue schedule. > - A customer declined a subsequent phase of the one project that the customer formally accepted. The company booked part of the value as other income. > - Cash is low and the company relies on milestone-based payments. Thus one late milestone makes the runway much shorter. > - Operating margins are positive. At the same time, the company does not include the costs of deployment in cost of goods sold. ## The conditions for recurring revenue Annual maintenance fees and subscriptions give recurring revenue only to the degree of the renewal rate. The customer group is new. Thus most renewals have no test at this time. Some customers receive a subsidy. For these customers, renewal at full price also has no test ([[Institute Channel and Subsidy Dependence]]). > [!tip] The important reconciliation > The reconciliation compares three values for each customer. The three values are signed value, cash that the company collects, and recognized revenue. The reconciliation shows the subsidy contingencies. [[Open Questions and Numbers to Reconcile]] gives more information. ## Why it matters Models of investor profit in this category use exit multiples from comparables of industrial software and of AI-native platforms. The multiple for revenue from services is lower. Thus the analysis must show the range of profit for the two frames. [[Platform or Solutions - The Central Question]] gives more information. ## Related - [[Foreman MOC]] - [[Delivery Economics and the Scaling Ceiling]] - [[Industrial AI Unit Economics]]