Embedded Analytics Rule: Price the Dashboard as a Retainer Line, Not a Build
Should an agency sell embedded analytics as a one-time integration project or as a recurring line item inside the client's existing retainer? Attach embedded analytics to the recurring retainer as a metered feature line, and refuse to scope it as a standalone one-time build.
By InnovaAI ResearchPublished
“Should an agency sell embedded analytics as a one-time integration project or as a recurring line item inside the client's existing retainer?”
Attach embedded analytics to the recurring retainer as a metered feature line, and refuse to scope it as a standalone one-time build.
Quoting embedded analytics as a fixed-scope integration project, then absorbing every new dashboard request, tenant addition, and schema change into unpaid maintenance. The build fee gets collected once while the customization work continues indefinitely, and the agency has given away the recurring revenue that made the category worth entering.
The category's economics come from subscription attachment, not integration labor: platforms such as Luzmo, Qrvey, and Solien ship multi-tenant isolation, connectors, and billing layers so the agency's remaining work is configuration and branding rather than infrastructure. That shifts the billable unit from hours to tenants, which is why a per-tenant line inside an existing retainer compounds while a fixed build fee does not. The same pattern is visible across the wider stack, where Forrester's Q3 2026 research found workflow integration, not model capability, is the bottleneck separating agencies that scale AI profitably from those running one-off experiments.
- •The client already pays a monthly platform or software retainer and the analytics feature would sit inside that product
- •The agency is choosing between a fixed-fee build quote and a per-tenant or per-seat recurring charge
- •More than one end customer will log into the same dashboard deployment under separate tenant logins
- •The client's procurement team treats new software features as operating expense rather than capital project spend
- •The agency has no internal data engineering team and would rely on a vendor's connectors and hosting