Plans and engagement models
Five levers, not three bundles
The platform, the depth in your market, the work to deploy it, the service around it, and the scope you add later. Bundling those produces impressive contracts and disappointing delivery, so we price them apart.
Platform
The subscription, priced on how much of the organization it covers.
Vertical Intelligence Pack
The maintained depth in one market.
Implementation
Connecting your data and describing it correctly.
Service model
Who you get, and what we commit to contractually.
Scale
Domains, sources, environments and rights, added as you grow.
What every agreement includes
Trust is the product, so none of it is a premium feature. There is no cheaper edition that produces less defensible answers, and no tier at which a broken thing is treated casually.
- Ask, Views and Evidence
- Governed, read-only querying
- Validation before execution, verification afterwards
- Access controls applied before a question is planned
- A production Roots environment
- Incident handling, service restoration and security escalation
01
Platform
Editions differ by how much of the organization is in scope and which modules are included — never by how defensible the answers are.
Core
Narrower, not weaker. This tier is about how much of the organization is in scope, not how far you can trust the result.
- One domain or use case
- Two or three standard sources
- Around 25–75 active
- DeepGarden-managed
- Included
Scale
Broad enough to matter across a function, without promising an organization-wide transformation.
- Two to four domains or major use cases
- Around five to ten
- Around 100–500 active
- Shared management with your team
- Advanced cross-source questioning
Enterprise
Greater scale, governance and contractual commitment. Enterprise does not mean unlimited customization — that is what Managed Intelligence is for.
- Multiple domains and business units
- Contracted enterprise scope
- Broad internal, or external-facing deployment
- Full enterprise governance
- Organization-wide
02
Vertical Intelligence Packs
The platform above is the same wherever you are. This is the part that is not: one market's data collected, normalized, resolved and kept current, with a certified metric library over the top. Licensed per market, alongside the platform or on its own.
Healthcare — Medicare Advantage
Six maintained CMS datasets, crosswalked and joined. Useful on day one without a single connection to your systems.
- Enrollment, penetration, Star Ratings, landscape, crosswalk and benefits
- Contracts, plans, organizations and counties resolved
- Certified market-share and share-shift definitions
Energy efficiency
Available through implementationTechnical Reference Manuals reconciled across jurisdictions and program years. The measure model is built with your team rather than maintained as a DeepGarden dataset.
- Measure characterizations for the jurisdictions in scope
- Measures resolved across jurisdictions and program-year revisions
- Certified savings, realization and adoption definitions
- Available through implementation
- The capability exists. Getting it working for you is part of the implementation, because it depends on your sources and definitions.
03
Implementation
Priced separately because it is not the same job twice. Three clean warehouse tables and a health plan with claims, enrollment and PHI controls are different engagements.
Standard implementation
One domain, a small number of clean sources, and a defined use case to prove out.
Roughly six to ten weeks
Advanced implementation
Several domains, more complex semantic modeling, and change management for a wider group of users.
Roughly ten to sixteen weeks
Enterprise programme
Multiple business units and a real data architecture. Run in phases rather than as one delivery.
Several months, in defined phases
A validated starting library of business metrics and definitions
Every implementation produces one, and it is the part that decides whether any of the rest works. It is the approved language and calculation logic DeepGarden is permitted to use when it answers a question — settled once, in writing, by the people entitled to settle it.
Size matters much less than status. Twenty definitions your finance director has ratified are worth more than two hundred somebody drafted.
04
Service models
Who you get, how often you see them, and what we commit to contractually. Fixing broken functionality is not what these buy — that is included at every level.
Standard
Pooled coverage rather than a named person.
Partnered
A named person who knows your setup, on a monthly cadence, with contractual response targets behind them.
Strategic
A team rather than a contact, working to a joint plan, with the strongest commitments we offer.
05
Scale
Scope grows in defined units, so adding a domain or a dataset does not mean reopening the whole agreement.
- Additional data source or custom connector
- Additional business domain
- Additional environment
- DeepGarden Data dataset licence
- Embedded or customer-facing deployment rights
- White-label rights
- Extended support coverage
- Managed Intelligence retainer
Proof of Value
DeepGarden needs real data and real definitions to show you anything worth seeing, so the way in is a fixed-scope engagement rather than a login. Part of the fee is credited against a first-year subscription.
- Eight to twelve weeks
- One business problem, one domain
- One or two data sources
- Success criteria agreed before we start
- No production service commitments
- No unlimited customization
Managed Intelligence
A monthly retainer for teams who would rather we operated the analysis function than staffed one. Defined delivery capacity against a backlog you prioritize — never unlimited customization.
- New metric and domain-model development
- New Views and recurring intelligence
- Query and prompt optimization
- Data-quality investigation
- Analyst support for specific questions
Get a quote that reflects your scope.
Tell us the scope you have in mind and we will come back with a number, and with what we would cut first if it needs to be smaller.