§01The first step
Agentforce Readiness Assessment
Find out where AI will pay off in your Salesforce, which tool should do each job and what has to be fixed first. Fixed scope, fixed price, and an answer you can use with or without us.
§02At a glance
An answer before a build.
- Who it’s for
- Salesforce leaders, admins, RevOps, service and operations leaders being asked “what's our AI plan?” who need an answer that covers Agentforce, Claude and the other options.
- Format
- Fixed scope, fixed price.
- Duration
- Express: one week. Standard: two weeks. Enterprise: three weeks.
- Access we need
- Read-only access to production or a full-copy sandbox. About 4 to 6 hours of your team's time.
§03The problem
Check the foundations first.
There are several ways to put AI on top of Salesforce: native Agentforce agents, Salesforce in Claude, marketplace listings and custom agents. Choosing one before checking data quality, access, content and licences can stall a project or expose more information than intended. We check those foundations before recommending a tool.
§04Your report
What you get.
- A readiness scorecard across seven areas: data quality, security and access, automation health, knowledge and content, licences and platform, AI tools in use, and people and process.
- A shortlist of opportunities ranked by value and feasibility. Each gets a recommendation: Native Agentforce, Salesforce in Claude, an AgentExchange listing, an Arkeforce agent, or not yet.
- A people-or-records decision for each use case: help someone asks for, or a job that must happen on every record.
- A fix-first list with effort estimates, including what an AI assistant would be able to see as each of your users.
- A cost picture split three ways: what you would pay Salesforce, anyone else such as Anthropic, and Arkeforce.
- An ROI estimate built from your own case, lead, opportunity or record volumes.
- A 90-day roadmap sequencing the fixes and first use cases.
- A leadership readout and a written report you own.
§05The process
How it works: Standard tier, two weeks.
| When | What happens | Your time |
|---|---|---|
| Before kickoff | Confirm scope, AI and data-use terms, and read-only access | 30 minutes |
| Week 1 | Kickoff interviews, automated org scan, data quality sampling, and a licence, feature and AI tool review | 3 to 4 hours for interviews |
| Week 2 | Use-case scoring and tool fit, ROI and cost model, fix-first list, roadmap and readout | 1.5 hours for review and readout |
§06Seven areas
What we look at.
| Area | What we check |
|---|---|
| Data quality | Duplicate rates, completeness of key fields, stale records and picklist sprawl on the objects your use cases depend on. |
| Security and access | Admin-level and view/modify-all access, sharing, guest and integration users, and Security Health Check. Broad user access becomes broad AI access. |
| Automation health | Flow and Apex errors, overlapping automation, old API versions and inactive or unused automations. |
| Knowledge and content | Coverage of your top case reasons, freshness, ownership and approved answer sources. |
| Licences and platform | Edition, Agentforce entitlements, Flex Credits, Data Cloud, Slack, Sales and Service add-ons, and eligibility for Salesforce in Claude. |
| AI tools in use | The tools teams already use or pay for, their connections and integration users, and your written AI policy. |
| People and process | Sales methodology, case handling, approvals, owners for AI outcomes and change readiness. |
§07Decisions you can use
What you get at the readout.
- An agreed first use case, the tool it should run on and a business case.
- Specific risks to fix before anything is switched on.
- A written reason for every recommendation, including where a native or marketplace option is enough.
- A roadmap with effort, owners and dependencies.
§08Fixed scope
Assessment tiers.
Prices in USD.
| Tier | Your org looks like | Duration | Fixed price |
|---|---|---|---|
| Express | One cloud, one business unit, under 50 users. One target area, such as service. | 1 week | $2,500 |
| Standard | One or two clouds, up to 250 users. Two target areas. | 2 weeks | $6,500 |
| Enterprise | Multiple clouds or business units, 250+ users. Complex security or integrations. | 3 weeks | $9,500 |
The fee is credited against any Arkeforce engagement signed within 60 days, up to $2,500 USD and never more than half the engagement price.
§09Your choice
What happens next.
Follow the report into a pack, Claude for Salesforce Teams, both or neither. You own the report if you build with someone else, use Salesforce’s own tools or decide to wait.
§10Access and control
Trust and safety.
- Read-only: we never change your org during the assessment.
- We work in your environment, sample only what is needed and keep no client data after the readout.
- We confirm your AI and data-use policies, and your own clients' where they apply, before any tool connects.
- We recommend native agents, Salesforce in Claude, a marketplace listing or nothing whenever that fits better than something we would build.
§11Next step
Bring the question.
Tell us where your team is starting and which jobs you have in mind. We’ll confirm the assessment scope before you sign.