research

​

Overview

​

Usage

arckit research "<topic, e.g. 'CRM platforms for charity', 'API management tools'>"

​

How It Works

1. Reads project requirements - Extracts functional and non-functional requirements
2. Identifies research categories - Determines what needs to be researched (compute, data, integration, etc.)
3. Conducts web research - Uses WebSearch and WebFetch to gather:
   • Vendor pricing and features
   • Product reviews and comparisons
   • Open source alternatives
   • UK Government Digital Marketplace options (GOV.UK services)
   • Compliance and security data
4. Build vs buy analysis - Compares building in-house vs purchasing solutions
5. 3-year TCO comparison - Calculates total cost of ownership for each option
6. Writes research document - Saves findings to projects/{project}/research/ARC-{PROJECT_ID}-RSCH-v1.0.md
7. Spawns reusable knowledge - Creates vendor profiles and tech notes (unless --no-spawn flag is used)

​

WebSearch and MCP Server Usage

• WebSearch - To discover vendors, products, and services matching requirements
• WebFetch - To extract detailed information from vendor websites, pricing pages, documentation
• MCP servers (if configured) - For cloud-specific research (see aws-research and azure-research)

• Vendor identification searches
• Pricing page fetches
• Review site searches (G2, Capterra, TrustRadius)
• Digital Marketplace searches
• Compliance documentation fetches

​

Output

​

Research Document

• Categories researched - Technologies, services, and products identified
• Vendor landscape - Commercial vendors, open source, cloud services, GOV.UK platforms
• Build vs buy recommendations - Strategic analysis for each category
• 3-year TCO comparison - Cost breakdown by option (build, SaaS, open source, cloud)
• Requirements coverage - How each option satisfies requirements
• Top vendor shortlist - Recommended vendors for procurement
• Risk analysis - Vendor lock-in, compliance, support risks
• Next steps - Recommended follow-up actions

​

Spawned Knowledge Files

• One per vendor evaluated in depth (3+ data points)
• Reusable across projects
• Updated rather than duplicated if already exists

• One per significant technology finding (2+ substantive facts)
• Captures technical insights for future reference

​

Build vs Buy Analysis

• Genesis/Custom evolution stage (< 0.50 on Wardley Map)
• Competitive advantage potential
• No suitable market solutions
• Strategic IP worth developing

• Product/Commodity evolution stage (> 0.50)
• Mature market with multiple vendors
• Standard business capability
• Faster time to market needed

• Commodity services (> 0.75 evolution)
• Cloud-first strategy
• Minimal customization needed
• Subscription model preferred

​

TCO Comparison Example

### 3-Year Total Cost of Ownership

| Option | Year 1 | Year 2 | Year 3 | Total | Notes |
|--------|--------|--------|--------|-------|-------|
| Build In-House | £120K | £80K | £80K | £280K | Dev costs Year 1, maintenance thereafter |
| SaaS Vendor A | £45K | £48K | £51K | £144K | Annual 6% price increase |
| Open Source | £30K | £25K | £25K | £80K | Support subscription + integration |
| Cloud Managed | £36K | £38K | £40K | £114K | AWS managed service + data transfer |

​

Flags

​

Integration with Other Commands

• requirements - Requires ARC-*-REQ-*.md to exist
• data - Uses data model for database/storage research
• stakeholders - Uses stakeholder needs for vendor evaluation

• wardley - Evolution positioning feeds into Wardley Maps
• sobc - TCO data feeds into Economic Case
• sow - Vendor requirements feed into RFP generation
• hld-review - Validates technology choices against architecture

​

UK Government Specific Research

• GOV.UK Notify (notifications)
• GOV.UK Pay (payments)
• GOV.UK Design System (frontend)
• GOV.UK PaaS (hosting)
• GOV.UK Verify (authentication)

• G-Cloud search for SaaS/IaaS/PaaS
• DOS (Digital Outcomes and Specialists) for bespoke development
• Framework compliance checks

• Point 3: Open Source options
• Point 5: Cloud First compliance
• Point 8: Share and Reuse opportunities
• Point 11: Procurement strategy

​

Examples

​

Example 1: CRM Research

arckit research "CRM platforms for charity with donation tracking"

• SaaS CRM vendors (Salesforce Nonprofit Cloud, Microsoft Dynamics 365)
• Open source CRM (CiviCRM, SuiteCRM)
• Custom build estimate
• 3-year TCO comparison
• Digital Marketplace G-Cloud options

​

Example 2: API Management

arckit research "API management tools for government services"

• Cloud API gateways (AWS API Gateway, Azure API Management)
• Open source (Kong, Tyk)
• GOV.UK API standards compliance
• Security and authentication options
• Government cloud compatibility (IL3, OFFICIAL)

​

Example 3: Quick Research Without Spawning

arckit research "payment processing options --no-spawn"

​

Next Steps

1. Create Wardley Map - arckit wardley to visualize evolution positioning
2. Build Economic Case - arckit sobc to feed TCO data into business case
3. Generate RFP - arckit sow to create Statement of Work for procurement
4. Validate Architecture - arckit hld-review to check technology choices

​

Related Commands

• aws-research - AWS-specific service research using MCP
• azure-research - Azure-specific service research using MCP
• wardley - Strategic mapping from research findings
• sobc - Economic case with TCO data
• sow - RFP generation from vendor requirements