Vendor Evaluation

Evaluating vendors — whether for software, services, or partnerships — involves gathering scattered information and making apples-to-apples comparisons. This recipe structures the process so agents handle the research while you focus on the decision.

The Pipeline

DEFINE    → Set evaluation criteria and weight them
RESEARCH  → Gather data on each vendor
SCORE     → Rate vendors against criteria
RECOMMEND → Final comparison with recommendation

Step 1: Define Criteria

Start by creating a database to structure the evaluation:

@Assistant create a database called "Vendor Evaluation - CRM" 
with columns: vendor (text), pricing_score (number), 
features_score (number), ease_of_use_score (number), 
support_score (number), integration_score (number), 
total_score (number), pricing_notes (text), 
features_notes (text), recommendation (text)

Then define what you're looking for:

@Assistant create a document titled "CRM Vendor Criteria" with:

Evaluation criteria for a CRM serving a 50-person sales team:
1. Pricing (weight: 25%) - Total cost for 50 seats, contract terms
2. Features (weight: 30%) - Pipeline management, reporting, automation
3. Ease of Use (weight: 20%) - Onboarding time, UI complexity
4. Support (weight: 15%) - Response time, channels, documentation
5. Integration (weight: 10%) - API quality, existing integrations

Vendors to evaluate: HubSpot, Salesforce, Pipedrive, Close, Freshsales

Upload Existing Data

If you already have vendor data in spreadsheets, upload them to get a head start:

1. Upload the spreadsheet (CSV or Excel) to a folder in the data tree
2. Click Convert to Database — Saltare infers column types automatically
3. Rows import in the background, ready for agent analysis

@Analyst I've uploaded and converted "Vendor Shortlist.csv" to a database.
Add a scoring column and fill in the evaluation criteria from the CRM Vendor Criteria document.

You can also upload vendor proposals, spec sheets, or contracts as PDFs. Agents will read the extracted text:

@Researcher review the uploaded "Acme Corp Proposal.pdf" 
and extract pricing, SLA terms, and integration capabilities 
into the Vendor Evaluation database

Step 2: Research Each Vendor

Create research tasks for each vendor:

@Researcher for each vendor in the CRM Vendor Criteria document,
create a subtask and research:
- Current pricing for 50 seats (monthly and annual)
- Key features and any missing features from our requirements
- G2/Capterra ratings and common complaints
- Integration options with our existing stack
- Support options and SLA
- Any recent news (acquisitions, outages, pivots)

Write findings into a document titled "[Vendor Name] - CRM Evaluation"

The agent will create 5 research sub-tasks and work through them systematically, creating a document for each vendor.

Step 3: Score and Compare

Once research is complete, populate the scoring database:

@Analyst review each vendor evaluation document and score 
each vendor in the Vendor Evaluation database:
- Score each category 1-10 based on the research findings
- Calculate total_score as weighted average using the criteria weights
- Add brief notes explaining each score
- Set recommendation to "strong yes", "yes", "maybe", or "no"

Step 4: Final Recommendation

Generate the decision document:

@Writer create a document titled "CRM Vendor Recommendation" using 
the scoring database and individual vendor documents:

1. Executive Summary (recommended vendor and why, 2 sentences)
2. Scoring Comparison Table (all vendors side by side)
3. Top Pick: Detailed Analysis
   - Why this vendor wins
   - Key risks and mitigations
   - Implementation timeline estimate
4. Runner-Up: Why it's second
5. Vendors Not Recommended: Brief explanation
6. Next Steps (trial, POC, contract negotiation)

Making the Decision

The dossier gives your team everything needed for a decision meeting:

@Assistant post in #leadership: 
"CRM vendor evaluation complete. Recommendation document ready 
for review. Key finding: [top vendor] scores highest on features 
and integration, with competitive pricing at $X/seat/month. 
Please review before Thursday's decision meeting."

Variations

Software Tool Evaluation

Same pattern but add criteria for:
- Security and compliance (SOC2, GDPR)
- Data migration complexity
- API documentation quality

Agency/Consultant Selection

Adapt criteria to:
- Portfolio quality and relevance
- Team expertise and availability
- Pricing model (fixed vs. hourly)
- References and case studies

Cloud Infrastructure Comparison

Focus criteria on:
- Performance benchmarks
- Geographic availability
- Cost modeling at different scales
- Managed service offerings