|
1
|
--- |
|
2
|
name: supplier-research |
|
3
|
description: Research, evaluate, and compare suppliers and vendors for B2B procurement |
|
4
|
--- |
|
5
|
|
|
6
|
# Supplier & Vendor Research |
|
7
|
|
|
8
|
Research, evaluate, and compare suppliers and vendors for B2B procurement. Build vendor shortlists, evaluation matrices, and RFP frameworks. |
|
9
|
|
|
10
|
## When to Use |
|
11
|
|
|
12
|
- User needs to find suppliers for a product, service, or component |
|
13
|
- User wants to evaluate and compare vendors |
|
14
|
- User needs an RFP or vendor evaluation framework |
|
15
|
- User asks about procurement best practices |
|
16
|
- User wants to assess supplier risk or negotiate better terms |
|
17
|
|
|
18
|
## When NOT to Use |
|
19
|
|
|
20
|
- Personal product shopping (use personal-shopper skill) |
|
21
|
- Competitive market analysis (use competitive-analysis skill) |
|
22
|
- Software/SaaS evaluation only (use deep-research skill) |
|
23
|
|
|
24
|
## Methodology |
|
25
|
|
|
26
|
### Step 1: Requirements Definition |
|
27
|
|
|
28
|
Before researching suppliers, clarify: |
|
29
|
|
|
30
|
**What you need:** |
|
31
|
|
|
32
|
- Product/service specification (be specific) |
|
33
|
- Volume/quantity requirements |
|
34
|
- Quality standards and certifications needed |
|
35
|
- Timeline and delivery requirements |
|
36
|
|
|
37
|
**Constraints:** |
|
38
|
|
|
39
|
- Budget range |
|
40
|
- Geographic preferences (domestic, nearshore, offshore) |
|
41
|
- Compliance requirements (ISO, SOC2, GDPR, industry-specific) |
|
42
|
- Minimum order quantities |
|
43
|
- Payment terms preferences |
|
44
|
|
|
45
|
### Step 2: Supplier Discovery |
|
46
|
|
|
47
|
Match the directory to the sourcing geography. Use `webSearch` + `webFetch` against these: |
|
48
|
|
|
49
|
| Platform | Coverage | Best for | Caveat | |
|
50
|
|----------|----------|----------|--------| |
|
51
|
| **Thomasnet** | 500k+ North American suppliers | US/Canada industrial — machinery, plastics, metals, custom components. Free for buyers. Filter by ISO certs + CAD availability. | US-only; no pricing shown | |
|
52
|
| **Alibaba** | 200k+ suppliers, 200M+ SKUs | China/Asia, prototype sampling, MOQ benchmarking across 5,900 categories | Many "manufacturers" are trading companies — verify with customs data | |
|
53
|
| **Global Sources** | Asia, audited | Electronics, consumer goods. Stronger supplier audits than Alibaba. | Smaller catalog | |
|
54
|
| **IndiaMART** | India | Textiles, chemicals, pharma intermediates, generics | Data quality varies widely | |
|
55
|
| **Kompass / Europages** | EU | EU-based sourcing when GDPR/CE compliance matters | Limited free tier | |
|
56
|
| **ImportYeti** (free) | US ocean freight records | **Verification, not discovery.** Look up a supplier to see real US customs shipment history — who they actually ship to, how often, what volume. Exposes trading companies posing as factories. | Sea freight only, US imports only | |
|
57
|
| **Panjiva / ImportGenius** | Global trade data | Competitor supply chain mapping — find out who your competitors buy from | Paid, learning curve | |
|
58
|
|
|
59
|
**Agent search patterns:** |
|
60
|
|
|
61
|
- `site:thomasnet.com "[product] manufacturer" [state]` — direct directory scrape |
|
62
|
- `site:alibaba.com "[product]" "verified supplier" "trade assurance"` — pre-filtered for badges |
|
63
|
- `webFetch: importyeti.com/company/[supplier-name]` — verify real export activity before engaging |
|
64
|
- `"[competitor product name]" "made in" OR "manufactured by"` — reverse-engineer competitor supply chains |
|
65
|
- `site:alibaba.com "[product]" MOQ` — quickly benchmark minimum order quantities across suppliers |
|
66
|
|
|
67
|
**2025 geography shift:** Vietnam, India, and Mexico are the primary China+1 alternatives. Mexico benefits from USMCA (no tariffs, 3-5 day freight vs 30+ from Asia). Vietnam is strong in furniture/electronics assembly. India is strong in textiles/pharma/software. |
|
68
|
|
|
69
|
Target: 8-12 candidates for RFI, narrow to 3-5 for RFQ. |
|
70
|
|
|
71
|
### Step 3: Vendor Evaluation Matrix |
|
72
|
|
|
73
|
Score each vendor across weighted criteria: |
|
74
|
|
|
75
|
| Category | Weight | Criteria | |
|
76
|
|----------|--------|----------| |
|
77
|
| **Quality** | 25% | Certifications, defect rates, QC processes, samples | |
|
78
|
| **Cost** | 20% | Unit price, total cost of ownership, volume discounts, hidden fees | |
|
79
|
| **Delivery** | 20% | Lead times, on-time delivery rate, shipping methods, inventory | |
|
80
|
| **Capability** | 15% | Production capacity, scalability, technology, R&D | |
|
81
|
| **Reliability** | 10% | Financial stability, years in business, references, insurance | |
|
82
|
| **Compliance** | 10% | Regulatory compliance, certifications, ESG practices, data security | |
|
83
|
|
|
84
|
**Scoring scale:** 1 (poor) to 5 (excellent) per criterion. |
|
85
|
|
|
86
|
**Total Cost of Ownership (TCO):** |
|
87
|
Don't just compare unit prices. Include: |
|
88
|
|
|
89
|
- Purchase price |
|
90
|
- Shipping and logistics |
|
91
|
- Import duties and taxes |
|
92
|
- Quality inspection costs |
|
93
|
- Inventory carrying costs |
|
94
|
- Switching costs |
|
95
|
- Risk costs (what if they fail to deliver?) |
|
96
|
|
|
97
|
### Step 4: Verification & Risk Assessment |
|
98
|
|
|
99
|
**Verify the factory is real (the #1 failure mode in overseas sourcing):** |
|
100
|
|
|
101
|
- **Customs data cross-check**: `webFetch` the supplier on ImportYeti — consistent monthly shipments to recognizable brands = real factory. Zero export history or shipments only to shell companies = trading company or fraud. |
|
102
|
- **Certificate verification**: Don't trust uploaded PDFs. ISO 9001 certs have a cert number — verify on the issuing body's site (SGS, BV, TÜV, Intertek all have public lookup tools). `webSearch: "[cert body] certificate verification [cert number]"`. |
|
103
|
- **Business license**: For China, request the Unified Social Credit Code (18 digits) — verifiable on the National Enterprise Credit system. For US, check state Secretary of State filings. |
|
104
|
- **Address verification**: `webSearch` the factory address — Google Maps satellite view should show an industrial facility, not a residential block or office tower. |
|
105
|
- **Alibaba badges**: "Verified Supplier" means a third party (SGS/BV) physically visited. "Gold Supplier" just means they paid a fee — it verifies nothing. |
|
106
|
|
|
107
|
**Risk dimensions:** |
|
108
|
|
|
109
|
| Risk type | Check | Red flags | |
|
110
|
|-----------|-------|-----------| |
|
111
|
| **Financial** | D&B report, years in business, customer concentration | <3 years operating, >40% revenue from one customer, requests 100% upfront payment | |
|
112
|
| **Operational** | Factory count, capacity utilization, QC process docs | Single facility, no in-house QC team, won't allow video factory tour | |
|
113
|
| **Geopolitical** | Tariff exposure (Section 301 for China), sanctions lists (OFAC SDN list), port stability | Sourcing region on UFLPA entity list, currency controls, single-port dependency | |
|
114
|
| **Compliance** | UFLPA (Xinjiang forced labor — US *presumes* guilt for flagged regions), EU CSDDD due-diligence rules (2024+), conflict minerals (3TG) | Can't provide tier-2 supplier list, cotton/polysilicon from Xinjiang, no chain-of-custody docs | |
|
115
|
| **Supply chain** | Tier-2 dependencies, raw material source, seasonal capacity (Chinese New Year = 4-6 wk shutdown) | Won't name their material suppliers, capacity claims exceed facility size | |
|
116
|
|
|
117
|
### Step 5: RFP/RFQ Process |
|
118
|
|
|
119
|
If conducting a formal selection: |
|
120
|
|
|
121
|
**RFP structure:** |
|
122
|
|
|
123
|
1. Company overview and project background |
|
124
|
2. Scope of work / product specifications |
|
125
|
3. Volume and timeline requirements |
|
126
|
4. Quality and compliance requirements |
|
127
|
5. Pricing format (line item breakdown) |
|
128
|
6. References (3+ similar clients) |
|
129
|
7. Evaluation criteria and weights |
|
130
|
8. Timeline for responses and decision |
|
131
|
|
|
132
|
**Evaluation process:** |
|
133
|
|
|
134
|
1. Distribute RFP to shortlisted vendors (3-5) |
|
135
|
2. Allow Q&A period |
|
136
|
3. Score responses against evaluation matrix |
|
137
|
4. Conduct reference checks for top 2-3 |
|
138
|
5. Request samples or pilot project |
|
139
|
6. Negotiate final terms with preferred vendor |
|
140
|
7. Award and onboard |
|
141
|
|
|
142
|
### Step 6: Negotiation Preparation |
|
143
|
|
|
144
|
**Levers ranked by typical yield:** |
|
145
|
|
|
146
|
1. **Volume commitment** — annual forecast (even non-binding) usually unlocks 8-15% off spot pricing |
|
147
|
2. **Payment terms** — overseas default is 30% deposit / 70% pre-shipment. Pushing to 30/70 *after* delivery, or Net 30 from shipment, is worth 2-5% of unit cost in cash flow |
|
148
|
3. **Incoterms** — know the difference: FOB (you pay freight + insurance from port), CIF (supplier pays to your port, but *you* bear risk in transit — worst of both), DDP (supplier handles everything including customs — most expensive, least risk). FOB is the standard for experienced buyers. |
|
149
|
4. **MOQ flexibility** — first-order MOQ is almost always negotiable down 30-50% if you frame it as a paid trial. "We'll pay the higher per-unit price on 500 units to validate, then commit to your 2,000 MOQ." |
|
150
|
5. **Tooling/mold ownership** — for custom parts, negotiate that *you* own the mold after paying for it. Otherwise you're locked in forever. |
|
151
|
|
|
152
|
**Contract terms that matter most:** |
|
153
|
|
|
154
|
- **Price escalation clause** — cap annual increases at a named index (e.g., PPI for the material category), not "supplier discretion" |
|
155
|
- **Quality SLA with teeth** — define AQL (Acceptable Quality Level — typically 2.5 for general goods, 1.0 for critical components), specify who pays for third-party inspection (QIMA, SGS), and define the remedy (rework at supplier cost, not just credit) |
|
156
|
- **Lead time + late penalties** — industry norm: 1-2% of order value per week late, capped at 10% |
|
157
|
- **IP protection** — NNN agreement (Non-disclosure, Non-use, Non-circumvention) for China, not a US-style NDA — US NDAs are unenforceable in Chinese courts |
|
158
|
- **Exit clause** — right to terminate with 60-90 days notice, obligation to complete in-flight orders, tooling transfer terms |
|
159
|
|
|
160
|
## Output Format |
|
161
|
|
|
162
|
Always present key findings and recommendations as a plaintext summary in chat, even when also generating files. The user should be able to understand the results without opening any files. |
|
163
|
|
|
164
|
```text |
|
165
|
|
|
166
|
# Vendor Evaluation: [Category] |
|
167
|
|
|
168
|
## Requirements Summary |
|
169
|
[Key specs, volume, timeline, constraints] |
|
170
|
|
|
171
|
## Shortlisted Vendors |
|
172
|
|
|
173
|
### 1. [Vendor Name] |
|
174
|
|
|
175
|
- Website: [url] |
|
176
|
- Location: [city, country] |
|
177
|
- Specialization: [what they do] |
|
178
|
- Key Strengths: [2-3 points] |
|
179
|
- Concerns: [1-2 points] |
|
180
|
- Estimated Cost: [range] |
|
181
|
|
|
182
|
### 2. [Vendor Name] |
|
183
|
... |
|
184
|
|
|
185
|
## Evaluation Matrix |
|
186
|
| Criteria (Weight) | Vendor A | Vendor B | Vendor C | |
|
187
|
|-------------------|----------|----------|----------| |
|
188
|
| Quality (25%) | 4/5 | 3/5 | 5/5 | |
|
189
|
| Cost (20%) | 5/5 | 4/5 | 3/5 | |
|
190
|
| ... | | | | |
|
191
|
| **Weighted Total** | **X.X** | **X.X** | **X.X** | |
|
192
|
|
|
193
|
## Recommendation |
|
194
|
[Top pick with reasoning, runner-up, and suggested next steps] |
|
195
|
|
|
196
|
``` |
|
197
|
|
|
198
|
## Best Practices |
|
199
|
|
|
200
|
1. **Never single-source critical components** — maintain a qualified backup at 10-20% of volume even if unit cost is higher |
|
201
|
2. **Sample → pilot → scale** — paid samples first, then a pilot run of 5-10% of target volume, then commit. Never skip to full MOQ. |
|
202
|
3. **Third-party inspection before final payment** — QIMA, SGS, or Bureau Veritas run pre-shipment inspections for ~$300. Cheaper than one bad container. |
|
203
|
4. **Back-channel references** — find their customers via ImportYeti shipment records and cold-email them. Supplier-provided references are curated. |
|
204
|
5. **Plan around Chinese New Year** — factories shut 4-6 weeks (late Jan/Feb). Orders placed in December ship in March. Build buffer inventory by November. |
|
205
|
6. **Landed cost, not unit cost** — a $2.00 unit from China can land at $3.50 after freight, 25% Section 301 tariff, duty, and inspection. A $2.80 unit from Mexico under USMCA might land at $3.10. |
|
206
|
|
|
207
|
## Limitations |
|
208
|
|
|
209
|
- Cannot access paid databases (Panjiva, D&B, ImportGenius full data) — ImportYeti free tier is the workaround for US import verification |
|
210
|
- Cannot physically inspect facilities or samples — always recommend third-party audit for orders >$10k |
|
211
|
- Cannot verify certificate authenticity directly — provides the lookup URLs for the user to check |
|
212
|
- Tariff rates and trade rules change frequently (Section 301, UFLPA scope) — verify current rates at time of order |
|
213
|
- Pricing from directory listings is indicative only — real quotes require RFQ with full specs |