Once you have a list of suppliers, the next step is qualifying them. Traditionally that means a procurement team visiting each supplier’s website, checking certifications one by one, searching news archives, pulling financial data from separate databases, and chasing contact details. For a list of 50 suppliers, that’s days of work, sometimes weeks.
Speya automates that research. You describe what you want to know, and Speya gathers it across all suppliers in your list in parallel, with sources cited for every cell. Enrichments help you evaluate which of those candidates to actually engage with.
Your list comes together in a discovery search, where requirements decide which suppliers make it in. From there, two things add research data to the suppliers already in your list: enrichments, where Speya researches your questions across public data, and registries, which pull verified data straight from an external authority. This article covers mainly enrichments, describing how you research suppliers to decide which candidates to engage with.
In this article
Two ways to add enrichments
Standard columns with + Research
Step-by-step: Enrich Suppliers with Speya
When is your enrichment done?
Best practice tips
Prompt library
FAQ
Two ways to add enrichments
You can add enrichments in two ways:
The + Research button: standard columns covering the most common data points sourcing teams research. A good starting point.
Write with Speya: the way to go when you want to customize the research to your case and your evaluation.
Both produce the same outcome: a column that researches every supplier in your list, with a source behind every result.
1. Standard columns
Click + Research in the top-right corner above the list, then click Enrich full list. Pick the enrichment you want to research and click + Add to queue. You can select up to 3 enrichments at a time.
The standard columns cover the most common things sourcing teams research about a supplier. This is also where you save your own enrichments as Custom templates, this way you can reuse them in other search projects and share them with your team.
Company profile
Find Contacts: email contacts for each supplier in the list.
HQ Location: the city and country of the company’s primary headquarters.
Website URL: the company’s canonical homepage address.
Year Founded: the founding year.
Employees: the latest total employee figure.
Annual Revenue: the latest annual revenue, drawing on filings, registries, news, and published estimates, with an estimated range when no figure is published.
Ownership: classified as Public, Private, State-owned, Subsidiary, or Joint-venture.
Ticker: the exchange-qualified ticker for publicly listed companies.
Production Facilities: the locations of the company’s production sites, including subsidiary and joint-venture plants.
Quality and compliance
Certifications: current third-party certifications, such as ISO 9001, ISO 14001, IATF 16949, and BSCI.
Quality Standards: additional quality systems and customer-specific standards, such as AS9100, VDA 6.3, GMP, and HACCP.
FDA Approval: FDA approvals, clearances, or facility registrations covering the company’s products.
Offering and markets
Core Products: the main product lines or service categories.
Key Industries: up to three primary customer industries.
Sustainability and risk
ESG Rating: a level from Excellent to Not Rated, based on third-party ratings and the company’s sustainability reporting.
ESG Controversies: significant environmental, labour, or governance incidents from the last three years, reported only with a source.
Financial Health: overall financial stability, rated Excellent, Good, Moderate, or Weak.
Supplier Risk Estimate: overall supplier risk classified as Low, Medium, High, or Critical, with a one-sentence rationale.
Operations
Production Scale: the company’s production scale from public signals such as stated capacity, plants, and production lines.
Export Markets: up to five leading export destination countries or regions.
Key Customers: up to five marquee customers with a public reference.
Add verified data with registries
Alongside the standard columns and anything you research with Speya, you can add registries. A registry is a direct connection to an external authority that holds verified data, so instead of Speya reasoning out an answer from across the web, the value comes straight from the organisation that maintains it. Registries are the way to confirm the facts that need to be auditable, such as a supplier’s legal identity, its verified climate targets, or its sanctions status.
Click + Research in the top-right corner above the list, then click Registries. Pick the registry you want and click + Add registries. We will add more sources soon, and if there’s a registry you need, let us know what you’d like us to add. Today you can add:
GLEIF / LEI: confirms a supplier is a real, uniquely registered legal entity, with its official identifier, registration, status, and home jurisdiction.
SBTi: confirms whether a supplier has climate targets validated by the Science Based Targets initiative, across near-term, net-zero, and long-term targets.
UN sanctions: confirms whether a supplier appears on the UN Security Council’s consolidated sanctions list, and under which regime.
2. Step-by-step: Enrich Suppliers with Speya
Writing to Speya works better when you want to customize the research, or you already know the enrichments you need. Both ways produce the same outcome; the chat gives you more power to reason about the columns, and Speya helps formulate them based on your case.
1. Type to Speya what you want to evaluate. Write it the way you’d brief a colleague: one data point per column, and include the answer format you want back, such as a short sentence, likely/unlikely, or low/medium/high.
“Add text columns for: [certification], [years in business], [production capacity], [contact details].”
Creates one column per data point.“Research if they have production sites in [country].”
Anchors the research to what the answer depends on, here where they manufacture rather than where they sell.
2. Not sure what to enrich? Ask Speya. It knows your case from the conversation and the list, so let it propose the data points:
“What other data points should we consider in this case?”
Surfaces research you haven’t thought of.“Do a risk analysis. Come up with the data points most important for evaluating risk on these suppliers.”
Makes Speya design and run the risk research for you: bad press, labour safety, sanctions, financial stability.
3. Read your results.
Hover over a column header to see the instruction that column runs for every supplier, so you always know what a result is actually answering.
Click a result to read Speya’s reasoning and open the exact sources it used. The instruction is included there too.
“No data found” means no result is available. Speya found no data that addresses this for that supplier.
4. Type to Speya to refine. This step is optional: based on what you read in the cells and what your evaluation criteria need, you can adjust the research or ask Speya to add more detail and dig deeper.
Reword the column with a sharper instruction, such as “Find the strongest public evidence that this company [does X]. Return a short sentence.”
Add a reasoning column when the data is thin: “I know some of these companies offer [this]. Reason about which ones likely do.” Speya weighs everything it knows about each supplier instead of requiring a published statement, which rescues plausible “No data found” rows.
5. Fill the gaps before you shortlist. Before you ask for a shortlist, Speya needs the full picture: every supplier evaluated on the same information. The easiest way to get there is Research missing, which runs a column’s empty cells only. Here’s how:
Hover over a column header and click the three vertical dots (…).
Click Research missing.
To re-run a single result, click Run cell.
6. Ask Speya to shortlist and tier the best suppliers. Once your columns hold data, you don’t have to read the table row by row. Ask Speya to shortlist or tier, and it reasons across every cell and column at once:
“I’m a [role] at [company] sourcing [product]. Which suppliers are most suitable? Give me the top-tier with reasoning.”
Sorts the whole list into Tier 1 to Tier 3.“Write an executive summary of the top 5 suppliers for this project.”
Returns a summary in the chat outlining what makes each supplier a good fit.
When you ask for a shortlist or tiers, Speya writes a summary of the top suppliers in the chat and sorts the whole list into tiers, adding a Tier column that ranks each supplier from Tier 1 to Tier 3.
Read the reasoning by clicking the info icon in the Tier column, and hover over the column header to see the instructions Speya used to decide. Speya sets those evaluation criteria automatically from your past case inputs, so to change them, tell it what matters: “Evaluate the top suppliers based on [dimensions].” You can be specific about the values or base requirements a supplier has to meet. Tiers are live, so they update as you add columns or research, or ask Speya to re-tier.
Speya reasons across the whole table. It has access to every cell and every column, so it reasons across the whole table, not just one row at a time. The more research your list holds, the better its shortlists, comparisons, and summaries become.
When is your enrichment done?
Enrichment is never 100% complete, and it doesn’t need to be. What you’re aiming for is enough comparable, source-backed data to separate a shortlist from the rest, with the remaining unknowns turned into questions. You’re there when most of these are true:
Your columns cover the few factors that actually decide your case, not every attribute you could think of.
Each column returns consistent and comparable data, so suppliers can be evaluated at a glance.
You’ve identified the suppliers with the most promising results, opened their sources to read more, and verified the results that came back from the research.
What’s still unknown is private information, such as prices, stock, and exact specifications. That’s expected: no public source can settle it.
Those unknowns are your RFI questions. The next step is to send them to your shortlist and let the answers land back in the list.
Next: Sending an RFI to suppliers from your list. (How to send RFIs with Engage)
Best practice tips
One data point per enrichment column. Asking Speya to research and reason about several things at once can impact performance, so focus each column on one thing. Separate columns come back as answers you can compare and check, one click to the source each. You can still create and load 3 to 4 different columns at the same time.
Benchmark against a supplier you know. “Compare each company to [current supplier], who is high quality but expensive. Is it likely or unlikely they are more cost-efficient? Answer likely or unlikely, with reasoning.” Anchoring the research to a supplier you know turns every answer into something you can calibrate.
Ask for verdicts, not scores. The model is not deterministic: on each pass it reasons from the data it finds to get the most likely answer, so an exact number like 73/100 can land differently between runs. Coarse verdicts such as likely/unlikely or low/medium/high absorb that variation, stay stable, and are much easier to compare across a list.
Pair a strict column with a reasoning column. Run one column that confirms a data point from a source or reasons about it, and a second one that reasons about the suppliers where nothing is published. You get the verified facts and the informed judgement side by side, and strong candidates with thin websites stay visible.
Research the drivers of cost, not the price. Real prices live in inboxes, not on websites. Ask Speya what drives cost for this product, such as location, energy, labour, scale, and process, and research those. If a column returns a number, open the source: a listed marketplace price is data, an estimated price is reasoning. For a number you can act on, send an RFI.
Let Speya design the research when the case is complex. Describe your challenge, your goal, and what you want to avoid, and ask Speya to formulate the enrichments. Because it builds them from your case context, the columns gather more specific data and label the output to support your evaluation. This is also the fix when columns come back too generic or “No data found”. Frame this as the challenge and ask Speya to suggest a better approach that gets as close as possible to what you need to evaluate accurately.
Columns persist. Keep working in the same list. New searches add suppliers to the same list and every research column stays. Use Research missing to fill the new supplier rows, this way your comparison keeps growing.
Prompt library
Every prompt here is something you type to Speya. They’re grouped by where you are in your research, with a description of what each does.
Decide what to research
“What other data points should we consider in this case?”
Lets Speya suggest research you haven’t thought of, based on the case.“Do a risk analysis. Come up with the data points most important for evaluating risk on these suppliers.”
Speya designs and runs the risk research for you: bad press, labour safety, sanctions, financial stability.
Design the research with Speya
“Here’s my challenge: [describe the problem]. I want to achieve [goal], and I want to avoid [what you don’t want]. Look at the information in the list and help me formulate the enrichments to best achieve this.”
Reason with Speya before enriching. Instead of you writing the enrichment, Speya designs it from your case context. The research it builds gathers more specific data from the suppliers and labels the output to support your evaluation.
Create enrichment columns
“Add text columns for: [certification], [years in business], [production capacity], [contact details].”
Researches several data points in one go, one column per data point.“Find the strongest public evidence that this company [does X]. Return a short sentence.”
Returns a concise, checkable answer per supplier instead of a long text to read.“Add a column to research bad press.”
Surfaces negative news you’d otherwise only find by reading the press yourself.“Research if they have production sites in [country].”
Separates where a supplier manufactures from where it sells or delivers.“Add a research column for contact details: sales email first, general email if none.”
Finds the right inbox for each supplier. An email found this way lands as text in the column; before sending, add it as the supplier’s contact and set it as primary (More: How to send RFIs with Engage).“Compare each company to [current supplier], who is high quality but expensive. Is it likely or unlikely they are more cost-efficient? Answer likely or unlikely, with reasoning.”
Benchmarks the whole list against a supplier you know. The likely/unlikely format gives consistent, comparable answers.“I know some of these companies offer [this]. Reason about which ones likely do.”
Soft re-evaluation that rescues plausible “No data found” rows where the supplier just doesn’t publish the detail.
Read and shortlist
“Read the results of the research and flag the ones that [produce X].”
Makes Speya read the finished columns for you and mark the rows that matter.“I’m a [role] at [company] sourcing [product]. Which suppliers are most suitable? Give me the top-tier with reasoning.”
Ranks across every result in the list and returns a shortlist tailored to your role and case.“Give me the top-tier alternatives.”
Quick shortlist from everything researched so far. Sorts the whole list into Tier 1 to Tier 3 and adds a Tier column, with the reasoning readable on each cell.“Write an executive summary of the top 5 suppliers for this project.”
Produces a summary you can share with colleagues who haven’t seen the list.“Find alternatives to [benchmark supplier] within this list.”
Compares against a named reference inside the current list, without starting a new search.“Evaluate the top suppliers based on [dimension].”
Sets the criteria Speya uses to tier, replacing the ones it chose automatically. Be specific about the values or base requirements a supplier must meet.
FAQ
What’s the difference between requirements and research columns?
They do different jobs. Requirements belong to a search: they decide which suppliers enter your list, and a verified requirement shows as a match. Research columns work on the suppliers already in your list: they answer the questions you choose, in the format you choose, and you can ask them to reason about a supplier even when nothing is published. Keep the requirements that define the category in the search, and check everything else as a column afterwards.
Can Speya find information that isn’t public?
No, and by design. Research columns read what suppliers leave as a public trace across 250+ data sources, so sensitive information such as stock levels, contracts, and internal audits is out of reach because it isn’t published anywhere. The exception is prices: in some cases Speya finds a price from a retailer or marketplace listing. That’s the list price before negotiations or bonus structures, not your purchase cost. For the real answers, ask with an RFI or RFQ instead, and the supplier’s reply lands back in your list.
How is a registry different from a research column?
A registry returns a verified value straight from the authority that maintains it, so it’s the source of truth for auditable facts such as legal identity, climate targets, or sanctions status. A research column reasons across public sources to answer the question you set, which is far more flexible and covers much more ground, but each run reads the sources fresh so wording can vary. Use a registry when you need an auditable fact, and a research column for everything you want to evaluate and compare.
Can Speya research prices?
It can research what’s published, which is usually not a real price. A listed marketplace or retailer price is genuine data, but it’s the list price before negotiations or bonus structures; an estimated price is reasoning, so open the source before relying on any number. The dependable pattern is to research the drivers of cost, then get actual numbers through an RFQ.
Why does the result say “No data found”?
It means no result is available: Speya found no data that addresses this for that supplier. Use Run cell or Research missing to try again, or reframe the column and ask Speya to suggest a better approach.
Why did the same research give a slightly different answer when I re-ran it?
Because each run reads the sources fresh rather than replaying a saved answer, wording and borderline judgements can vary between runs. Coarse verdicts such as likely/unlikely stay far more stable than numeric scores, which is one reason to prefer them. If a result your decision leans on looks different from what you expected, open the source and read the reasoning.
Can Speya score suppliers on a scale?
Yes, but verdicts compare better. A 0 to 100 score suggests precision the public data can’t support, and suppliers with unverified data can end up looking mid-table instead of unknown. If you want a ranked view, ask for tiers with reasoning, and treat “unverified” as its own category.
Do my columns carry over when I search again?
Yes. New searches add suppliers to the same list and every column stays. New rows arrive with empty cells for your existing columns; use Research missing to fill them in one go.
Can I reuse the same columns in another list?
Yes. Save them as Custom templates from the + Research menu and apply them in other projects. They’re shared with your team. (More: How to manage your workspace)
How is this different from researching each supplier in ChatGPT or Google?
Scale and traceability. One instruction researches every supplier in your list at once, every answer links to the source it came from, and when Speya can’t confirm something it says so instead of guessing. You get a comparable table built for a sourcing decision, not a stack of answers to fact-check one by one.