A water utility bought replacement pumps last year. Your company makes the seals inside those pumps. Who belongs in your prospect list? The utility that owns the plant, the contractor that installed the machinery, and the workshop that replaces worn components have different jobs. An award notice can give you a starting point for sorting them out. The name printed most prominently may be the least useful place to send a catalogue.
The EU Publications Office has announced a hybrid TED-together meeting in Luxembourg and online on 24–25 November 2026. Its announced programme includes notice validation, proposed website improvements and updates on the TED Open Data Service. For exporters researching public-sector customers, the immediate opportunity is to bring a real search problem to the event. The announcement establishes an agenda. It does not establish that a proposed feature has shipped or that purchasing rules have changed. Read the organiser’s announcement.
Start with the sale you could actually fulfil
Write down one deliverable, one buyer role and one geographic constraint before opening the search box. A manufacturer of corrosion-resistant seal kits might serve repair contractors across western Europe by parcel. The same manufacturer might offer on-site commissioning only within driving distance of its workshop. Those propositions lead to different searches, even though the catalogue photograph looks identical.
Give the researcher a question that can produce a decision. “Find water projects” invites an impressive spreadsheet. “Which contractors maintain rotating equipment at municipal treatment plants within reach of our overnight spare-parts service?” makes every result contestable. A design consultancy that drew the new facility may know its contractors. A civil engineering consortium that poured the concrete may have finished its involvement before the first bearing failed.
Agree on exclusions as well. Domestic garden pumps, swimming pools and construction dewatering can appear alongside wastewater equipment. A laboratory buying a benchtop peristaltic pump orders in smaller quantities and expects different documentation. Keep a small rejection log. A colleague starting the next territory should understand why a familiar manufacturer or apparently attractive contract disappeared from the list.
Search broadly enough to learn the buyer’s vocabulary
TED supports searches using buyer names, procurement classifications and notice text. Its help page shows how to apply Common Procurement Vocabulary codes and how the latest-version option affects results. Begin with the interface and save the expression you actually used. Someone reviewing your work in a fortnight needs to reproduce the sample. Consult TED’s search and browsing guide.
Try three passes. First search a plain product description, including the local term a maintenance engineer might use. Next examine a promising notice for its classification and search neighbouring categories. Finally search an organisation named in the results to discover how it describes similar work elsewhere. A bundle filed under mechanical maintenance can contain a substantial pump overhaul. A narrow equipment code can miss the contractor entirely.
Look at ten or twelve notices before collecting hundreds. Mark each as potentially relevant, ambiguous or outside scope. If most are ambiguous, inspect the lots and the place of performance. If most are plainly wrong, change one search variable at a time. Dropping a single adjective teaches you more about coverage than simultaneously widening geography, switching language and replacing the classification.
Follow one procedure through its documents
Treat a procedure as a folder containing a sequence of events. The original competition describes the intention to buy. A published result may identify a winner for a particular lot. A correction can alter a value or spelling. Open linked versions and note which document supports each field you record. A result for electrical panels gives little support to a claim about the supplier of mechanical seals.
Consider this fictional example. Northbank Water groups a pumping-station refurbishment into three lots. Alder Mechanical wins the rotating-equipment package. Rowan Controls supplies instrumentation. A notice lists a central purchasing agency as the formal buyer, while the operating site belongs to Northbank. Your component business now has a reason to investigate Alder’s aftermarket team. You have no evidence that Rowan buys seals, that Northbank specifies your alloy, or that the purchasing agency approves substitute parts.
A missing award amount deserves an empty cell and a note. Copying an estimated framework ceiling into a field called annual spend creates false precision. Multiple winners can share a framework, individual orders can vary and the notice may describe several years. Preserve the original currency, period and wording. If a manager needs an addressable revenue estimate, build it separately from quantities, compatibility and a documented assumption about replacement frequency.
Keep a short evidence sheet that survives a handover
Give each candidate a row containing the legal organisation name, role, relevant lot, location, notice publication number, source link and review date. Add a sentence explaining the fit in ordinary language. “Won maintenance of pumps used at three eastern district treatment sites” is more informative than a score of eighty. Set aside a second sentence for the largest unresolved question.
Resolve identities before enriching contacts. Compare the registered town, website domain and business description. A subsidiary may trade under its parent’s brand while operating a separate service desk. A joint venture might exist only for the contract. Similar initials can refer to unrelated firms on opposite sides of a border. Preserve accents and legal suffixes in your evidence even if your CRM displays a shorter trading name.
Assign every inference an owner. A sales engineer can judge whether the stated fluid and temperature make the product plausible. A logistics colleague can check overnight delivery to an island or mountain district. An account manager can identify an existing distribution agreement that affects who should approach the contractor. Each can say exactly what would change the decision. That makes the sheet usable when the original researcher goes on leave.
Turn a contract result into a sensible next question
Read the winning contractor’s own service pages. Does it refurbish pumps in a workshop, sell complete units, or subcontract maintenance? Look for the difference between a photograph of a machine and a description of work performed. A company advertising vibration analysis may diagnose a failure but send the rotating assembly elsewhere. Its purchasing role remains uncertain until you can explain who orders the replacement.
For Alder, an opening message might read, “Your team is listed for the mechanical package at Northbank. We supply short-run seal kits for abrasive wastewater duties. Does your workshop select replacement assemblies, or does the plant specify them centrally?” This gives the recipient a reason for the question and an easy correction. Avoid mentioning a supposed failure or a budget you inferred from the award. The aim of the first exchange is to find the responsible function.
Choose the communication route to suit the stage. During a live competition, use the buyer’s designated clarification channel for tender questions and respect the procedure described in its documents. After an award, a technical conversation with a contractor can explore its general service needs. Keep confidential specifications out of a first email. A link to a public material compatibility sheet can establish relevance without asking anyone to disclose another supplier’s drawing.
Work backwards from an unpromising result
Imagine a second search returns a university campus renovation. The headline contains “pump station,” but the lot description covers landscaping around a drainage basin. Narrowing to wastewater treatment would remove it. Before doing so, check whether your team also supplies drainage contractors. A result can be wrong for the present campaign and valuable to another business line. Route it deliberately or close it with a reason.
A third result concerns an operator in the right region whose contract includes twenty-four-hour emergency attendance. Your workshop can deliver components quickly but has no local technicians. Investigate the service provider responsible for overnight callouts. Perhaps a stocking agreement with that provider would meet the practical need. If no suitably equipped partner exists, park the account. Geographic proximity alone cannot compensate for a missing capability.
Notice what these examples change in your research brief. You now need records of service coverage, repair responsibility and purchasing discretion. A larger download of names will supply none of them automatically. Expandia’s market research service describes how to start from a commercial question. Bring that question and a few rejected examples when asking someone else to extend your search.
Test the shortlist before automating collection
Take a handful of accepted and rejected records to the people who will use them. Ask a field salesperson to choose the first call and explain why. Have an engineer flag unsupported compatibility claims. Ask a colleague unfamiliar with the project to find the underlying notice from the saved link. If they cannot repeat the reasoning in two minutes, improve your evidence sheet before adding another country.
Track outcomes that distinguish coverage from usefulness. Count how many organisations have a confirmed buying role, how many belong to an existing customer group and how many need external verification. Record the reason a candidate moved from research to outreach. A list can shrink substantially while becoming more valuable. Five verified maintenance workshops can support a territory visit that forty anonymous department names cannot.
If the manual exercise works, the TED Search API gives developers access to published notices for analysis and reuse. Hand over your tested expression, examples, identifiers and rejection reasons. Ask the developer to preserve notice versions, duplicate detection and links back to original records. Compare the automated output with your manual sample before relying on a scheduled export. Read the official Search API documentation.
Bring a reproducible question to Luxembourg or online
The meeting offers a timely occasion to investigate a problem that actually affects your pipeline. Save a public notice where you could identify a winning organisation but struggled to connect it to the correctly described lot. Write down the interface language, search terms, filter settings and sequence of clicks. Remove internal account scores and personal contact information. Other attendees can then reproduce the behaviour using public material.
Phrase your question precisely. “How should we follow a replacement notice when a translated company name changes?” invites a more concrete discussion than asking for better data. If someone describes a planned improvement, capture its status, supporting documentation and intended availability. Validate it in the published interface before changing instructions for your team. Give one person ownership of checking the answer after the event.
Decide whether a market deserves a field visit
Suppose the shortlist includes a regional operator, two engineering contractors, a specialist overhaul shop and a multinational framework holder. Plot the organisations against a realistic travel day. A map centred on registered headquarters might send your representative to five administrative buildings. Instead, verify workshop addresses, visitor arrangements and the presence of someone who understands the equipment. Procurement geography and operating geography often diverge.
Build an itinerary around decisions. At the overhaul shop, ask which components technicians inspect during a rebuild and who signs off substitutions. At the contractor, ask whether emergency spares are held centrally or purchased by individual depots. At the utility, seek publicly available supplier guidance. Write a useful outcome beside every appointment. A referral to the correct depot or a clear disqualification can justify the journey even without a quotation.
Bring a modest evidence pack: a dimensional sketch, examples of compatible working conditions, lead times for standard versus custom items, and the limits of your warranty. Leave the generic corporate presentation for anyone who requests it. A technician examining a worn sleeve can show precisely where a proposed replacement would bind. An administrator reviewing a brochure cannot settle that question. Capture what you learned, who can verify it and whether it changes the commercial fit.
Budget research time by uncertainty
Separate cheap desk checks from expensive enquiries. An address mismatch might take minutes to resolve on an organisation’s website. Verifying that a contractor can machine a particular surface finish might require a specialist conversation. Reserve those conversations for accounts where a positive answer would lead to a concrete next step. Running every prospect through the most elaborate qualification process delays promising opportunities and exhausts technical colleagues.
A simple stopping rule helps. Close a record when the evidence contradicts an essential requirement, assign a follow-up when one answer could settle a material doubt, and promote it when buying responsibility and product fit are sufficiently clear for a relevant approach. Put a review date on deferred candidates. A change of contractor, a new depot or an updated catalogue can alter the picture. Endless research without a decision consumes time just as surely as indiscriminate outreach.
At the end of the week hand over three things: the saved search, a small shortlist with evidence and the outstanding question beside each name. A colleague should be able to explain why Alder Mechanical belongs on the list, why Rowan Controls was excluded and what information could reverse either decision. That is a practical test of whether public procurement research has earned a place in your sales routine.
