Research method
How the research classes and QA work
Relevance comes first. A company must meet the written targeting criteria; a project signal is an additional evidence requirement for one class, not the definition of every lead.
Our method keeps company fit, observed evidence, interpretation and uncertainty separate. It does not infer private buying intent.
Start with a research unit and a written profile.
- Define the brief. Agree the industries, territory, company criteria, exclusions and evidence requirements. Specify a company account or a facility record; do not count several sites as separate companies.
- Check source use. Identify sources that may be reviewed and what may be included in customer-facing output. Public access does not settle reuse rights.
- Resolve the entity. Check which company and site the evidence concerns; review parent/subsidiary relationships, duplicates and conflicts.
- Record the fit. Support each material criterion with a source reference and relevant dates. State why the account is included and what remains unresolved.
- Apply the class check. A project-signal record needs evidence of a relevant project or development as well as account fit.
- Review before release. Check the criteria, source posture and claims. A research pass is not, by itself, authority to deliver data or contact anyone.
Two classes, with project research status kept separate.
Industry-matched leads (Look Alike)
A rules-based match to the agreed profile. No current or upcoming project is required. Where project research was not part of the brief, say not assessed—not required. Where an agreed bounded search found none, say no qualifying signal identified in the bounded review. Do not turn either statement into “no project exists”.
Opportunity-signal leads
A profile-matched account with a qualifying project or development signal supported by dated evidence. State the actual stage and the unknowns; do not infer an open tender, unawarded package or purchasing contact.
Both classes require source-supported fit. Our 90% industry-matched / 10% opportunity-signal mix is a target, not a guarantee. Counts and class treatment are agreed in writing. Neither class is automatically a marketing-qualified lead.
Read the record in three layers.
- Observed fact
- What the source actually supports, attributed to that source and its date. An announcement of a proposed expansion is not evidence that construction or procurement has begun.
- Interpretation
- Why the fact may be relevant to the supplier’s range. An application hypothesis is not a verified technical match.
- Unknown
- Information not established: for example specifications, timing, procurement route or package availability. State the next useful check rather than filling the gap.
Keep publication dates distinct from retrieval or review dates. A recent review does not make an old source current. Reconcile conflicting sources before relying on a material claim.
A score cannot replace the fit decision.
The written criteria determine whether a record is accepted, held for review or rejected. Keep the decision reason visible. A score cannot rescue an excluded account or supply missing project evidence.
The detailed fictional opportunity brief includes an illustrative prioritisation score. It is not a buyer-intent score, a probability of sale or a quality ranking between the two lead classes.
Sources and data have boundaries.
Company-level research is the default described on this site. Named-contact fields and outreach require a separate agreed scope and applicable source, privacy, suppression and security controls.
Research may use approved public sources and, only where contracted and permitted, licensed sources. Evidence retained internally is not automatically available for customer copying. Unknown source or customer-use rights are a hold, not a permission.
No confidential client information belongs in a public example. The published samples are fictional and do not demonstrate live-data clearance, a software platform or client results.
Review, correct and learn within the agreed scope.
Review dates and triggers are agreed for the research, rather than presented as continuous monitoring. If a material fact is challenged, identify the record and claim, inspect the evidence, and record the correction or unresolved question.
Acceptance, replacement windows and any replacement limits are agreed in writing. A missing opportunity signal must not be silently replaced with an industry-matched record to make a batch look complete.
Commercial outcomes can inform later work only when suitable client feedback is available. No meeting, pipeline or conversion improvement is claimed here.