B2B LEAD GENERATION GUIDE
Good lead generation starts with a specific market and ends with records a salesperson can assess. Here is the work between those two points.
Write down the industries, locations, company characteristics, buying situation and roles that matter. Add exclusions early: existing customers, competitors, suppressed domains and segments your team cannot serve.
Use company websites, directories, sector sources and databases as discovery inputs. Check the company itself rather than inferring fit from a single category label. Record the evidence that explains the match.
Choose a person whose role fits the decision and the offer. A correct company with an unrelated contact is not a qualified lead. Keep role, name and professional contact details tied to the same evidence trail.
Check work-email status, duplicate records, CRM conflicts and suppression rules. Mark uncertain cases for human review instead of hiding uncertainty in a confident-looking export.
B2B lead generation helps a business identify and engage potential business customers. Inbound channels attract people who may enquire; outbound research identifies people a team may choose to approach. Both need qualification, but they start with different evidence of interest. A researched contact becomes a sales opportunity only when the sales process establishes a real potential purchase.
| Channel | Useful starting point | What still needs checking |
|---|---|---|
| Organic search | A buyer searches for a problem, guide or service. | Whether the enquiry matches your market and can be served. |
| Referrals and partnerships | A relevant introduction provides context. | Permission for the introduction, need and ownership of follow-up. |
| Events and communities | A topic or relationship connects you. | What was actually discussed and what next step was agreed. |
| Paid acquisition | A defined offer reaches a selected audience. | Qualified-enquiry cost, attribution and conversion into sales. |
| Outbound research | A company and buyer fit a written brief. | Evidence, contact route, exclusions and appropriate outreach. |
“UK businesses that need more sales” is too broad to review consistently. Specify the offer, service area, company characteristics, buyer responsibilities, evidence required and exclusions. Decide what counts as unknown and who resolves borderline cases. Keep the original brief version with each delivered batch so a later targeting change does not rewrite the acceptance standard retrospectively.
Separate must-haves from preferences. If commercial installation work is mandatory, a domestic-only contractor fails even if its size and location are attractive. If employee count is only a preference, missing evidence should be visible rather than replaced with an estimate presented as fact.
Illustrative workflow, not a customer result. A supplier targets UK contractors that maintain industrial roofs. The agreed buyer is an Operations Director or equivalent. Researchers discover companies, inspect service pages, retain source URLs and check buyer roles. A residential specialist is rejected; a company with unclear service coverage is held for clarification.
For the remaining candidates, the team verifies contact details, checks duplicates and exclusions, and records why each match belongs. The reviewer approves the current record revision. Delivery rechecks suppression and places accepted records with an owner and a next action. A reply expressing a project need starts sales qualification; the earlier research did not establish that need.
Use software when your team has the skill and time to run the workflow. Commission research when you need an agreed output and someone else to carry out the checking. Consider an installed system when you want your team to own the operation and can maintain accounts, source access, exception handling and review. Managed outbound adds responsibility for approved sending, replies and routing; it should be assessed separately from research delivery.
Compare these choices against the same brief, accepted-record standard and total operating cost. The tools guide provides a decision matrix.
Track discovered records, accepted records, qualified enquiries, sales opportunities and sales as separate counts. Accepted yield is accepted records divided by reviewed records. Cost per accepted record includes research, data, verification and review costs divided by accepted records. Record the denominator and reporting period so percentages remain interpretable.
For inbound acquisition, retain source and landing page through enquiry, qualification, proposal and sale. For delivery, measure time to first review, unresolved exceptions and replacement requests. Report zero sales as zero; do not turn a large export count into a revenue claim.
AI can assist interpretation and drafting, but generated text is not source evidence. Keep hard requirements, duplicate rules, suppression and external actions under explicit controls. Use the AI workflow guide to separate assistance from acceptance.
Review the source, licence, intended use and appropriate handling of business contact data before activation. Document the actual data flow and how corrections and opt-outs are handled. The responsible-use page explains Keystone’s operating boundaries; it is not a substitute for reviewing your particular campaign.
No. A useful record, a relevant message, buyer interest and a booked meeting are separate outcomes.
Choose a bounded experiment based on how the target buyer finds and evaluates the offer. Record the result you expect to learn and how you will measure it.
Return the feasible yield and explain the constraint. Broaden the brief only by agreement; do not fill a quota with weaker matches.
APPLY KEYSTONE TO YOUR MARKET