What Is Lead Generation? Finding B2B Leads With Web Data

Published:

15 minute read

Acar Diveroli
Written by: Acar Diveroli
A field of company tiles with building icons and signal bars fading into depth; one blue tile lifted below a check badge

A company that makes commercial floor-cleaning machines has three sales reps and a trade-fair catalogue with 1,500 exhibitors. Some of those firms run warehouses, hotels or hospitals and could use a machine; most never will. Somewhere in the same catalogue, a logistics firm has just posted twenty jobs for a warehouse that opens in March. Finding that firm, and knowing why to call it this month rather than next year, is what lead generation is about.

This post covers the definition and its terms (MQL, SQL, inbound, outbound), the B2B process step by step, the public signals that suggest a company may be buying, data quality, the e-mail rules in the EU, the US and Türkiye, and where proxies fit.

What is lead generation?

Lead generation is the work of attracting or finding potential customers and gathering enough information to start a sales conversation with them. A lead is a person or company that has shown interest in what you sell, or fits the customer you want, and can be contacted.

In business-to-business (B2B) sales, a lead has two layers. The account is the company, with its industry, size and location; the contact is the person who evaluates or approves the purchase. The account decides whether a deal makes sense, the contact decides whether a conversation happens, and the law treats the two very differently.

Prospecting is a related term: the sales-side work of researching a specific list of accounts and contacting them one by one. Lead generation decides which accounts belong on that list.

What are MQLs and SQLs?

An MQL and an SQL are two stages of qualification: marketing judges that a lead is promising, then sales confirms that it is a real opportunity.

  • Marketing qualified lead (MQL): fits your target and has engaged enough to pass on, like a hotel group's facilities manager who downloaded a spec sheet and visited the pricing page twice.
  • Sales accepted lead (SAL): an optional middle step, where sales agrees to work the lead.
  • Sales qualified lead (SQL): sales has confirmed a need, a timeline and someone who can decide. On a call, the same manager says the group opens two hotels next spring and has a budget.

Teams decide what counts as qualified with a framework or a score. The BANT framework asks about budget, authority, need and timeline. Lead scoring gives points for fit (industry, size, country) and behaviour (pages visited, replies) and passes a lead to sales above a threshold. Agree on these definitions with sales in writing; otherwise marketing reports leads that sales does not recognise.

How does B2B lead generation work?

B2B lead generation works as a loop: describe the ideal customer, find companies that match, check whether they are likely to buy now, contact the right person lawfully and hand ready leads to sales. Six steps cover most programmes:

  1. Write the ideal customer profile. The ideal customer profile (ICP) describes the companies you serve best: industry, size, countries and the problem you solve. The cleaning-machine maker's ICP might read: warehouses over 10,000 m², hotel groups with five or more properties, hospitals in its delivery countries.
  2. Choose the sources. Start with your own data (past customers, website forms, event badge scans), then add company registries, procurement portals, company websites and licensed datasets.
  3. Build the account list. Collect companies, not people, at this stage. Remove duplicates and match each account to a stable key such as its web domain or registration number; What Is Data Enrichment? explains how missing fields such as industry and size get filled in.
  4. Qualify fit and timing. Score each account on how well it matches the ICP and on signals that it may be buying now, then sort the list into tiers: the top tier gets personal research, the rest a lighter touch.
  5. Contact lawfully. Find the role you need (head of facilities, operations director), choose a channel and check the rules for that channel and country before the first message.
  6. Hand over and learn. Pass each SQL to sales with its context (why this account, which signal), track how many leads become deals and adjust the ICP.

How do inbound and outbound lead generation differ?

Inbound lead generation waits for the buyer to come to you, outbound goes to the buyer first, and data-driven prospecting is the research layer that decides whom to approach and when.

Inbound covers search, content, webinars, referrals and forms: the buyer finds you and leaves details. Outbound covers cold e-mail, calls, trade fairs and ads aimed at a chosen list. Data-driven prospecting is not a channel of its own; it uses public company data to pick and time the outbound work and to decide which inbound leads to answer first.

InboundOutboundData-driven prospecting
First moveThe buyerYour teamYour team, at a chosen time
Typical sourcesSearch, content, formsLists, events, callsRegistries, job ads, tenders
Buying intentHighUnknownInferred from signals
FitMixedChosen by youChosen by you
Main costContent and timeRep timeResearch and upkeep
Legal focusConsent at the formRules for cold contactPersonal data in research

Which public signals show that a company may be buying?

Public signals at company level, such as new hiring, new sites, tenders and changes to visible technology, show that an account may need something now. They describe what a company does in public, not what a person does, which keeps the research on safe ground.

  • Job postings. A logistics firm hiring forklift drivers and shift leads in a new town is opening a warehouse, and the required skills often name the software and equipment a team uses. Job Posting Data covers the fields and sources.
  • New locations. Store, hotel, factory and office openings appear on a company's own news page and in the local press.
  • Public tenders. Public bodies publish what they plan to buy, with dates and specifications: in the EU on Tenders Electronic Daily (TED), in Türkiye on the EKAP platform of the Public Procurement Authority. A hospital tender for cleaning equipment is about as direct as a signal gets.
  • Technology on the website. The public page source often shows the shop platform or booking engine a company runs; a switch can mean a new project.
  • Registry news and launches. New registrations, capital increases and mergers appear in official registries and gazettes; a new product line or dealer map shows where a company is expanding.

Record each signal with its source link and date, and let it expire: a hiring wave from last year says little about this quarter. A signal is a reason to look closer, not proof of intent.

How do you keep lead data accurate?

You keep lead data accurate by matching records on stable keys, storing a source and a date for each field and removing everyone who asked not to be contacted before every send. Lead data ages quickly: people change jobs, companies rename and move, and shared inboxes stop working.

  • Deduplicate on keys, not names. A group and its subsidiary can share a name yet need separate records.
  • Keep a source and date per field. A role confirmed last month beats one copied from an old list.
  • Act on bounces. Remove a bounced address; do not guess a new one from a name pattern.
  • Keep a suppression list. Everyone who opted out or objected goes on it, and every new list is screened against it.
  • Measure conversion, not volume. The share of leads that become opportunities says more than the number of names.

What does the law say about B2B lead generation?

B2B lead generation is legal, but two sets of rules apply as soon as people are involved: data protection law for the personal data you hold and marketing law for the messages you send. Company facts such as a registered address or an industry code are generally not personal data; a named work address such as jane.doe@example.com is, because it identifies a person.

EU, data protection. Using a contact's personal data needs a lawful basis, for marketing usually consent or legitimate interest. Recital 47 of the GDPR says direct marketing "may be regarded as carried out for a legitimate interest", which still means weighing your interest against the person's. Under Article 21, anyone can object to direct marketing at any time, and you must then stop. If the data did not come from the person, Article 14 requires telling them where it came from within one month, or at your first message if that is sooner.

EU, e-mail. The ePrivacy Directive adds a rule for the message itself. Under its Article 13, marketing e-mail to natural persons needs prior consent, except that a business may e-mail its own customers about its own similar products if they could object when giving their details and can object in every message. For companies as recipients, the directive leaves protection to each member state, so the rule for a cold e-mail to a business address depends on the recipient's country. The sender must always be identifiable and give a valid opt-out address.

United States. The CAN-SPAM Act makes no exception for B2B e-mail, but it does not require prior consent. It requires accurate headers and subject lines, a clear statement that the message is an advertisement, a valid postal address and a working opt-out; the FTC's compliance guide says opt-out requests must be honoured within 10 business days.

Türkiye, commercial messages. Law No. 6563 on the Regulation of Electronic Commerce (in Turkish) allows commercial electronic messages only with the recipient's prior consent, except for messages to tradespeople and merchants. Any recipient can refuse at any time, and the sender must stop within three working days. Under the implementing regulation, the sender registers business recipients' addresses in IYS, the national message management system, and checks there whether they have refused.

Türkiye, personal data. Under the data protection law (KVKK, Law No. 6698), data a person made public can be processed without consent, and the Personal Data Protection Board reads this narrowly: only for the purpose for which it was made public. In decision 2022/861 (in Turkish), the Board fined a company 150,000 TL for taking a lawyer's work e-mail from search results and sending promotional e-mails: the address was public for legal work, not for marketing, and lawyers cannot act as merchants, so the 6563 exception did not apply.

The rule for every market: "publicly visible" does not mean free to use. Keep personal fields to the role you need, record where each came from and drop the rest; Personal Data in Scraped Datasets shows how to cut them down and protect what remains.

Where do proxies fit in lead generation?

Proxies help with one narrow part of lead research: seeing public company pages the way they appear in each target country.

Company websites, job boards and local directories often show each country a different version: a careers page lists only local vacancies, a store finder shows nearby branches. Checked from one office, every page returns that office's version. A Residential Proxy lets a researcher load the page through a home connection in the chosen country or city, at a pace the site can easily handle.

Proxies do not change what you may collect or whom you may contact. Logged-in platforms and professional networks have their own terms and sales tools; use those. Teams that need public company pages collected at scale can use managed page extraction through our Web Scraping API, available through our sales team.

Where is lead generation used?

  • B2B software: finding companies of the right size and sector as they outgrow their current tool.
  • Manufacturers and distributors: finding dealers and large buyers in a new export market.
  • Agencies and consultancies: spotting companies that launch products or rebuild their websites.
  • Recruitment firms: identifying companies that hire for the roles they place.
  • Facility and logistics suppliers: following new sites, moves and expansions.

Common mistakes

  • Buying a list without asking how it was built. You inherit the seller's methods and errors.
  • Collecting personal e-mails and phone numbers from websites. That is personal data, and often an unlawful start to a campaign.
  • Ignoring objections. Without a suppression list, the same person is contacted again from the next list.
  • No shared definition of a qualified lead. Marketing and sales argue about numbers instead of fixing the funnel.
  • Contacting without a reason. A message tied to a real signal gets read; a generic pitch does not.
  • Researching every market from one country. Local pages and job listings come out wrong for the others.

Decision guide

NeedRecommendation
A few large target accountsResearch each by hand; follow its signals
A steady flow of interested buyersInbound content; a clear form with recorded consent
Entering a new countryCheck local e-mail rules first; local-language pages
Knowing when to callTrack hiring, new sites and tenders per account
Buying a contact listAsk for source, consent and date; screen it
E-mailing individuals in the EUConsent or the existing-customer exception
E-mailing businesses in TürkiyeMerchant exception only for merchants; check IYS
Thousands of company pages a monthManaged extraction of company-level fields

Frequently asked questions

What does lead generation mean in sales?

In sales, lead generation means filling the pipeline with potential customers who fit your target and can be contacted. Marketing usually brings in and scores the leads, and sales takes the qualified ones into conversations and deals.

What is the difference between lead generation and demand generation?

Demand generation creates interest in a market or a problem; lead generation turns that interest into identifiable prospects. A webinar on warehouse hygiene is demand generation; its sign-up form, with consent recorded, is lead generation.

What is a qualified lead?

A qualified lead fits your ideal customer profile and has shown a real reason to buy. Marketing qualifies by fit and engagement (an MQL); sales qualifies in conversation by confirming need, budget, authority and timing (an SQL).

It can be. The GDPR needs a lawful basis for the contact's data, often legitimate interest with a documented balancing test, and each member state's e-mail rules decide whether a message to a business address needs consent. Always identify yourself and offer an easy opt-out.

Buying company data is generally fine; buying personal contact data moves the legal risk to you. Ask the seller where each record came from, which consent or legal basis covers your use and when it was last checked, then screen the list against your suppression list. Without those answers, do not use it for e-mail.

How is AI used in lead generation?

AI tools score leads, summarise company news, draft first messages and merge duplicate records. They do not change the rules: an AI agent that gathers personal e-mails is still collecting personal data, and a person should check its output before anything is sent.

Summary

Lead generation finds the companies that fit your offer, notices when they are likely to buy and starts a conversation with the right person. In B2B it runs as a loop from the ideal customer profile to qualification and hand-off, and it works best on company-level public signals such as hiring, new sites and tenders. Personal contact data brings in data protection and e-mail law in every market, so collect less, record your sources and honour every objection. For checking public company pages in each target market, our market research proxies are built for that kind of research.

Ask ChatGPTAsk Claude