Lead generation: what it is and how it works for a small business

Someone has sent you a spreadsheet with 300 “leads”. You call the first ten: several numbers go unanswered, one person never submitted a request, and another was looking for work your company does not do. The situation is hypothetical, but the question is practical: what exactly were you sold? Lead generation is the process of finding people or companies with a relevant need and a real reason to start a conversation with you. A contact list can be one input to that work. It becomes an outcome only when you know who these people are, why it makes sense for them to get in touch, and what happens after the first contact.

If your company sells to other companies, the mistake is expensive. A sales manager can spend a week on an unchecked cold list while a genuine enquiry sits unanswered in the inbox. Before spending on ads, a database or a new tool, agree on one word: “lead”.

In everyday use, a lead can mean anyone who has left their contact details. That definition is too broad for a sales team. A contact tells you how to reach someone. A lead adds a reason: the person or company may need your service, fits the conditions you can serve, and has given a signal that the conversation is worth continuing.

That signal can take different forms. Someone may ask for a quote. Someone else may ask whether you can deliver a project in a particular region. In B2B, a person might reply to a tailored message and describe a task their team needs to solve. Even then, it is not yet a sale. You still need to clarify the need, budget, timing and decision-maker.

It is easier to work with a few clear stages than to argue over whether every name “counts”:

Each business sets its own stage definitions. If a manager counts only requests for a proposal as leads while the ad platform counts every click on “Contact us”, the reports will not match. Write down definitions that marketing, sales and the owner can apply in the same way.

The question “what are leads?” often gets the short answer “potential customers”. That is fine as a quick definition. In sales, you also need to know why a particular contact might become a customer. Someone can follow your page for years without ever needing your service. A company on a list has not necessarily asked for anything either.

PILLAR sells solar carports for large parking sites. A list of every company with a car park can help estimate the market, but it does not show a sales funnel. Some sites are too small; for others, the owner is unclear. Even a relevant site does not prove that the company is planning a purchase now.

Describe three signals for your service:

1. The task. What does the buyer need to do, or what problem do they need to solve? For a carport, that could be a decision about using parking space or energy infrastructure. A car park by itself does not confirm that need.
2. Fit. Can you actually help given the scale, region, budget and procurement process? A salesperson should not have to spend a long call uncovering an obvious mismatch.
3. A reason to talk now. Is there an action, event or public signal that makes contact relevant at this moment? A reply to a message, a request for a calculation and a relevant tender indicate different levels of intent. They should not all go in one “hot” column.

This definition also protects the buyer. They will not receive the same “we’d love to collaborate” message just because their address appeared in a database. The manager should be able to explain why this conversation could be useful. If there is no reason, keep researching.

In the PILLAR parking map case, the original research task was to find large parking sites in Germany where a conversation about solar carports might be relevant. The team mapped 40,594 parking sites of at least 5,000 m². In the snapshot dated 13 July 2026, the system marked 6,456 priority sites with a score of 80 or higher. The data is refreshed weekly.

Pause before turning that into a headline like “6,456 leads”. These are priority targets for research and outreach, not 6,456 people who asked for a proposal. The score helps sort sites; it does not tell you whether the owner plans to buy or who makes the decision. The next step is to check the company, the person’s role and the reason for a conversation.

For comparison, look at how the PILLAR tender radar research prototype works. Across April-July 2026, dated snapshots contained 146 unique tender IDs. Nine weekly digests included 98 shortlisted rows; those were used to prepare tailored outreach packs for LinkedIn and phone. The radar added an important element: a specific event to ask about. The published case does not count tenders won, replies to messages or revenue. A shortlisted tender should not automatically be recorded as a qualified lead either.

The stages have more accurate names:

All possible sites → sites that fit the criteria → priority sites to check → companies with a relevant reason to talk → substantive replies → qualified conversations.

The map makes the early steps repeatable. The salesperson handles what comes next: checking whether the information is current, contacting the right person and listening to the reply. The case shows preparation for sales. Its figures do not tell us how many deals followed.

This distinction matters beyond complex B2B products. If someone offers you “ready-made leads”, ask which stage those people have reached. Did they submit an enquiry? Agree to a conversation? Or were they simply identified by company criteria? Each result can be useful, but the work and cost that follow are different.

There are more routes than ad dashboards. A buyer may find a page in search, follow a recommendation, see an ad, read an article and ask a question. This is inbound activity: the person has shown some interest, though their readiness varies. Check not only how many forms arrive, but also whether the page makes clear who the service is for and how to start a conversation.

There is also outbound work. The team identifies suitable companies, looks for relevant roles, checks events and makes individual contact. PILLAR’s parking map and tender radar show this kind of work. Do not expect the first message to close a sale. Its more modest job is to give someone a reason to reply with useful information. For a company selling to other businesses, a relevant B2B conversation on LinkedIn can be one option if someone can run it and the offer has evidence behind it.

Ad platforms offer additional ways to collect enquiries. Google Ads describes lead forms in ads, and LinkedIn offers Lead Gen Forms. These forms make it easier to get in touch, but they do not confirm that a future project is a fit. If someone entered only a name and email, your team still needs to understand the task.

Choose a channel based on how the buyer makes a decision. One person may start a complex purchase, another may evaluate it, and a third may approve it. In that situation, tailored research and explanation can matter more than reach. When someone is already searching for a specific service, they need a page that answers the question and makes it easy to contact you. Ads can bring that page in front of more buyers only if the offer and follow-up process can handle the resulting enquiries.

Starting every channel at once makes it hard to tell which one brought a suitable conversation and which simply produced more names. Pick one route, record its steps, check how an enquiry reaches the owner, and then decide whether to scale it.

Back to PILLAR. The map shows a company with a large parking site. A researcher finds the relevant role and checks for a public event or another sign that the topic may be timely. After an appropriate message, the person replies: “We’re considering options like this for next year. Could you send the site requirements?” That reply is hypothetical; it is not a result from the case. It illustrates the point at which a research record could become a substantive conversation.

Teams can easily lose that handoff. Someone sees a LinkedIn reply, someone else sends a presentation, and nobody is assigned the next follow-up. A week later, the report says “lead”, while the buyer is left wondering whether anyone will get back to them.

Make the next step part of your lead definition:

• who receives the reply and where it is recorded;
• what need they clarify;
• who prepares the material or estimate;
• when the person will hear back;
• how you record whether the conversation continued or ended.

Set a response time your process can meet, rather than borrowing a promise from an ad presentation. One business may need to respond on the same working day; another may need to check the details before giving an accurate technical answer. Name an owner and a deadline the team can keep. If enquiries arrive through a form, check that it reaches a person. If they are scattered across chats, set up a simple tracking process in a spreadsheet or CRM.

The cost of one contact and the cost of one useful conversation can be very different. Both figures can help if they answer different questions. Calculate them consistently:

Before dividing, define the period, included costs and attribution rule. If ads ran in one month but a deal closed the next, “this month’s spend / this month’s sales” can distort the picture. For a long B2B sales cycle, keep the source of the first contact and the dates when a record changed stages. Do not credit a deal to one channel just because its dashboard reported a conversion.

Google Analytics documentation distinguishes the recommended events generate_lead and qualify_lead. These are measurement labels, not ready-made rules for deciding whom your business should call a lead. In Google Ads, the “Leads” goal is tied to the action a campaign is optimised for, such as a form submission. The platform recommends defining the action you expect. If your forms often bring unsuitable enquiries, a form count cannot replace a fit assessment.

Keep at least two records: “what the channel brought in” and “what happened after sales received it”. Analytics and enquiry sources can help check the first. The second needs an owner, a status, a reason for rejection and a next action. If your ads are running but enquiries are not, trace the path from the ad to your team’s reply before changing the budget.

Create a short card for one service. Marketing and sales should fill it in together, then use it on several real enquiries. Copy this template:

1. Who we sell to: company type, person’s role, region and project scale.
2. The task we take on: signs of a need we can help with.
3. What shows interest: an enquiry, reply or other concrete signal; mark the researcher’s assumptions separately.
4. Who does not count as a lead: conditions that make a request unsuitable and where to direct it instead.
5. Who owns the response: a specific role or person, where the record lives and a backup owner.
6. When and what happens next: the promised response time, the first follow-up question and the next action after the conversation.

For PILLAR, the card would start with the type of parking site and the relevant role. A site from the map would go on the research list, not straight into a report called “new leads”. If the company replied about a specific project, the record could move forward with the question it asked. The manager would have the context, and the buyer would not have to explain the task twice.

The card is useful when buying a lead-generation service too. Ask the provider for one anonymised example of a record at each stage and a clear explanation of what they charge for. Paying for a delivered contact is a valid model when it is described that way. If they promise “qualified leads”, the qualification criteria should be visible before launch.

After researching a list, it is tempting to send everyone the same email. Start instead with one verified reason and a short question. The template below is hypothetical; it is not a quote from PILLAR correspondence:

Hello [name]. I saw [a public event or specific site] and am checking who on your team handles [topic]. We work with [brief relevant context]. Would it be useful if I sent you [one specific resource or question to assess], or is someone else responsible for this?

Before sending, check three things. Does the event actually concern this company? Is this person connected to the topic? Will the material you offer help them make a decision? If you do not know, research further. Do not replace a reason for contact with “we’re market leaders”. The buyer did not ask you to rate your own title.

For a tender, your first question might refer to published requirements or the responsible role. For a large parking site with no announced project, use a careful tone: the presence of a parking area does not mean anyone plans to build a carport. Check the rules of the platform and whether contact is permitted for your market and channel.

Choose one B2B service and, over the next week, trace the path from the first signal to a reply. You do not need to buy a new CRM or increase the budget first. Fill in the “Our lead” card, then sort your latest enquiries into the stages above. Mark which were just contacts, which showed a substantive need and where the team lost the next step.

Then check one form or one outbound channel. Submit a test enquiry and see for yourself whether the record appears where the owner looks for it and whether that person knows how to respond. For outbound work, choose a few companies and write down the reason for each contact before sending a message. You will then have a clearer sense of whether you need more enquiries or first need to improve how you handle the ones you already have.