B2B buying signals: which signals show who wants to buy right now?

The short answer

The strongest B2B buying signals come from your own data: a reply to an email, a meeting, several visits to the pricing page, an old proposal that comes up again. External signals such as job changes, job postings, funding rounds or a change in the tech stack mainly provide the occasion, not the intent to buy.

On its own, hardly any signal says much. Only together, weighted and bundled per account, do they add up to a temperature: hot, warm or cold. In data protection terms, the most important difference is whether a signal is attached to a company or to a person.

What is a B2B buying signal?

A buying signal is a hint that something is moving at a target customer right now that fits your offer. It can be a direct reaction, such as a question about a proposal. But it can also be an event that makes a need more likely, for example a new head of sales or a system change.

In B2B, a single person rarely buys. Usually several roles decide together, the so-called buying committee. That's why we score signals per account, that is per company, not per contact. Three visits from three people at the same company say more than ten visits from one person.

Roughly, signals fall into two groups. First-party signals come from your own systems: CRM, inbox, calendar, website. Third-party signals come from outside: job boards, press releases, registers or data providers.

Which signals are in your own data?

Replies to emails

A reply is the clearest signal of all, because a person actively reacted. It becomes especially telling when it comes from a new role, for example when the finance lead suddenly writes after the business unit.

Meetings and calls

A booked meeting, a second call or an invitation to more colleagues shows that a topic is being carried further internally. Conversely, cancelled and postponed meetings are a warning sign for open deals.

Website visits, especially on the pricing page

Someone looking at prices, integrations or references is usually deeper into research than someone reading a blog post. The pattern is what counts: several visits in a short time, on pages close to a purchase, from several people. In Germany you generally need consent to analyse individual visitors, more on that below.

Deals and history in the CRM

Your CRM knows lost deals, proposals with a follow-up date and customers whose contract runs out soon. An account that failed on budget a year ago and now shows up on the website again belongs right at the top. Without a system, these signals are the easiest to lose.

Which signals come from outside?

Job changes and champion changes

When your contact changes company, two things happen: you lose a contact at the old account, and at the new employer there is someone who already knows you. Such a fan, often called a champion in sales, is one of the best occasions for a conversation. Watch the source: LinkedIn's user agreement prohibits automated scraping of profiles by software or bots [5].

Job postings

Open positions show where a company is investing. If a target customer is looking for a head of RevOps or several people in sales, the topic is becoming important internally. But a job posting doesn't prove an intent to buy; it gives you a good hook.

Funding, new leadership, growth

A funding round, a new location or new management change budgets and priorities. New leaders often bring their own ideas and their own service providers. That opens doors, but sometimes closes some too.

Technology changes

Anyone switching their shop system, CRM or website platform often needs support around it. Software versions in the tech stack that are reaching end of life can also trigger a need. Which technology signals count depends heavily on your industry.

Intent data

Intent data are signals from data providers that are meant to show that a company is researching certain topics unusually often. The provider Bombora, for example, measures reading behaviour across many B2B websites and compares it with the company's usual level. It explicitly speaks of account-level signals [6]. So intent data show a trend, but not who exactly is researching and why.

How reliable and how GDPR-compliant are the signals?

The "strength" column is our assessment from project experience, not a measurement. The "data protection" column is a rough classification for Germany and not legal advice. The details follow below the table.

Company or person?

The GDPR protects the data of natural persons. Data about legal persons, such as company name, legal form or the company's contact details, are not covered by the regulation (recital 14) [1]. That a limited company posts a job or switches its shop system is therefore usually unproblematic. But as soon as a signal is attached to an identifiable person, it is personal data (Art. 4(1) GDPR) [1]. That applies to job changes, to the visit of a recognisable person to your website and to every contact in the CRM.

Legal basis and limits

Without consent, the main legal basis for advertising purposes is legitimate interest (Art. 6(1)(f) GDPR). According to recital 47, direct marketing can be such an interest, but it always requires a balancing against the interests of the people concerned [1]. The German data protection conference (DSK) draws a clear line here: detailed profiles, behavioural predictions and advertising scores from external sources such as social networks generally mean that the interests of the people concerned prevail. Then consent is required [3]. The DSK regularly considers it impermissible to use contact details from a website's legal notice for advertising [3].

In practice, in our view, that means: build the temperature at account level, use person-level signals mainly from your own business relationship, and don't build shadow profiles of individual people.

Inform people and respect objections

If you collect a person's data from somewhere other than the person themselves, you have to inform them, including about the source. This must happen at the latest with the first message to them and at the latest one month after receiving the data (Art. 14 GDPR) [1]. If someone objects to advertising, their data may no longer be processed for it (Art. 21(3) GDPR) [1]. This objection belongs in the CRM as a block, otherwise the contact shows up again with the next signal.

Website visits

Under § 25 TDDDG, cookies or similar technology that recognise a visitor are only permitted with consent. The only exception is what is strictly necessary for the service requested [2]. For the processing that follows, the DSK writes that legitimate interest is sufficient for tracking only in a few constellations [4]. That's why we only analyse website visitors with consent. German competition law also applies to the channel: email advertising requires consent in B2B too, more on that in the article Cold email in B2B: legal framework and success factors (German).

How do signals become an account temperature?

A list of 200 signals helps no one. Your sales team needs one simple call per account: is it hot, warm or cold right now, and what is the next step? That is exactly what the signal engine on your CRM does. It builds the temperature from emails, calls, meetings, deals and website visits and adds external signals. Every morning this produces a daily list with a reason, a contact and a hook.

These are the rules we use to weight signals:

• Fit before signal. A signal only counts for a target customer. If a company doesn't match the ideal customer profile, even a visit to the pricing page doesn't make it hot.
• Closeness to the decision. Your own interactions such as replies, meetings and the pricing page weigh more than external occasions such as job postings or intent topics.
• Recency. Signals lose weight over time. A visit from yesterday counts more than one from three months ago.
• Breadth. Several signals and several roles from the same account are stronger than a single spike.
• Learning. Reporting shows which signals actually lead to deals. The weighting is adjusted accordingly.

Example (fictional companies): At Musterwerk GmbH there were three visits to the pricing page, plus a reply from the finance lead. The account is hot; the next step is suggesting a meeting. Mustermann Handel switched its shop system, nothing else happened. It is warm, and a case study on the new platform is a good hook. Probe IT Services was hot 34 days ago and has been quiet since. It is cold, but lands on the list for reactivation instead of quietly disappearing.

Because the temperature has a history, you also see when an account cools down. If it drops, the engine reminds your team and suggests an occasion. What that looks like is shown by the sample daily list on GTM engineering and on the page How the signal engine works.

Which mistakes should you avoid?

Too many signals at once

Connect every available data source and you get a list in which everything is a little bit warm. From our project experience, it's better to start with a few signals that fit your market and only add more once the team is working with the list.

Noise instead of signal

Not every visit is buying interest. Applicants on the careers page, existing customers in the help section, your own employees and bots distort the picture. These cases have to be filtered out before the temperature is calculated.

Bad data in the CRM

A signal is only as good as its match. If an account exists three times, or a contact who left long ago is still listed, the signal lands on the wrong record. That's why cleaning the CRM always comes first.

Spray and pray

A signal is no free pass for a mass sequence. The same message to every warm account burns good accounts and your sending domain. Outreach should fit the signal and the role. A job change needs a different message than a visit to the pricing page.

A signal without a next step

A hint that no one can act on gets ignored. Every signal on the daily list needs a reason, a contact and a suggestion of what to do now.

Who is this for, and how do you start?

Systematically analysing buying signals pays off for B2B companies with a CRM and their own sales team whose deals are big enough that timing matters. If a single person does the selling or the contacts still live in Excel, a well-kept list is usually enough. The role a GTM engineer plays here is covered in the article What is GTM engineering?

With us, the way in is the CRM Check: €3,000 net at a fixed price, 1–2 weeks, for a CRM with up to 20,000 records. You get a data quality score before and after, a completed clean-up, the top fixes and a first list of accounts that already show signals today. What setup and operation cost afterwards is covered in the article What does GTM engineering cost?