8 min read

Where Should a Buying Signal Live in Your CRM?

Where Should a Buying Signal Live in Your CRM?

A rep opens the top account on today's list and cannot answer the only question that matters: why is this company here, and is that reason still true?

Capturing buying signals is the easy part now. Most tools push something into Slack or the CRM within minutes of a trigger firing. What breaks is everything after capture. The signal lands in a channel nobody owns, in a shape nobody can read later, with no date attached, and within a month it has stopped being evidence and started being clutter.

That is a storage and routing problem, and it is solvable in an afternoon if you decide what a signal record actually is before you connect another source.

What Does It Mean to Track Buying Signals in Your CRM?

Tracking a buying signal in your CRM means storing each observed signal as a dated, sourced record attached to an account, in a place both a rep and an automation can read. A signal that exists only as a Slack alert or as points inside a score is not tracked. It is announced, and then it is gone.

The test is simple. Six weeks after a signal fires, can someone who was not involved answer four questions: what happened, when it happened, where you saw it, and whether it still counts? If any of those answers requires finding the person who first spotted it, the signal is not in your CRM in any useful sense.

Most Advice on This Starts With Infrastructure You Do Not Have

Most published guidance on intent data and CRM assumes a data warehouse and an identity resolution layer sit in front of the CRM. ZoomInfo's 2026 CRM integration guide treats account matching and enrichment as prerequisites before intent signals can act, which is the right sequence when you ingest millions of page-level events a week and have to resolve them to accounts before anything downstream works.

Most teams reading this are not in that situation. A fifteen-person sales org watching hiring pages, funding announcements, podcast appearances and competitor reviews is not processing millions of events. It is processing perhaps thirty or forty meaningful signals a week, most of them already attached to a company name a human recognized. At that volume the warehouse is not the constraint. The record format is, and that is the part almost nobody writes down.

The Six Fields a Buying Signal Record Needs

A usable signal record carries six pieces of information: what happened, where you saw it, when you saw it, what type of signal it is, how it was confirmed, and when it expires.

  • What happened, in one plain sentence. Write "posted two RevOps roles and a lifecycle marketing role on November 12" rather than "hiring activity." A rep writing a message at 8am should be able to use that sentence as it stands.
  • The source URL. This is the field most teams skip and the one that decides whether anyone trusts the record. A signal without a link is a rumor, and the link is also how the next person checks whether the posting is still live.
  • The date you observed it. Keep this separate from the date the event happened. A funding round announced in September and found in November is a stale signal wearing a fresh timestamp.
  • The signal type. This is the field you score, filter and report on, so keep the list short and stable. Ten to fifteen types is enough for most teams, and what each signal type actually proves matters more than how many you track.
  • How it was confirmed. Vendor detected, rep confirmed, or inferred. Collapsing those into one score is how a team ends up treating a guess and a press release as the same evidence.
  • The expiry date. Set it when the record is created, not when someone gets around to reviewing it.

Ownership does not need a field of its own. The signal inherits the account owner, and duplicating that creates two places to be wrong.

Should a Signal Be a Field, a Timeline Event, or a Custom Object?

Use fields for the current state, the timeline for history, and a custom object only when you need to analyze signals as a population.

Fields on the account answer what is true right now. Four are usually enough: last signal summary, last signal date, active signal count, and top signal type. Those are what a rep reads before a call and what routing rules can act on. HubSpot treats the record itself this way, where the intelligence tab shows enrichment and visit data to inform qualification, routing and outreach decisions.

The timeline answers what happened here before. Individual events belong there rather than in fields, because fields overwrite and timelines accumulate. Third-party signals usually arrive in that shape already. HubSpot's Bombora integration, for instance, places Company Surge signals on the company record timeline alongside your other intent signals.

A custom object answers which signals actually produce meetings. It is the only structure that lets you query signals independently of the accounts they hang off, and it is also the most expensive choice. Salesforce allocates 50 custom objects and 100 custom fields per object on Professional Edition, so the decision competes with everything else the business might want to model. Build the object when you are ready to report on signal performance, and not before. Until then, fields plus the timeline carry most teams further than they expect.

A Signal With No Expiry Date Becomes Background Noise

Every signal has a window in which it justifies an interruption, and that window closes.

This is why the expiry field matters more than it looks. Without it the active signal count only ever rises, every account eventually looks warm, and the list stops ranking anything. With it, accounts drop off on their own and the team's attention follows the signals that are still live.

Sensible starting windows differ by type. A job posting stays relevant until it is filled or roughly two months pass. A funding announcement carries for a quarter. A leadership change can run longer, because a new executive's first real projects often start months after the announcement. Treat those as defaults to calibrate rather than as benchmarks. Your own closed-won data will correct them faster than any published number, and how long a buying signal stays actionable is worth measuring directly. If you also weight signals in a score, recency decay belongs in the scoring model rather than in the record.

Storage Is Not Routing

A stored signal changes nothing until someone is responsible for acting on it.

Three rules cover most of this. Every signal gets one owner, inherited from the account, so it is never owned collectively. Every signal type gets one named first action and a window, for example a hiring signal means the owner sends a LinkedIn message within 48 hours. And every signal that passes its window without an action lands in a visible queue.

That third rule is where implementations fail quietly. Signals keep arriving, the count keeps climbing, and nobody notices that none of them were touched. The steps between a signal firing and a reply landing are where most intent programs lose their value, and routing is usually the first one to break.

Define triggers narrowly enough that the queue stays readable. HubSpot lets you define buyer intent triggers in plain language, from hiring updates to RFP postings, with an account-level monthly limit on how many are tracked. That limit is a feature. A trigger set that fires on everything produces a queue nobody opens.

What a Finished Record Looks Like

Here is a hiring signal stored as a record rather than announced as an alert:

  • What happened: posted two revenue operations roles and a lifecycle marketing role on November 12
  • Source: the careers page listing those three roles
  • Observed: November 13
  • Type: hiring, revenue operations
  • Confirmation: rep confirmed, all three postings live at time of check
  • Expires: January 12

A rep opening that account in December sees a signal that is still live, knows exactly what to reference, and can verify it in one click. The message almost writes itself from the summary field: three new revenue and lifecycle roles usually means reporting and handoffs are being rebuilt, not simply that headcount is growing. That is a specific hypothesis the record supports, and a prospect can correct it in one sentence, which is the point.

Compare that with the same event stored as a score moving from 71 to 84. Nothing in that number tells the rep what to say.

Where Alsona Fits

Alsona is built around the record rather than the alert. It tracks unstructured buying signals across the web, including hiring, funding, podcasts and competitor moves, scores each account against your ICP with the reasoning visible rather than just a number, and produces a ranked daily list of accounts to work. The signals it surfaces are the ones that arrive without any structure at all, which is exactly why they need a record format before they reach a rep.

On the CRM question specifically, Alsona syncs two ways with Salesforce and HubSpot, and setup includes a mapping step where you define how prospects, accounts, signals and outcomes move between Alsona and your systems. That mapping is the decision this article is about: choosing once where a signal lands and what it carries with it, instead of letting each source decide for you. Scoring sits on top of the same records, so the account ranking can show its reasoning instead of asking reps to trust a number.

The Takeaway

The useful question is not which signal source to buy next. It is whether a signal that arrives today will still be legible in six weeks, to someone who was not watching when it fired.

Decide the six fields. Choose field, timeline or custom object deliberately. Set expiry at creation. Give every signal type one owner and one first action, and put the misses in a queue someone reads. That work takes an afternoon and it decides whether everything you buy afterward is worth anything.

Build outbound around signals your team can still read a month later. See how Alsona scores accounts on real buying signals and shows the reasoning behind each one.

Frequently Asked Questions

Should buying signals be stored on the contact or the account?
Do you need a data warehouse to track buying signals?
How many buying signal types should a team track?
How long should a buying signal stay active?
What is the difference between a buying signal and an intent score?

Time is Money.
Alsona saves you both.

Stop building sequences. Start building pipeline.