A definition you can quote, first-party against third-party signals, an illustrative worked example of surge, and an honest account of how much intent data can tell you.
Free Forever • No Credit Card Required
Quick answer
An intent score estimates how actively an account or person is researching a purchase in your category at the moment, based on observed signals rather than on who they are.
It is a timing measure. Fit answers whether an account is worth selling to; intent answers whether this is the week to call them. Keeping the two separate is what lets a team act sensibly on a perfect-fit account that is currently quiet.
Behaviour on your own properties: pricing pages viewed, documentation read, a demonstration requested, an email replied to, a trial started. Reliable, tied to a known person or account, and available to anyone willing to instrument their own site.
Activity observed across other publishers and communities, resolved to a company. Its advantage is reach into accounts that have never visited you. Its limitation is precision: resolution to an organisation is approximate, and it identifies a company rather than a person.
Not all activity is research. Careers pages, customer support documentation, login pages, and traffic from your own staff should be excluded or scored negatively, otherwise the model will reliably promote your existing customers and people applying for jobs.
Surge = recent-period activity for the account and topic ÷ that account's own baseline activity for the same topic, compared against a threshold.
The choice of denominator is the whole idea. A large organisation generates more activity than a small one at all times, so comparing raw volumes across accounts would simply rank companies by size. Comparing an account against its own history isolates the change.
These figures are illustrative and are not benchmarks. An account's baseline consumption of a topic across the preceding twelve weeks averages four content interactions per week. Over the last two weeks it records eighteen and then twenty-two, averaging twenty per week. Surge = 20 ÷ 4 = 5 times baseline. Against a threshold of three times, the account is flagged.
What that flag actually supports is modest: this account is worth a call this week rather than next month, on a topic that is probably relevant. It does not say who inside the company is researching, whether a budget exists, or whether a purchase process has begun.
Intent decides sequence. Given a target list of two hundred good-fit accounts and capacity to work forty properly this quarter, intent is how you choose which forty. That is a genuinely useful decision and it is the main thing intent data is good for.
It also shapes the opening topic. If the signals cluster around a particular problem, leading with that problem tends to produce a better conversation than leading with a product overview, without ever referring to the signal itself.
First-party signals are more reliable, individually attributable, and already yours. Teams often purchase third-party data while their own pricing page generates no recorded signal at all.
Intent without recency is history. A score that only accumulates will keep flagging accounts whose research finished months ago.
A company-level flag says nothing about which of two hundred employees was reading. Representatives who assume otherwise open calls with the wrong person and the wrong premise.
Telling a buyer that you noticed their organisation researching a topic is uncomfortable to receive, frequently wrong at the individual level, and depending on the data source and jurisdiction may raise privacy questions worth taking advice on.
Combined with fit into one total, intent becomes invisible, and the team loses the ability to distinguish a good account that is quiet from a poor account that is busy.
A useful intent setup has classified signals, a decay function with a stated half-life, first-party signals instrumented before any feed was purchased, scores stored with their contributing signals, and a demonstrated difference in conversion between high-intent and low-intent contacts within the same fit band.
A poor one has an expensive feed, a list of flagged companies with no named contacts, scores that never fall, and representatives who quietly ignore the whole thing because acting on it produced awkward calls.
| Source | Strength | Limitation |
|---|---|---|
| Pricing page views | Strong, specific, and tied to a known visitor | Only covers people already aware of you |
| Demonstration requests | Explicit and unambiguous buying interest | Low volume, and arrives late in the process |
| Email replies | Genuine engagement from a named person | Easily confused with polite acknowledgement |
| Third-party topic surge | Reach into accounts that never visited you | Company-level, approximate, and no named contact |
| Email opens | Cheap and abundant | Unreliable, as privacy features can trigger them |
Store the intent score on the account with its contributing signals and a timestamp, and pair every flagged account with a named contact before it reaches a representative. A flag with a person, a topic, and a date is a call worth making. A flag on its own is a company name, and company names are not a work queue.
Continue with AI CRM capabilities, lead management software, what a CRM is, sales automation, CRM for small business, and product features.
Every team in this space loses revenue to the same recurring gaps. Here is what they cost you and how HelloGrowthCRM closes each one.
A third-party feed was bought, and representatives cannot tell what to do with the accounts it flags.
Third-party intent identifies which accounts to contact, not what to say or who to ask for. Pair each flagged account with a named contact from your own data, a specific topic the surge relates to, and a first-party check for any activity on your own site. Without those three, the feed produces a list nobody can act on.Signals paired with a next action
Intent scores never fall, so the same accounts stay flagged for months.
Apply decay so that signals lose weight with age, and set a half-life short enough to reflect a real buying window. Intent describes a moment. A score built by accumulating signals without decay stops measuring activity and starts measuring how long an account has been in your database.Decay with a stated half-life
Support traffic and job applicants are inflating intent for existing customers and non-buyers.
Classify pages and signals deliberately. Careers pages, customer support documentation, and login pages should be excluded or scored negatively rather than counted as research. A model that treats all activity as interest will consistently promote your own customers and people applying for jobs.Signal classification
Representatives open calls by referring to research the buyer never told them about.
Use intent to decide timing and topic, not as something to quote back. Opening with knowledge a buyer did not share is uncomfortable and can be a compliance issue depending on the source and jurisdiction. The useful version is an ordinary, well-timed conversation about the topic the signals suggest.Timing, not surveillance
AI-powered CRM with the features you need to close more deals.
Common questions about using HelloGrowthCRM in your industry.
Join small businesses that close more deals with HelloGrowthCRM.
Free Forever • No Credit Card Required
Free Forever • No Credit Card Required
Prefer email? Write to sales@hellogrowthcrm.com