A pricing-page visit can be worth 20 points in one lead-scoring model, five in another, and zero in a third.
That difference is where enterprise lead management gets messy.
A visitor may be evaluating a purchase. They may also be a competitor, existing customer, or someone researching months before a budget exists. The click alone tells you little. What matters is who did it, what else they did, how recently they did it, and whether other people from the same account are showing similar interest.
Buyers are doing a significant amount of evaluation before sales ever enters the conversation. 6sense’s 2026 research found that buyers typically wait until around 61% of the buying journey before speaking with sellers, while roughly four out of five vendors on their shortlist are already there from day one [1].
So, the useful question is: Which signals should actually change what we do next?
One Score Should Not Answer Three Different Questions
Scoring gets unreliable when fit, interest, and sales readiness are pushed into one number.
Keep three decisions separate:
- Fit: Is this a company you want to sell to? Look at industry, company size, region, product match, and customer status.
- Intent: Is something happening now? Look at recent product research, pricing visits, event activity, comparison content, and repeat engagement.
- Action: What should happen next? Use product interest, territory, current account owner, buying role, and open opportunities to decide.
Adobe’s B2B predictive-scoring documentation takes an account-level approach by aggregating person activity into account scores and tying scores to defined conversion outcomes [2].
This helps prevent a high-fit account with weak current interest from receiving the same treatment as one where several people have started researching the same product this week.
Score Patterns, Not Random Clicks
One ebook download should rarely decide a handoff.
Look for combinations across three dimensions:
Recency: Did the activity happen yesterday or two months ago?
Depth: Was it a general blog visit or repeated research on pricing, product, implementation, security, or comparison pages?
Spread: Is one person active, or are several people from the account showing related behavior?
Consider two accounts. Account A is a strong fit, but one director downloaded a guide 45 days ago and has done nothing since. Account B is also a strong fit, and this week an IT leader returned to a product page twice, another contact viewed security documentation, and a third registered for a product webinar.
Account B gives the team a much clearer reason to act.
Old activity should also lose influence over time. A lead should not stay “hot” for months because of research that never continued.
Route on the Reason, Not Just the Score
A score of 85 can tell you a lead deserves attention. It cannot tell you who should own it.
Routing needs its own logic.
If the account already has an owner, keep continuity. If the activity points to a particular product, route accordingly. If contacts from an active opportunity suddenly become engaged, alert the opportunity owner instead of creating another lead for a different rep.
The rules could look like this:
High fit + recent high-intent activity + no open opportunity → assign to the correct owner and alert sales.
High fit + active opportunity → notify the current opportunity owner and add the new activity to the account context.
Strong interest + poor fit → keep the contact in marketing instead of forcing a sales handoff.
That final rule matters. Automation should reduce wasted follow-up as well as speed up good follow-up.
Personalization Should Follow What Changed
Personalization gets weak when the system knows a buyer’s name and industry but ignores what they just did.
Someone repeatedly reading security and implementation content does not need another broad industry email. Their behavior is pointing toward a question.
A buyer researching pricing may need commercial context. A technical evaluator reading security material may need architecture, compliance, or implementation detail. Several people researching the same product may justify coordinated account-level messaging instead of separate generic nurture journeys.
A useful rule is: when the signal changes, check whether the message should change too.
Build the Signal-to-Action Map Before the Workflow
Map the few signal patterns that should genuinely change treatment before adding more automation.
| Signal pattern | Score | Route | Campaign |
| Repeat product or pricing research from a target account | Raise current intent | Alert existing owner or assign by territory | Shift to product-specific content |
| Several contacts research the same topic | Raise account-level intent | Surface the buying group to sales | Coordinate around the shared topic |
| High fit, little recent activity | Keep fit high, let intent cool | No immediate handoff | Continue lighter nurture |
| Strong activity from a poor-fit account | Intent rises, fit stays low | Hold from sales routing | Keep in an appropriate marketing path |
The exact rules will differ by business. The standard is more useful: every signal needs a reason, every score change needs a consequence, and every automated action needs an owner.
Start With the Decision That Keeps Going Wrong
You do not need to automate every buyer signal at once.
Start with one decision your current process keeps getting wrong. Maybe strong account activity sits untouched for too long. Maybe leads with very different levels of interest receive the same follow-up. Or perhaps multiple contacts from one account keep getting treated as unrelated leads.
Trace the signals behind that problem, decide what should happen when they appear, and automate that specific response first. Once the scoring, routing, and follow-up work as intended, expand from there.
If you want another set of eyes on where buyer intent signals are getting lost between marketing and sales, email our team at info@growthnatives.com. We can help identify the first scoring, routing, or campaign decision worth fixing.
Sources
[1] 6sense