How to qualify inbound leads: fit, intent, timing

A full pipeline can hide a weak sales process. That's the warning at the top of Pipedrive's September 2026 guide by Syed Ali Nemath. Reps spend hours on leads that were never going to close. The best prospects slip through while they do it.

So the first problem isn't the leads. It's the definition. HubSpot's 2025 Sales Trends Report, quoted in HubSpot's own qualification guide, found 73% of sales reps rate their marketing leads as high or very high quality. And 27% say they need better leads from marketing. Same leads, different rules. Fix that before you build any score.

Inbound needs its own rules because the buyer came to you. Abe Dearmer's Sendspark playbook cites Harvard Business Review: inbound leads convert 8x more often than outbound, but only if you reach them within five minutes. Most B2B teams take five hours, the same playbook says. A slow, careful process loses the very lead it was built for.

Don't treat every form fill alike. A demo request and an ebook download are different signals. The first company wants to talk to sales. The second wants a PDF and may never want more. Put them on separate paths from day 1: demo requests go to a rep, downloads go to nurture until the company's behaviour changes.

This doesn't apply if only a handful of inbound leads arrive each month. At that size a founder can read every one by hand. A scoring model is overhead you don't need yet. Either way, list every form on your site and mark each one sales-ready or nurture. That list is the first version of your rules.

Pipedrive's guide has 4 steps: define an ICP, score for fit and behaviour, apply a framework in discovery, then commit or disqualify. Sendspark's playbook uses 5. The list below follows Pipedrive's order, split into single actions. Routing goes last because of Sendspark's five-minute figure.

1. Write your ICP down: the industry, company size, region and buying role you can actually serve.
2. Write a separate ICP line for each market you sell into, such as US companies and Indian companies.
3. Score fit on its own: does this company match the ICP?
4. Score behaviour on its own: what did this company do on your site and with your emails?
5. Keep the 2 scores apart, so a perfect-fit company that read 1 blog post stays in nurture.
6. Pick one framework for the first call: BANT for high-volume SMB sales, Meddic for complex enterprise deals.
7. Use that framework across the whole team, so a score means the same thing to every rep.
8. Commit or disqualify after discovery, and write the reason in the lead record.
9. Route every new demo request automatically, through a scoring rule or an enquiry form that asks the qualifying questions as it arrives.

The per-market ICP line matters more than it looks. A US company and an Indian company can match on industry and size. They can still differ on budget, on who signs and on the privacy rules you work under. If you need a starting point for the ICP itself, the ICP guide covers it.

The 2 scores carry most of the weight. See, fit says whether you could ever sell to this company. Behaviour says whether it's looking right now. Pipedrive keeps them apart for that reason, because a blended number hides which one's missing.

For the framework, Sendspark adds CHAMP and FAINT to the list. Don't run Meddic on a low-priced self-serve product, though. A long discovery call costs more than the deal's worth. BANT will do.

Commit or drop is the step teams skip. A lead dropped with a reason teaches your ICP something. A lead left open teaches nothing. It also makes the pipeline look fuller than it is, which is exactly what Pipedrive warns about.

The ranking guides score behaviour as email clicks, content downloads and discovery answers. Pipedrive's split is right. But that behaviour mostly happens after the lead hands over an email address.

What the guides leave out is the most direct buying signal you own. Which companies came to your site, which pages they read, and in what order. It's company-level data. That's enough, because fit is a company question anyway.

Page order is the signal. Take Company A and Company B, both inside your ICP. Company A reads a blog post on lead routing, then your pricing page, then a comparison page, then your docs. Company B reads 1 blog post and leaves.

Next week, both download the same ebook. Score them on the form alone and they look identical. Score them on page order and they don't. Company A is evaluating you. Company B is researching a topic.

So give each page type a meaning before you score anything. Pricing means a company is checking cost. A comparison page means it's shortlisting. Docs mean its team is testing whether your product will work for them. Return visits count too: a company that keeps coming back to pricing is further along than one that came once.

You can run this on your own site this week:

• pull last month's visits that resolved to a company, and drop every company outside your ICP
• mark pricing, comparison, docs and case-study pages as signal pages
• rank what's left by signal pages read and return visits
• give the top of that list a Priority tier, and compare it with the form fills you actually got

That last comparison is the useful one. If companies at the top of your list never filled a form, your form's the bottleneck. Not your traffic.

Now the limits. A visit doesn't always resolve to a company. A name from visit data can be wrong, too, so check it against your ICP before you treat it as a lead. For US and Indian sites alike, requirements vary by jurisdiction and your consent setup; use your own legal review.

It also won't work on a site with little traffic. There isn't enough page-order data to rank anything. Stay with form fills and discovery calls until traffic grows. Your decision here: write the list of signal pages before you look at a single company.

The best score is worth little if the lead waits. Sendspark's five-minute figure from Harvard Business Review is why. Set it against the five hours Sendspark says most B2B teams take, and a careful manual process is already too slow.

The first trap is a shared inbox. Demo requests land in it, someone sorts them after lunch, and the window's gone. Routing has to run by itself, through a scoring rule or an enquiry form that asks the qualifying questions the moment a lead arrives. Assign the owner by rule. Not by whoever's free.

The second trap is turning the first call into an interrogation. Utmost Agency's guide says it plainly: most leads aren't ready to buy. That's a reason to ask fewer questions early, not more. Use your framework to decide what you must learn on call 1, and leave the rest for later.

The third trap is routing on fit alone. A company that matches every ICP line but has shown no intent jumps the queue. A weaker fit that kept coming back to your pricing page waits behind it. Keep the behaviour score in the routing rule.

The five-minute rule is for demo requests and other sales-ready forms. It isn't for an ebook download. Calling a company 5 minutes after a PDF download treats a research signal as a buying one, which is the form-fill mistake again.

So test the right thing. Submit your own demo form from a work address and time the reply. If it takes longer than five minutes, fix routing before you touch the scoring model.

Pipedrive's guide has a whole part on what to do when qualification stops delivering good opportunities. The useful reading is simple. When good-looking leads stop closing, check the definition first. The reps come second.

Start with your disqualification reasons. If the same one keeps appearing, such as no budget or wrong company size, your ICP is letting the wrong companies through. Tighten that line. If most dropped leads carry no reason, you can't diagnose anything. Make the reason a required field first.

Then go back to the sales and marketing gap. HubSpot's 73% and 27% split shows each team scoring against a different picture. Put both teams in a room with the last batch of disqualified leads. Agree which ones should never have reached sales, and change the rules to match.

The market split matters here too. A US B2B SaaS team and an Indian one can sell the same product and still need different ICP lines. Ask the budget question in the currency the buyer plans in: USD for a US company, rupees for an Indian one. Don't convert and compare raw numbers. None of these sources gives a benchmark for either market, so build your own from your disqualification reasons.

This doesn't apply to a brand-new product with no closed deals. At that stage your ICP is a guess. Qualify loosely, and talk to more companies than the rules allow. Let the first wins write the ICP. For now, pick a review date and change at least one ICP line based on what the dropped leads show.

PageFox is the wrong choice for a site with little traffic: it tells you which companies are already visiting, it does not bring new visitors.

Decide which pages on your site count as a buying signal, and write your ICP down with a separate line per market before the next demo request arrives. Then submit your own demo form and time how long the reply takes.