Published by CXJ Labs, an automation company. Not independent. Read our disclosure.
The Automation Index

Can lead qualification be automated?

Largely automatable: 60% of task time

The answer

Partly. Collecting form answers, researching the company, scoring against clear criteria and logging the outcome automate well today. In our model about 60% of qualification time can move to software. The qualifying conversation for borderline leads stays human, because buying intent, authority and timing are rarely stated plainly on a form.

The tasks, and how much of each is automatable today

Six tasks make up qualification. Reviewing form answers and logging outcomes are 85 to 100% automatable. Company research is about 75% automatable with enrichment tools and AI. Scoring is 60% automatable if your criteria are written down. The conversation with borderline leads stays about 80% human and takes the most time.

TaskActivity typeMin / weekAutomatable
Review form answersRules and AI read structured answers reliably. Free-text answers that contradict the dropdowns still need a human glance.Collecting data
base rate 64%		12085%
Research the company and contactEnrichment tools return size, industry and location. AI summarises the website. Small or new businesses often return little data.Collecting data
base rate 64%		16075%
Score the lead against your criteriaAbove McKinsey's 18% base for expertise because written criteria turn most of the judgement into rules. Unwritten criteria stay human.Applying expertise
base rate 18%		8060%
Qualifying conversation with borderline leadsAI can prepare questions and a summary. Hearing real budget, authority and urgency needs a person.Interacting with people
base rate 20%		20020%
Disqualify politely and route to nurtureA templated message and a nurture tag. A person should check large accounts before they are turned away.Processing data
base rate 69%		4090%
Record the outcome and reasons in the CRMScore, lane and the reasons behind it are written automatically, so sales can see why a lead was booked.Processing data
base rate 69%		40100%
Whole process64060%

Base rates are McKinsey Global Institute's technical automation potential for each activity type. The share for each task is our adjusted estimate. Method.

Hours saved per week

For a team qualifying about 80 inbound leads a week, the work takes roughly 10.5 hours of a setter's week. Automating research, scoring and logging returns about 6.5 hours a week. The setter keeps the conversations with leads the score cannot decide, which is where qualification skill actually matters.

Modelled on: A B2B services firm receiving about 80 inbound leads a week from its website and ads, with one setter qualifying before booking calls.

  • 80 inbound leads a week, each with a short form.
  • About 1.5 minutes to review form answers and 2 minutes to research the company per lead.
  • About one in three leads is borderline and needs a conversation of around 7 to 8 minutes.
  • Qualification criteria (company size, budget range, service fit) can be written down in a short list.

10.5 hours a week by hand, 6.5back with automation. Formula: the sum of each task's minutes multiplied by its automatable share, divided by 60, rounded to the nearest half hour.

Tools that do the job

You need a form that asks qualifying questions, an enrichment source for company data, and somewhere to apply a score. Typeform or a GoHighLevel form collects answers. Clay enriches the company. An AI agent applies your written criteria. n8n, Make or Zapier writes the score and reasons to HubSpot or Pipedrive.

Typeform
Asks the qualifying questions, with logic that skips irrelevant ones.
Clay
Enriches each lead with company size, industry, location and web presence.
AI agent
Applies your written criteria and explains the score in two sentences.
n8n
Runs the enrichment, scoring and lane routing, and writes everything to the CRM.
HubSpot
Holds the score, lane and reasons on the contact and deal records.

A simple workflow

The workflow starts with a form submission. The lead is enriched with company data, scored against written criteria by rules and an AI agent, and sorted into three lanes: clear fit goes straight to booking, clear miss goes to nurture, and the uncertain middle goes to a setter with a research summary attached.

  1. TriggerA lead submits the form or is created in the CRM from an ad.
  2. EnrichLook up the company and summarise the website in three lines.
  3. ScoreApply hard rules first (country, size, service), then an AI score against written criteria, with reasons.
  4. SortClear fits get a booking link; clear misses get a polite note and nurture; the middle goes to a setter.
  5. LogWrite the score, lane and reasons to the CRM, and review a sample of AI decisions weekly.

Keep these human

  • Conversations with borderline leads. Most real disqualifiers, such as no authority or no urgency, only come out in conversation.
  • Final say on large accounts. A wrong automatic 'no' on a big prospect is expensive.

Where it goes wrong

  • Unwritten criteria. If the team cannot write down what a good lead is, the AI will invent a standard. Write the criteria first.
  • Silent false negatives. Nobody reviews the leads that were turned away. Sample ten rejected leads a week and check them.
  • Scoring on form answers alone. People round up budgets on forms. Weight behaviour and enrichment data alongside self-reported answers.

What most guides miss

Most guides treat qualification as a score. The useful design is three lanes, not a number: clear yes, clear no and uncertain. Automating the two clear lanes is safe. Sending only the uncertain middle to a person concentrates human judgement where it changes outcomes, and makes the setter's time far more valuable per conversation.

The evidence