MQL vs SQL: where marketing's job ends and sales' begins.
Both terms are usually explained as slide definitions. In your CRM they are fields, and only one mainstream platform ships them out of the box. Here is what each one actually is, in five systems, with the thresholds that move a lead across.
Toni MedicSalestructSeptember 30, 202612 min readSales Operations
MQL and SQL get taught as two boxes on a funnel diagram. In a working CRM they are not boxes, they are a field value, a rule that sets it, and a stopwatch that starts the moment it changes. Only one of the five platforms in this post ships that field by default. Everyone else builds it, and most teams building it get the mechanics wrong before they ever argue about the definition.
The short version
- Of the five platforms checked here, only HubSpot ships "Marketing Qualified Lead" and "Sales Qualified Lead" as native values. Salesforce, Pipedrive, Zoho CRM and Adobe Marketo Engage all require you to build the equivalent yourself.
- Salesforce's own tutorial content uses a starting automation rule of engagement score above 150 and fit grade above B+ before a lead auto-routes to a rep, with rule-of-thumb minimums of score 100 and grade B-.
- A score alone tells you interest, not fit. Salesforce's default grade for a record with no fit data is a D, which is the platform's own way of saying an ungraded lead should not pass.
- HubSpot states an MQL-to-SQL conversion range of 10% to 20% on its own blog, without publishing the study or sample behind it. Treat it as a rough anchor, not a benchmark to be held to.
- The response time owed once a lead crosses into SQL is a separate problem from the definition. We cover the mechanics in a dedicated post on speed to lead.
What is the difference between an MQL and an SQL?
An MQL (Marketing Qualified Lead) is a lead marketing believes is worth a salesperson's time, based on behaviour, fit data, or both. An SQL (Sales Qualified Lead) is a lead a salesperson has actually reviewed, usually by talking to them or reading their record, and accepted as worth pursuing. The difference is not seniority or intent, it is who last touched the record and what they were willing to sign off on. Marketing can generate volume. Only sales can generate acceptance.
The practical version: an MQL is a guess with evidence behind it. An SQL is a guess a human has already checked.
How this comparison was made
Every claim below is checked against the vendor's own documentation, help centre, or blog, fetched directly on 29 September 2026. Where a vendor states a number without showing its source, that limit is named rather than repeated as fact. Salestruct has no commercial relationship with HubSpot, Salesforce, Pipedrive, Zoho or Adobe. Full sourcing is in the Sources section.
Where MQL and SQL actually live in each CRM
| Platform | Native MQL/SQL field | What you build instead | Source |
|---|---|---|---|
 | Yes. "Marketing Qualified Lead" and "Sales Qualified Lead" are default values on the Lifecycle Stage property. | Nothing required for the label itself. The work is defining the criteria that trigger the stage change. | HubSpot Knowledge Base, "Use lifecycle stages," last updated 17 Jul 2026, fetched 29 Sep 2026 |
 | No. Account Engagement (formerly Pardot) has Score and Grade fields, not MQL/SQL labels. | A scoring and grading model, then an automation rule on the combination of the two that you name yourself. | Salesforce Trailhead, "Get Started with Lead Grading," fetched 29 Sep 2026 |
 | No. | A custom field to hold a score or a stage label, plus a custom filter and an automation to act on it. | Pipedrive Blog, "Marketing qualified lead: a practical guide for SMBs," dated 23 Sep 2026, fetched 29 Sep 2026 |
 | No. | A custom field, wired into a Blueprint (Zoho's staged-process tool), so a record cannot advance until it is filled in. | Zoho CRM Tutorials, "Blueprint for Lead Qualification," fetched 29 Sep 2026 |
| Adobe Marketo Engage | No. | A person-scoring program combining behavioural and demographic points, released past a threshold you set. | Adobe Experience League, "Build Person Scoring Models for Marketo Engage Programs," fetched 29 Sep 2026 |
The pattern holds across every platform that does not ship the field: you are always building two things, a way to measure the lead and a rule that acts on the measurement. The vendors differ in how much of that they hand you pre-built.
HubSpot: the one platform that ships the label
HubSpot's Lifecycle Stage property comes with a fixed list of values, and "Marketing Qualified Lead" and "Sales Qualified Lead" are two of them, sitting between "Lead" and "Opportunity." HubSpot's own knowledge base describes the stage as one a contact moves through in order, set either manually or by a workflow you build on top of properties like lead score, form fills, or page views.
That is the entire native contribution: a place to record the answer. HubSpot does not ship a default definition of what makes a lead marketing-qualified. You still decide the criteria (a demo request, a pricing-page visit plus three email opens, a specific job title) and build the workflow that flips the property. The value of the native field is that everyone in the account, sales and marketing, is looking at the same property, with the same five allowed values, instead of five different homemade tags.
Salesforce: score and grade, built separately, combined on purpose
Account Engagement (Salesforce's marketing automation product, the one most people still call Pardot) does not have an MQL field. It has two separate fields, Score and Grade, and Salesforce is explicit in its own Trailhead training that neither one alone is a qualification decision.
Score is a number that goes up with engagement: opening emails, visiting pages, filling forms. Grade is a letter, A+ to F, based on how well the lead's profile fits your ideal customer on criteria you weight yourself, commonly company size, industry, geography, job title and department. Salesforce's own default grade for a prospect with no profile data at all is a D, which is the platform quietly telling you that an ungraded lead should not be treated as qualified by default.
Salesforce's own worked tutorial ("Get Cloudy") sets up an automation rule that only assigns a prospect to a sales rep once they clear both a score above 150 and a grade above B+, and separately states a starting rule of thumb of score above 100 and grade B- or better for teams building their first model. The reason for combining the two is stated directly in the training: a high score with a bad grade is someone very engaged who is a poor fit, and a good grade with no score is someone who fits the profile but has shown no real interest yet. Either one alone produces false positives.
Pipedrive: field plus filter plus automation, no packaged scoring engine
Pipedrive has no native MQL status. Its own blog walks through building the equivalent yourself: set up a custom field to hold a lead score, apply a custom filter or automation to tag leads as MQLs once they cross your threshold, and use workflow automation to move the lead into a pipeline stage or notify the assigned rep (Pipedrive Blog, "Marketing qualified lead: a practical guide for SMBs," 23 Sep 2026). The platform gives you the parts, not the assembled machine.
The same post's table of common MQL mistakes names a misaligned definition between marketing and sales first, alongside relying on outdated criteria and an unclear handoff process. All three are configuration failures a team creates, not something the product forces on you.
Zoho CRM: qualification as a Blueprint stage
Zoho's own tutorial content builds lead qualification as a Blueprint, Zoho's tool for enforcing a staged process with mandatory fields at each transition. Zoho's own worked example configures a Blueprint around a "Call Status" field that walks a rep through a scripted call as a lead moves toward a booked demo, and the record cannot advance to the next stage until required fields, in that example the demo time and address, are filled in. The mechanism itself is not specific to scoring fit or engagement, it is a general-purpose gate on whatever field you choose to make mandatory, so a team can point the same Blueprint at qualification data instead.
As with Pipedrive, there is no packaged score-plus-grade product here. The qualification logic is whatever you define inside the Blueprint, and the enforcement comes from Zoho refusing to let a record move stages until the required fields are filled in, which is a different kind of guardrail than a numeric threshold but solves a similar problem: it stops a record from advancing before someone has entered the data it needs.
Adobe Marketo Engage: a scoring program with a release threshold
Marketo (Adobe Experience League's current documentation calls it Marketo Engage) uses a scoring program, combining behavioural points (engagement: visits, downloads, email opens) and demographic points (fit: title, industry, company size) into a single running total per lead. Adobe's own documentation on scoring programs describes the model as intentionally two-part for the same reason Salesforce separates score and grade: behaviour alone rewards curiosity, not fit.
Once a lead's combined score crosses the threshold you set, Adobe's own tutorial recommends triggering an alert to sales rather than leaving the rep to notice the score on their own. Adobe frames scoring as an iterative exercise: the model is reviewed and adjusted as you learn which point values actually correlate with closed deals, not set once and left alone.
What a good MQL-to-SQL conversion rate actually is
HubSpot's own blog states that "MQL to SQL conversion rate typically ranges from 10% to 20% across industries," with higher and lower figures cited for specific sectors. We could not find a published study, sample size, or methodology behind that range on HubSpot's site. Treat it as a rough industry anchor from a vendor with a large customer base to observe, not a verified benchmark. If you want a number to hold your own team to, measure your own conversion rate for two full quarters before you set a target, because the range depends entirely on how strict your MQL definition already is: a loose definition inflates MQL volume and drags the conversion percentage down without anything about your pipeline actually changing.
Who sets the criteria, and who owns them when they change
The criteria that separate an MQL from an SQL are not a marketing decision handed to sales, or the reverse. Pipedrive's own guidance names a missing shared definition as one of the most common failure modes it sees, and Salesforce's training treats score and grade thresholds as something to review together as pipeline data comes in, not a one-time setup step.
In practice this needs one owner who is accountable for keeping the definition current, usually a RevOps function or whoever owns the CRM, with marketing and sales both able to request a change and see why the current thresholds are set where they are. We cover how to structure that decision right in who can change what in your CRM, and the broader case for a shared owner in what revenue operations actually does.
What response time do you owe once a lead crosses into SQL?
Crossing the SQL line is not the finish line, it starts a clock. How fast a rep has to respond, and what happens if nobody does, is a separate mechanical problem from the definition covered in this post, and we've written it up in full in speed to lead. If your qualification criteria are solid but leads still sit untouched after they clear the gate, that post is the one to read next, not this one.
Where the MQL/SQL model breaks
- Marketing can hit its MQL number by lowering the bar. Nothing in any of the five platforms above stops you from setting a threshold so low that most leads qualify. That does not fix the pipeline, it just moves the disagreement one stage downstream, from "is this a real lead" to "why is sales ignoring our qualified leads."
- A score without a grade rewards curiosity, not fit. Salesforce, Marketo and any homemade scoring model built on Pipedrive or Zoho face the same risk: a prospect who reads everything you publish scores high whether or not they could ever buy from you. Every platform in this post that has a real scoring concept (Salesforce, Marketo) pairs it with a separate fit measure for exactly this reason.
- Grading only works if the fit data actually exists. Salesforce's own default grade for an empty profile is a D. If your sales team is not capturing job title, company size or industry on the leads coming in, every record defaults to the same low grade regardless of how good the lead actually is, and the grade stops meaning anything.
- A scoring model decays. Adobe's own documentation treats scoring as something to review and adjust as you get more closed-deal data, not something you configure once. A threshold that made sense on day one will misclassify leads a year later if nobody has looked at it since.
- The label does not replace the handoff. None of the five platforms guarantee that a rep actually acts on a lead once it is marked SQL. The field changing is necessary and not sufficient. The rest is process, ownership, and the response-time question covered above.
Verdict by situation
| Your situation | What to do |
|---|---|
| You already run HubSpot | Use the native Lifecycle Stage values for MQL and SQL. Do not build a parallel custom field. Spend your effort on the criteria that trigger the stage change, not on the label. |
| You run Salesforce with Account Engagement | Build a score and grade model using Salesforce's own default grading criteria as a starting point, and set your first automation rule at score above 100 and grade B- or better, tightening it as data comes in. |
| You run Pipedrive | Add a custom field for score or qualification status, a saved filter, and an automation on top of it. Agree the definition with sales in writing before you build the filter, since that is the most common failure Pipedrive itself documents. |
| You run Zoho CRM | Build the qualification stage as a picklist field inside a Blueprint, and use the Blueprint's mandatory-field enforcement to stop a lead advancing without the data needed to judge it. |
| You run Adobe Marketo Engage with no strong CRM-side opinion | Build a person-scoring program combining behavioural and demographic points, and treat the release threshold as a setting you revisit, not a constant. |
| None of the above fit yet | The label does not matter until you can name who reviews a lead, which field they change, and how fast a rep is expected to act once they do. Fix that first. |
FAQ
Common questions
An MQL and an SQL are only useful if the field, the threshold, and the response time behind them are all built on purpose.
Sources
HubSpot, Use lifecycle stages in HubSpot (2026)
HubSpot, MQL vs. SQL: What they are and how they differ (2025)
Salesforce, Get Started with Lead Grading (Trailhead) (2026)
All 7 sources and how they were checked
Salesforce, Support a Lead Qualification Model with Account Engagement (Trailhead) (2026)
Pipedrive, Marketing qualified lead: a practical guide for SMBs (2026)
Zoho, Blueprint for Lead Qualification (2026)
Adobe, Build Person Scoring Models for Marketo Engage Programs (2026)
Every figure and mechanism on this page was read from the vendor's own documentation, knowledge base, or blog on 29 September 2026 and is linked above. Vendors update these pages; if a detail here disagrees with the vendor's own page on the day you read it, the vendor's page is right. Salestruct has no commercial relationship with HubSpot, Salesforce, Pipedrive, Zoho or Adobe.
If you want a second pair of eyes on your own setup, Salestruct runs a free diagnostic.
Read next.
All articlesCRM governance: who is allowed to change what?
Most teams have never answered this, and assume the CRM is keeping a record anyway. On HubSpot Starter and Professional, it is keeping a record of logins and almost nothing else.
What is revenue operations? RevOps explained
RevOps is what happens when marketing, sales and customer success stop reporting from three different spreadsheets. What it owns, how it differs from sales ops, and the signs a company needs one before it has the headcount to hire one.
