MQL to SQL: The definition problem nobody fixes, and how to fix it

Most MQL-to-SQL fights are fights about a word nobody wrote down. Marketing calls a lead qualified because it crossed a score threshold. Sales calls it junk because nobody on the account has a timeline. Both are right, because "qualified" was never defined once, in one document, with both signatures on it.

A Gartner survey of 412 senior marketing and sales leaders found they collaborate on only three of 15 commercial activities, and 90% report conflicting functional priorities. Use a single written definition with fit, pain and timing criteria, reason-coded rejections, and a two-way SLA that your CRM enforces instead of your calendar.

The textbooks disagree before your team does

MQL and SQL definitions conflict, as do the qualification frameworks behind them:

MQL: One common definition says an MQL is "good fit but not ready to buy." Another says an MQL means marketing says the contact is ready for sales. Those are opposite statements about the same acronym in B2B SaaS.
SQL: Definitions range from fit plus ready to sales qualification as an opportunity moved into pipeline. Other versions require an estimated dollar value and close timeframe, or buying intent, decision information, budget and executive buy-in.
Qualification frameworks: Some sales development teams use BANT (budget, authority, need, timing) as their SQL definition; others use looser variants like ANUM (authority, need, urgency, money).

So if your RevOps lead learned the Marketo revenue model, your VP Sales learned MEDDIC, and your paid agency reports on HubSpot lifecycle stages, you have three definitions in the building before anyone has argued about a single lead.

Why it stays broken

Score thresholds are guesses. Points and thresholds are often assigned without propensity analysis. Downloading several whitepapers tells you almost nothing about whether someone will buy, but in a model built that way, it could easily push a contact over the threshold.
Incentives lock the guess in place. If marketing is measured on MQL volume, the cheapest path is the cheapest form fill. Under pipeline pressure, teams can game scoring thresholds. On the sales side, reps may use their own arbitrary rules to decide which leads get a call.
The feedback loop is missing. When teams don't record why a lead was discontinued or lost, marketing can't tell whether rejected leads were the wrong persona or just the wrong month. The scoring model never improves, and the argument restarts every quarter.

Together, those failures create exactly the coordination burden growth leaders face when paid, outbound and content teams optimize different definitions.

Published B2B SaaS MQL-to-SQL rates vary widely. None are sized by ACV, and nothing we've found benchmarks the 20K-100K band specifically. Use your own closed-won data cut by source and persona.

Four vendors, four definitions

The problem compounds when execution is spread across specialists. A paid agency might optimize to cost per lead and report form fills. Some outsourced SDR firms are paid per meeting; outbound SDRs tend to be measured on appointments with looser criteria than inbound SDRs, who are measured on sales-accepted leads. A content freelancer gates an ebook and the download lands in HubSpot as an MQL because the score crossed a threshold set two agencies ago.

Each vendor reports in its own dashboard against its own definition of "qualified," and the Head of Growth ends up reconciling four spreadsheets before a pipeline review. Within a quarter, pipeline reviews turn into arguments over whose leads were real. Their reports honestly reflect the metric in each contract, and nobody handed the vendors one shared definition to work from.

Write one definition, sign it twice

Pull closed-won and closed-lost from the last four quarters and look at what the winners had in common at the moment they entered pipeline. A steep MQL-to-SQL drop usually signals there's no shared definition.

Use the same ICP fields across paid audiences, outbound lists and content gates so each channel is working from the same evidence.

Define the MQL with fit plus one real signal. Fit means ICP criteria your sales leader signs: company size, industry, geography, tech stack. The signal is something a buyer does on purpose: a demo request, a pricing-page visit, a reply to outbound, a product trial that reached value.

For product-led SaaS, that trial signal is concrete usage, like a teammate invite or an API key generated. Classify a whitepaper download on its own as a nurture event and exclude it from MQLs.

Define the SQL with four lead-to-opportunity criteria: ICP fit, identified pain, engagement such as a discovery call or demo, and buying intent shown through a timeline, budget or trigger event. Forrester argues that requiring defined budget early is too aggressive for new-concept demand, and that qualification criteria should be weighted by demand type, including the buyer's ability to mobilize and prioritize resources.

Our recommended minimum bar for 20K-100K ACV is:

Fit grade met
Prospect-confirmed pain
A timing or critical-event signal
A plausible funding path
An identified champion

Together, these five checks establish a consistent opportunity gate across every channel. Fill deeper deal fields like economic buyer and paper process after the opportunity exists; exclude them from the gate.

Keep the reason-code list to five items: Unable to reach (only after SLA attempts were met), Inaccurate data, No interest/need/budget/authority, No fit, and Not ready within X months, which triggers a nurture task. Map each code to a fix: wrong persona tightens ICP scoring, company too small adds a firmographic disqualifier, bad timing refines intent signals, existing customer fixes CRM dedupe.

Sign a two-way SLA. Marketing commits to a volume; sales commits to a clock. A sensible starting point is to accept or reject within 24 hours with a reason code, make 5-8 touches over 7-10 days, record reasons on 100% of rejects, and review results weekly. Demo requests need a separate, faster clock. Auto-reassign leads that aren't touched in time, and let a rep who doesn't want a lead pass it to a hungrier one. That gives every channel and vendor one operating rule.

Sales leadership signs the SQL definition, and we'd review it monthly. SDRs who meet the SQL terms are performing as defined; low SQL-to-opportunity then points to a flawed definition.

Make the CRM enforce it

A definition without a mechanism is a memo. Configure the CRM around four controls:

Lifecycle movement: In HubSpot, automatic lifecycle updates only move a record forward. For an MQL that goes back to nurture, use a separate re-engagement status rather than moving the lifecycle stage backward.
Required rejection data: The Leads object ships with a default pipeline (New, Attempting, Connected, Qualified, Disqualified) and a Disqualification Reason property. Pipeline stage logic set to "Required" blocks a record from updating until a property has a value. Pairing Required stage logic with the Disqualification Reason property forces a reason code on every rejection.
One SLA timer: HubSpot tracks response-time metrics on the lead and on the contact separately, and they don't start at the same moment. Pick one for the SLA report.
Automatic routing: Score-to-MQL routing is a workflow. Enroll at the High band, round-robin assign and notify. Workflow-based assignment requires Sales Hub Professional or Enterprise.

These controls turn the shared definition into an operating process with one interpretation for vendors and sales reps.

In Salesforce, the Lead Status picklist and an active assignment rule handle routing. SLA timers run on a record-triggered Flow with Scheduled Paths. Account Engagement separates grade (fit) from score (activity). Calibrate both thresholds against your own closed-won data rather than a generic benchmark.

Where Understory fits

The cleanest way to keep one definition is to have as few hands as possible translating it. Understory runs LinkedIn, Google and Meta paid media campaigns. It runs signal-based outbound through Instantly, with HeyReach handling LinkedIn outreach on Clay-built lists. The signals include a recent CRO hire or funding round, or a tech-stack change. Understory also provides on-staff creative for B2B SaaS clients.

The same ICP criteria drive the ad audiences, the outbound lists and the content gates, so a form fill from paid and a positive reply from outbound mean the same thing when they hit your CRM. Your sales leader owns your SQL definition. We build to it and report against it.

If paid is already covered by another agency, we're happy to run outbound and creative alongside them against the same definition.