A marketing qualified account (MQA) is an account — not a person — that has demonstrated enough engagement and fit to justify sales attention. The MQA replaces the marketing qualified lead (MQL) as the primary handoff in account-based motions, because B2B buying decisions are made by committees, not by individuals filling out forms. If your team still operates on MQLs and wonders why the sales floor treats marketing's "leads" as noise, the MQA is the upgrade path.
Full disclosure: Abmatic AI builds account-level scoring and MQA workflows as core product. We have an opinion that account-based qualification is more accurate than person-based qualification for most B2B motions. The mechanics in this guide work regardless of which platform implements them.
A marketing qualified account is an account that meets a fit threshold (it matches your ICP) AND an engagement threshold (it has interacted meaningfully with your properties or signals). The combination — fit AND engagement — is the thing. Either alone is not enough. Fit-only is a target list. Engagement-only is anonymous traffic. The MQA is the intersection that earns sales attention.
The shift from MQL to MQA tracks the shift in B2B buying. A single VP downloading a whitepaper does not represent intent from the buying committee. Three people from the same account — a VP, a director, and an IC — engaging across pricing, comparison, and case-study content over two weeks does represent intent. The first is an MQL with a low conversion rate. The second is an MQA with materially higher pipeline value per public ABM-program reports.
See how Abmatic identifies and routes MQAs →
MQL frameworks emerged when B2B buying was modeled as a single-decision-maker process. A "lead" was a person; the marketing job was to attract and qualify enough of them; sales took the qualified ones forward. The framework worked when one person actually made the decision.
That stopped being how B2B works some time ago. Modern enterprise software purchases involve buying committees of multiple stakeholders, multi-quarter evaluation cycles, and decision processes where no single individual's form fill predicts the outcome. The MQL persisted as a metric anyway, because frameworks are sticky and dashboards measure what they always measured.
Teams that have run side-by-side analyses of MQL-driven motions versus MQA-driven motions consistently see materially different conversion patterns per public customer reports. The MQA captures the multi-stakeholder, multi-touchpoint reality that MQL frameworks structurally cannot. The MQL counts an individual; the MQA counts the buying committee's collective signal.
MQL handoffs are also sales-floor-expensive. SDRs receive a list of "qualified leads," call them, find that most are not in active evaluation, learn to deprioritize the queue, and let genuine signals get lost. MQA handoffs — "this account has three engaged stakeholders, two pricing-page visits, and a VP-level title in the buying committee" — come with enough context that the SDR's first call is meaningful.
An account becomes an MQA when it crosses thresholds across three dimensions, not one.
The account matches the firmographic, technographic, and segment profile of accounts your business serves well. Industry, revenue band, employee count, geography, tech stack. If the account does not fit, no amount of engagement makes it an MQA — it is a noisy non-customer.
The account has demonstrated meaningful interaction across owned properties — pricing visits, demo requests, comparison-page engagement, content downloads, product trials, email engagement. Engagement should be measured at the account level (people from the account, weighted by recency), not at the individual level.
The engagement is distributed across the buying committee, not concentrated in a single low-influence person. A pricing-page visit from a VP, a comparison read from a director, and a case-study consume from an IC together demonstrate committee-level interest. One IC visiting three pages does not.
Build the ICP scoring rules — firmographic and technographic criteria that map to your best-fit customer profile. Score every account in your CRM and target list against the ICP. Set a fit-threshold cutoff below which engagement does not earn MQA status.
Pick the events that count. Weight them by intent strength (a pricing-page visit weighs more than a homepage visit). Apply time decay (recency dominates). Roll up to account-level engagement scores using a rollup logic that fits your motion (sum, max, weighted average).
Compute the unique-stakeholder count and the seniority distribution of engaged people per account. Boost the score for accounts with multi-stakeholder engagement; cap or down-weight scores from single-person engagement bursts.
Tune the threshold against historical pipeline and conversion data. The threshold that maximizes pipeline-per-MQA, not MQA volume, is the right one. Most teams find their first cut too generous — tighten it.
An MQA handoff is not a Slack message that says "Acme Corp is now MQA." It is a packet: account name, engaged stakeholders with titles, recent activity timeline, recommended outreach motion, the AE the account is assigned to. The richer the packet, the higher the SDR-to-meeting conversion rate.
Track every MQA through the funnel. SQL conversion. Opportunity creation. Pipeline. Closed-won. The loop is what makes the model improvable. Without it, the threshold drifts and the team trusts the score less over time.
The framework is different, not just the unit of analysis. MQA logic that simply rolls up MQL counts to the account misses stakeholder coverage, fit-engagement intersection, and time-decay discipline.
"We need 200 MQAs a month for sales" is a marketing-output target. Using it to set the threshold backwards-engineers a model that produces 200 noisy accounts. Set the threshold by quality and let the volume be the output.
Engagement-only MQAs include accounts that are not actual customers. Sales burns time disqualifying. The fit gate is non-negotiable.
Single-person-engagement MQAs are not buying-committee MQAs. They are interested individuals, sometimes on personal-research missions. The stakeholder-coverage requirement separates committee signal from individual curiosity.
Inbound, outbound, customer-expansion, and partner-co-sell motions need different MQA models. The thresholds that work for inbound do not work for outbound; the engagement signals differ.
MQAs that are not tracked through to pipeline and closed-won outcomes degrade over time. Without measurement, the threshold cannot improve.
| Term | Unit of analysis | Owner | Question it answers |
|---|---|---|---|
| MQL | Person | Marketing | Did this person engage enough to deserve sales follow-up? |
| MQA | Account | Marketing | Is this account engaged enough across the buying committee to deserve sales attention? |
| SAL (Sales Accepted Lead/Account) | Person or Account | Sales | Has sales accepted the handoff and committed to follow up? |
| SQL (Sales Qualified Lead) | Person | Sales | Has the conversation happened and is the person actively considering? |
| SQO (Sales Qualified Opportunity) | Account | Sales | Is this account a real opportunity in the pipeline with a forecast? |
Most modern stacks use MQA for marketing-side qualification and SQO for sales-side qualification, with SAL as the handoff confirmation step. MQL and SQL persist where the underlying motion is still person-led; they coexist with MQA and SQO where the same team runs both inbound-individual and account-based motions.
An account that meets your ICP fit threshold AND has demonstrated meaningful engagement across the buying committee, qualifying it for sales attention.
An MQL is a person; an MQA is an account. The MQL framework treats individual form-fills as the qualification unit; the MQA framework treats committee-level engagement as the qualification unit. For account-based motions, the MQA is structurally more accurate.
Most enterprise and mid-market B2B teams should. SMB and product-led-growth motions where buying decisions are genuinely individual still have a place for MQL frameworks. The hybrid — MQA for enterprise, MQL for self-serve — is increasingly common.
Backtest against historical pipeline and conversion. Pick the threshold that maximizes pipeline-per-MQA, not MQA volume. Tighten until the field team trusts every account in the queue.
Yes. Abmatic builds the fit model from your ICP definition, captures engagement at the account level, computes stakeholder coverage, applies time decay, and routes MQAs to your CRM with the context packet sales needs.
First-party behavioral (web, product, email), CRM activity (deal stages, meeting attendance), and ideally third-party intent corroboration. Negative signal (unsubscribe, lost-deal) should pull scores down, not just stop adding.
Quarterly at minimum. The threshold and signal weights should be reviewed against pipeline outcomes; drift in the customer ICP, content mix, or buying motion should trigger out-of-cycle reviews.
To make the qualification dimensions concrete, here is a sketch of how an MQA threshold might compose. Numbers are illustrative — tune to your data.
An account in your tier-2 list, ICP-fit score 82, in week 6 of the quarter:
The same account two weeks earlier, with only the VP visit and one case-study read, would have engagement of 35, coverage of 30, and would not have crossed. Time and stakeholder breadth made the difference. That sequencing matters — the model captures movement toward purchase, not just static activity volume.
The MQA is the right qualification unit for B2B account-based motions. Fit AND engagement AND stakeholder coverage. Time-decayed. Threshold-tuned to pipeline outcomes, not output targets. Handed off with context, not just an account name. The teams that have moved from MQL-driven to MQA-driven motions consistently see better sales adoption and higher pipeline-per-handoff per public reports.
If you want to see what an MQA model looks like running on your own data — with your ICP, your engagement events, your CRM connected — book a 30-minute Abmatic demo. We will walk through the model live and show how the handoff packet reaches the SDR.