On this page
What are MQL and SQL, and how do they differ?
A marketing-qualified lead (MQL) is a contact or account record that has passed marketing’s declared prequalification screen for target-customer fit and recorded engagement, making it eligible for human qualification . A sales-qualified lead (SQL) is a contact or account that a sales development representative (SDR) or sales representative has verified, through direct background research and two-way dialogue, as ready to enter an active sales conversation and opportunity evaluation .
In academic marketing and sales research, the two labels correspond to two distinct screening steps that sit on either side of the boundary between marketing and sales . Sabnis et al. (2013, p. 55)4 define the upstream screen in their study of B2B lead follow-up: “Prequalification is a process by which the marketing department screens the leads it has generated and discards those it deems unattractive.” Terho et al. (2023, pp. 343, 344)5 examine the downstream screen inside eight B2B firms and trace how a specialized sales development function researches, engages, and qualifies leads before handing them over to outside sales as sales-qualified leads . As a Global Inside Sales Lead in their study stated (p. 343): “What a sales qualified lead looks like can be customized for the company […] but always have qualification criteria.”
Treating MQL and SQL as properties of the buyer rather than as rules of the selling organization is the primary source of confusion in pipeline reporting . The same buyer inquiry can be an MQL in a firm that passes every target-industry webinar attendee to an SDR queue, a rejected inquiry in a firm that requires an explicit demo request, or an SQL only after an SDR confirms that “the lead is a person who can carry out conversations with sales” (Terho et al., 2023, p. 344)5 .
Three adjacent concepts sit beside MQL and SQL and should be kept distinct:
- Sales-accepted lead (SAL): An intermediate operational state used in many marketing-to-sales handoffs, in which sales or sales development formally acknowledges receipt of an MQL within a declared service-level agreement (SLA) and agrees to attempt contact before deciding whether the record qualifies as an SQL .
- Product-qualified lead (PQL): A user or workspace in a self-serve or freemium product that has crossed an observed product-usage threshold rather than a marketing-content threshold; see what a product-qualified lead is .
- Lead scoring and qualification checklists: Lead scoring is the rule-based or model-based ranking instrument often used to set the MQL threshold, whereas BANT and MEDDIC are structured checklists used during discovery and opportunity management at and after the SQL gate .
Where do MQL and SQL sit in the B2B acquisition funnel?
Practitioners invented the compound labels marketing-qualified lead and sales-qualified lead because the unmodified word lead has no stable position in the B2B funnel literature . D’Haen and Van den Poel (2013, p. 545) review the customer-acquisition literature and note the contradiction directly: “A main difference, however, is where the studies place a prospect and a lead in the sales process”, with one group of studies placing the prospect before the lead and another group placing the lead before the prospect .
In the four-stage framework of D’Haen and Van den Poel (2013, p. 545), the funnel begins with suspects, defined as “Suspects are all potential new customers available.” Screening suspects against firmographic or profile rules yields prospects: “Prospects are suspects who meet certain predefined characteristics.” Because marketers face large company lists, “As a result, marketers often make selections using a set of arbitrary rules.” (p. 545) Only after a second screen do prospects become leads: “Leads are prospects that will be contacted, after they have been qualified as the most likely to respond. This qualification is often driven by gut feeling or self-claimed competence.” (p. 545) Finally, “leads who become clients of the company are customers.” (p. 545)
Monat (2011, pp. 179, 181)3 uses the same progression when he writes that several industrial frameworks “discuss the sales pipeline beginning with suspects and then progressing through prospects to qualified leads”, while defining a lead as a recorded expressed interest in a firm’s goods or services, “irrespective of whether that interest is from a new prospect or from an existing customer” . Jolson (1988, p. 191)2, by contrast, treats the incoming response card or phone inquiry as the raw lead and calls the screened contact a qualified prospect: “Based on the judgment or intuition of the salespeople, qualified prospects are people or organizations with needs or desires that can be satisfied by the seller’s products or services” .
Why B2B teams split “lead” into MQL and SQL across the handover
Figure 1Why B2B teams split "lead" into MQL and SQL across the handover
Source: Diagram from this essay. Sources and interpretation are given in the article.
Figure 1 is this guide’s synthesis of the terminology in D’Haen and Van den Poel (2013, p. 545), Jolson (1988, p. 191)2, Monat (2011, pp. 179, 181)3, Sabnis et al. (2013, p. 55)4, and Terho et al. (2023, pp. 343–347)5 . When one department uses lead to mean any captured email address (Tradition B) and another uses lead to mean a vetted account ready for a field rep’s calendar (Tradition A), handover meetings turn into arguments over vocabulary . Terho et al. (2023, p. 346) report this exact failure mode across their case firms: “units should clearly define and agree on the terminology “what is a lead, a sales qualified lead or a prospect,” as well as clearly define responsibilities who has the ownership of the lead along the sales process.”
How does sales development move a record from MQL to SQL across selling contexts?
In a small sales organization, the same account executive may review marketing inquiries, research the company, place the first qualifying call, and run the closing presentation. In specialized B2B sales organizations, an inside sales or sales development team sits between marketing and field sales to qualify records to the SQL threshold .
Terho et al. (2023, pp. 337, 341–345)5 studied this function through nine one-to-one interviews and a seven-person focus group with 13 informants across eight B2B firms (four SaaS providers, an industrial measurement manufacturer, a security and IT services enterprise, a management consultancy, and an advertising agency): “The authors adopted a qualitative field-study approach and interviewed 13 people from eight firms.” Across all eight firms, the sales development process followed three steps: researching the leads against ideal customer profile (ICP) fit and pain points, engaging leads across digital and telephone channels, and end qualification and handover to outside sales .
Crucially, however, what an MQL looks like at entry and what an SQL requires at handover depended on two dimensions (Terho et al., 2023, p. 343)5: “The data analysis further pointed out four key types of sales development that notably differ by the nature of the incoming leads and the complexity of the lead management process, as illustrated in Figure 2.” Specifically, “an SDR can pick up “inbound leads” generated by the marketing function or focus on sales function–directed “outbound prospecting” based on predefined criteria.” (pp. 343, 344) And “the activities of an SDR can focus on “individual prospects” or “larger enterprise accounts”, depending on the complexity of the sales situation.” (p. 344)
Four sales development paths to a sales-qualified lead
Figure 2Four sales development paths to a sales-qualified lead
Source: Diagram from this essay. Sources and interpretation are given in the article.
Figure 2 adapts the four-context typology from Terho et al. (2023, Figure 2, p. 342, and pp. 343–345)5 and shows why a single company-wide checklist rarely works for both transactional and enterprise motions :
- Inbound prospect-focused sales development: Marketing generates a high volume of warm inbound inquiries (MQLs) for a relatively simple offering. Speed and role verification dominate. As one interviewee explained, SDRs screen the contact’s marketing-automation and CRM history and ask (Terho et al., 2023, p. 344)5: “We tried to find out if it’s a case where we can sell something right now or maybe six months down the line” , while checking whether the inquirer is the right person or can name the right contact. Handover to SQL happens “once the SDR verifies that the lead is a person who can carry out conversations with sales” (p. 344) .
- Inbound account-focused sales development: Marketing captures content downloads or website visits from multiple employees inside a large enterprise account. Instead of treating each download as an isolated single-person MQL, sales development examines why the account’s buying committee is researching the topic, reaches out across stakeholders, and hands the record over as an SQL only “Once SD deems that the customer, as a decision-making unit, is ready to start a dialogue with the account executive, the lead is handed over to outside sales.” (p. 345)
- Outbound prospect-focused sales development: There is no upstream MQL from marketing; SDRs start from a cold list of target companies in markets where buyers rarely respond to cold field calls, verify ICP fit, run multi-touch sequences, and hand the contact over as an SQL once the prospect agrees to a sales conversation (pp. 343, 344) .
- Outbound account-focused sales development: Sales defines major target accounts, and SDRs spend months mapping business units and engaging stakeholders: “Complex purchase decisions require prolonged engagement of multiple and varied customer stakeholders at the account level rather than at the individual prospect level” (p. 344) . At SQL handover, the account executive receives not just a calendar invite, but a briefing on how the decision-making unit engaged, what content they read, and what insights emerged (p. 345) .
Across all four contexts, Terho et al. (2023, pp. 346, 347)5 found that the MQL-to-SQL handover breaks down without three organizational mechanisms: backward-calculated joint targets (“Many respondents emphasized joint target setting by working backward from sales targets to understand how many leads marketing needs to generate monthly and what the quality of those leads needs to be so that the sales funnel maintains a steady flow.”, p. 346), formal SLAs (“Many firms further used internal service-level agreements to formally govern the internal lead management process, for example, regarding the contact times for different leads”, p. 346), and an SQL reward metric tied to verified pipeline rather than raw meetings . Without clear ownership rules, salespeople argue over credit (“this was my lead, no I talked to them first, no it was mine”, p. 346) . And when SDRs are paid solely on meetings booked, they push premature records across the SQL line; as the Global Inside Sales Lead warned (p. 347): ““Booked meetings” is a classic metric, however, in my opinion a wrong one. It would be more important to think about the lifecycle of the client: how it was transformed from a subscriber to an opportunity.” Several case firms therefore checked “did the field sales enter a new sales opportunity into the firm’s CRM system after holding a sales meeting booked by an SDR” (p. 347) before crediting the SQL handover .
Why does the MQL screen govern whether sales reps follow up on marketing leads?
Even when marketing generates thousands of leads, sales representatives decide every week how much of their scarce selling time to spend on marketing’s MQLs versus prospecting on their own . Sabnis et al. (2013, pp. 52–67)4 explain why this tension is structural: marketing, with its longer-term perspective, likely considers “leads that demand careful nurturing to be more attractive than sales does, because sales reps must focus on short-term quota attainment” (p. 54) . Consequently, “the leads that marketing views as most attractive may not meet the immediate needs of the sales rep.” (p. 54)
Because “Sales reps cannot observe the objective quality of leads” before investing time to call and qualify them, they rely on their perception of marketing’s prequalification screen (Sabnis et al., 2013, p. 55)4 . Sabnis et al. (2013, pp. 58, 59) surveyed 2,666 sales representatives across four long-established B2B firms (51 reps in scientific instruments, 30 in chemicals, 100 in copiers, and 280 in computers), obtaining 500 usable responses and analyzing 461 reps after removing 39 respondents (37 of them at the computer firm) who spent zero time on marketing leads . Across all four firms, compensation was tied to quota achievement, commissions, and top-line bonuses: “Although the compensation schemes included bonuses tied to acquiring new customers, no component explicitly rewarded following up on marketing leads.” (p. 59)
In that 461-rep sample (Table 1 Panel B, p. 59), representatives allocated on average just 5.7% of their working time (0.057, SD = 0.091) to following up on marketing leads, 15.6% (0.156, SD = 0.167) to pursuing self-generated leads, and 78.8% (0.788, SD = 0.196) to non-acquisition work such as customer retention and administrative tasks . The authors’ opening mention on p. 52 that 70% of marketing leads are never followed up comes from two cited practitioner reports from 2002 and 2009, not from this 461-rep survey . Within customer-acquisition time alone, reps spent nearly three times as much of their time on leads they found themselves as on leads marketing sent them .
Figure 3Determinants of marketing-lead follow-up time share across 461 B2B sales representatives
Perceived MQL prequalification quality (0.64) and its interaction with experience (0.71) pull reps toward marketing leads; managerial tracking (-0.89) and its interaction with experience (-0.55) push the other way, while lead volume (0.44) crosses zero.
Source: Sabnis, Chatterjee, Grewal, and Lilien (2013), Table 2, p. 61; 461 B2B sales representatives in four firms.
Figure 3 plots the posterior parameter estimates and 95% Bayesian credible intervals from the Dirichlet component regression on the marketing-lead time share in Sabnis et al. (2013, Table 2, p. 61)4 . Three findings govern how an MQL definition performs in practice:
- Prequalification quality drives follow-up (
b = 0.64, 95% interval0.39to0.85): When sales representatives perceive that marketing screens leads rigorously before passing them over, they allocate a significantly larger share of their working time to marketing leads and a smaller share to self-generated leads (b = -0.38, p. 61) . That positive association is even stronger among more experienced reps (b = 0.71, 95% interval0.48to0.96) . - Sending more leads without tightening the screen does not move follow-up (
b = 0.44, 95% interval-0.28to0.73): The main effect of marketing lead volume has a credible interval that crosses zero, and its aggregate elasticity in Table 3 (p. 63) is-0.07. Flooding the CRM queue with loosely screened MQLs does not increase the share of time reps spend working them . - Managerial tracking goes with less follow-up, not more (
b = -0.89, 95% interval-0.96to-0.68): Across the pooled sample, close managerial monitoring of lead follow-up is associated with a smaller share of time spent on marketing leads, and the association is stronger among more experienced salespeople (b = -0.55, interval-0.74to-0.39) . The authors had predicted the opposite and read the result as reactance against behavior control, which sales reps paid on outcomes are likely to resist (pp. 61, 62) .
The firm-level elasticities in Table 3 (p. 63) point the same way: “The elasticity is positive across all four firms but varies considerably (from .58 to 1.64).” Across the pooled sample, “a 1% increase in perceived prequalification quality produces a 1.12% increase in marketing lead follow-up time allocation” (p. 62) . The lead scoring guide linked above charts those elasticities firm by firm.
What happens when a team sets the MQL or SQL gate too tight or too loose?
If sales representatives respond to a stricter prequalification screen by taking passed leads more seriously, why not tighten the MQL and SQL gates until only near-certain buyers pass? Because every real-world qualification screen is noisy, and tightening an imperfect gate trades a higher conversion rate among passed leads against lost sales among rejected leads .
Two of the held studies measure that trade directly, and the BANT vs MEDDIC guide linked above works through both in detail. In Jolson’s (1988, pp. 193–195)2 1987 field test at a regional security-system installer, two direct-mail offers of 20,000 packets each brought in 104 tightly qualified and 622 loosely qualified leads . The tight leads closed at 30.8% against 7.7% for the loose leads, yet the loose arm produced 48 sales and $95,320 against 32 sales and $64,200, because “The T sales group seemed to run out of places to go” (p. 194) . Monat (2011, pp. 189–191)3 applied his lead scorecard to 324 randomly selected leads at a US computer hardware and software firm, with data for only five of its eight purchase determinants and equal weights . At his chosen cutoff of -0.1, the 133 passed leads booked at 28.6%, but the 191 rejected leads still booked at 10.5%: 20 of the 58 bookings in the sample (34.5%) .
A third study shows what the record of those decisions is for. D’Haen and Van den Poel (2013, pp. 548, 549) had a US telecom provider’s sales force evaluate the top 356 prospects that a model ranked from “more than 16 million U.S. companies”: “Of these, 56 companies were qualified as good leads, corresponding to a conversion rate from prospect to lead of 15.73%” . Because “Information on decisions made, steps taken, strategies employed, and so on must be recorded and analyzed periodically.” (p. 545), the 56 accepted and 300 rejected evaluations become the training labels for the next round of the model .
This trade-off explains why neither marketing nor sales can set the MQL and SQL thresholds in isolation . When marketing is rewarded on raw MQL count, it has a reason to loosen the screen, yet in Sabnis et al. (2013, pp. 61, 62)4 more marketing leads did not go with a larger share of reps’ time on them, while a screen that reps rated highly did . Conversely, when sales demands a gate so tight that only buyers with immediate budget and urgency pass, the funnel starves selling capacity (Jolson, 1988, p. 194)2 and discards accounts that would have converted (Monat, 2011, p. 191; Sabnis et al., 2013, p. 53, noting from a 2006 practitioner report that “disagreements about classifying leads can cause sales reps to discard customers that would be highly likely to buy in the future”)3 .
What does the published evidence show, and what does it leave open?
Every study in this synthesis bounds its own scope explicitly, and none of the five provides a universal MQL score, a universal MQL-to-SQL conversion benchmark, or a causal field experiment on changing a firm’s SQL definition .
- Terho et al. (2023, pp. 342, 350)5 note that their qualitative field study of 13 interviewees in eight firms aims at analytical rather than statistical generalization (“In terms of generalizability, we have striven for analytical rather than statistical generalizability.”, p. 342) and that “First, this study is explorative and builds on qualitative data from in-depth interviews with managers.” (p. 350) .
- Sabnis et al. (2013, pp. 63, 64, 67)4 emphasize that their cross-sectional survey of 461 reps in four established B2B firms measures self-reported shares of working time rather than actual hours, lead counts, or lead-level conversions: “Lead-level data from firms’ customer relationship management systems might make such analyses possible.” (p. 64) They also bound their sample to mature industrial sellers: “our conclusions might not apply to smaller or younger firms whose sales processes are still evolving” (p. 67) .
- D’Haen and Van den Poel (2013, p. 550) tested only the first prospect-to-lead screening step with sales reps at one US telecom company and ran two pseudo-tests on reference data: “The main limitation of this study is that it was not possible to run a full, real-life test of the algorithm.”
- Monat (2011, p. 191)3 coded 324 archived email leads at a single computer hardware and software seller using equal weights and an in-sample threshold: “It did not use hold-out samples for validation” .
- Jolson (1988, pp. 191, 195, 196)2 tested two direct-mail offers with six inexperienced reps selling one security product in one metropolitan area, noting that in field practice “A lead is highly qualified only if the salesperson thinks it is.” (p. 191) and concluding that “the current results are surely not generalizable” (p. 196) .
| Study | Sample and setting | What it shows for MQL and SQL design | Where it stops |
|---|---|---|---|
| Terho et al. (2023) | 13 interviewees across 8 B2B firms (9 interviews and a 7-person focus group) | Three SDR steps (research, engagement, end qualification and handover) across four contexts (inbound or outbound by prospect or account); links alignment to shared MQL/SQL definitions, backward joint targets, SLAs, and CRM opportunity checks | Explorative qualitative field study; analytical rather than statistical generalizability |
| Sabnis et al. (2013) | 461 B2B sales reps across 4 established firms (500 usable responses; 39 zero-marketing-time reps excluded) | Reps spent 5.7% of time on marketing leads vs 15.6% on self-generated leads; perceived prequalification quality raised marketing-lead time share (b = 0.64, elasticity 1.12), while managerial tracking reduced it (b = -0.89) | Cross-sectional survey of time shares and perceptions in four mature firms; no lead-level CRM conversion data |
| D’Haen and Van den Poel (2013) | US database (>16M firms); 1 US telecom firm (107 of 389 customers as input; 356 prospects evaluated) | Documents prospect/lead order reversal across literature; sales reps qualified 56 of top 356 ranked prospects as good leads (15.73%); iterative loop records accepted and rejected decisions | No full multi-phase real-life test of the algorithm; funnel averages quoted on p. 545 come from a 2004 practitioner book |
| Monat (2011) | 324 random leads (58 bookings, 266 non-bookings) from ~750 leads at 1 US computer firm between 2003 and 2005 | At cutoff -0.1, 133 passed leads booked at 28.6% (38 bookings), while 191 rejected leads booked at 10.5% (20 bookings, or 34.5% of all bookings) | Single firm, retrospective email coding by one analyst, five of eight determinants coded with equal weights, in-sample cutoff with no hold-out validation |
| Jolson (1988) | 40,000 mailers (20,000 tight, 20,000 loose) to small businesses; 6 new reps in 2 teams at 1 security firm | Tight arm: 104 leads, 32 sales (30.8%, 300 lead cost per sale); loose arm: 622 leads, 48 sales (7.7%, 200 lead cost per sale); tight team ran out of leads | Single security product, one metropolitan region, six inexperienced reps; author states results are not generalizable |
Figure 4MQL and SQL evidence map: what each held B2B study measures and where it stops
Each held study examines one link in the MQL-to-SQL chain, from marketing prequalification and SDR context to tight-versus-loose thresholds. None supplies a universal conversion benchmark.
Source: the five held studies as cited; the limits column is this guide's reading.
How should a revenue team specify and audit its MQL and SQL gates?
Because both an overly loose screen and an overly tight screen carry measurable costs, a revenue operations team should treat its MQL and SQL definitions as a testable operating contract between marketing, sales development, and sales rather than a static CRM picklist . What follows is this guide’s operational synthesis and has not been tested :
- Declare the context and unit of record before writing the rule: Following Terho et al. (2023, pp. 342–345)5, separate single-prospect inbound flows from multi-stakeholder enterprise account flows . In an inbound prospect flow, the MQL unit is an individual contact who matches the ICP and requests a commercial interaction or crosses a validated engagement threshold. In an account-based flow, individual content downloads are signals attached to the account, and the record reaches SQL when the buying unit is ready for an account executive dialogue .
- Pair every stage gate with an owner, a response SLA, and a structured return code: Following Terho et al. (2023, p. 346) and D’Haen and Van den Poel (2013, p. 545), record every acceptance and rejection decision with a mandatory reason code (
wrong persona,outside ICP,literature collection / no project,timing > 6 months,competitor locked) . - Evaluate the SQL gate on opportunity creation and bookings, not booked meetings alone: Following the warning in Terho et al. (2023, p. 347), verify whether field or inside sales enters a documented CRM opportunity after the handover meeting, and size monthly MQL and SQL targets by working backward from sales capacity and quota targets (p. 346) rather than pushing raw lead volume (Sabnis et al., 2013, p. 61)4 .
- Audit a sample of rejected records every quarter: A team that observes outcomes only on accepted MQLs and SQLs cannot see the buyers its gates turn away, which is what the rejected leads in Monat (2011, p. 191)3 and the loose arm in Jolson (1988, p. 195)2 showed . Route a periodic sample of below-threshold or SDR-returned leads into a structured nurture or follow-up check; the BANT vs MEDDIC guide sets out how to assign and compare such a sample .
Figure 5An MQL and SQL stage-gate specification and audit worksheet
One row per selling motion: the unit of record, MQL entry rule, SLA and return code, SQL handover rule, CRM opportunity check, and false-negative sample rate.
Source: Author's operational worksheet grounded in Terho et al. (2023), Sabnis et al. (2013), D'Haen and Van den Poel (2013), Monat (2011), and Jolson (1988); reader inputs only, untested.
References
- D'Haen, J., & Van den Poel, D. (2013). Model-supported business-to-business prospect prediction based on an iterative customer acquisition framework. Industrial Marketing Management, 42(4), 544–551. https://doi.org/10.1016/j.indmarman.2013.03.006
- Jolson, M. A. (1988). Qualifying sales leads: The tight and loose approaches. Industrial Marketing Management, 17(3), 189–196. https://doi.org/10.1016/0019-8501(88)90002-8
- Monat, J. P. (2011). Industrial sales lead conversion modeling. Marketing Intelligence & Planning, 29(2), 178–194. https://doi.org/10.1108/02634501111117610
- Sabnis, G., Chatterjee, S. C., Grewal, R., & Lilien, G. L. (2013). The sales lead black hole: On sales reps' follow-up of marketing leads. Journal of Marketing, 77(1), 52–67. https://doi.org/10.1509/jm.10.0047
- Terho, H., Salonen, A., & Yrjänen, M. (2023). Toward a contextualized understanding of inside sales: The role of sales development in effective lead funnel management. Journal of Business & Industrial Marketing, 38(2), 337–352. https://doi.org/10.1108/JBIM-12-2021-0596