Why MQLs exist, and why small teams inherit them
The MQL stage is a large-company artefact that small teams inherit and never question. In a business with a marketing department, a sales department and a demand generation budget, a Marketing Qualified Lead is a handover device: marketing says this person is worth a conversation, sales says this person is worth an hour, and somebody in between is accountable for the disagreement. That is a real problem in a fifty-person company, and the label solves it.
In a five-person company the device has no second half. There is no marketing team to hand over to, no budget to protect, no dashboard to report against. What arrives instead is a pipeline with an extra stage in the middle, a form that routes to it, and a rep who now has one more thing to filter before they can talk to anyone.
That middle stage is not neutral. A lead sitting in MQL is a lead nobody owns. It sits between a marketing list and a sales call, accruing age, quietly costing you the response speed that actually decides whether you get the conversation at all. We have written separately about why small teams are the slowest at speed to lead, and an MQL stage is one of the reasons a small team can be slow while doing everything right.
So the useful question for a small team is not what counts as an MQL. It is what the stage is actually doing for you, and what you would do differently if it were not there.
What a small team can actually measure instead
Three mechanisms get called lead qualification, and they are not equally useful at small size.
| MECHANISM | WHAT IT ACTUALLY DOES | WHAT IT NEEDS |
|---|---|---|
| A status field on the contact | Marks a lead as worth a call, and nothing else | A definition you can write in one sentence |
| A routing rule | Sends a lead to a person or a queue based on something observable | One rule, and a trigger that fires |
| A scoring model | Scores a lead from behavioural or firmographic signals, and ranks a queue by that score | Enough history for the signals to mean anything |
The third row is the one vendors sell hardest, and it is the only one that needs a data asset most small teams have not built yet.
The first two are cheap and start working immediately. A status field on the contact record is a custom field, and a routing rule is an automation with a trigger and an action. Both live in the same product as the pipeline the lead is waiting in, so there is no sync step and no second system to keep alive.
What a small team can genuinely measure is narrower than what a scoring model produces, and it is the part that changes behaviour. Time from form submit to first contact. Contact rate per source. The number of leads that reach a real conversation. The number that become opportunities. Every one of those is a number you can act on, and none of them require a model.
The trap is that a score feels more rigorous than a field. It produces a number, and numbers look like they know something. But a score computed on a year of traffic tells you about a year of traffic, which is why the vendor that ships it also has to tell you how long it takes before the insight is worth anything.
What lead scoring is for, and when it is the wrong lever
Read the vendor pages on 6 November 2026 and you can see exactly where scoring sits and what it costs to get there.
Keap's pricing page lists lead scoring as a capability of the product, described plainly: prioritise your hottest leads and trigger automation based on likelihood to buy. That is the honest description of what a scoring model is for. It is a prioritisation aid for a queue that is already long, and it works best when the automation it triggers is something you would have done anyway if you had known.
HubSpot's own sales page is blunter about the catch. For more experienced teams, it says, advanced features like conversation intelligence and predictive lead scoring take your sales organisation to the next level, and then adds that while these features are available right from the start, it may take some time to gather valuable insights from them. That is a vendor saying the score starts out as a guess and improves with history. Worth weighing before you build a process around it.
ActiveCampaign goes further and puts scoring behind a paywall rather than in the base tier. Its pricing table lists lead scoring as included in the Pipelines add-on across the tiers that carry it, and not included in others, with win probability sitting behind a separate Sales Engagement add-on. Read that as a statement about maturity: scoring and pipeline are sold as an enhanced layer on top of a product, not as a default.
Which is the honest case for scoring: you have a lot of inbound, it arrives from several sources, the sources behave differently, and you cannot hold every conversation. At that size, prioritising the queue is a real gain. The dishonest case is buying a score because a competitor's dashboard has one.
Honest caveats: when you should buy real lead scoring
There are three situations where we would tell a small team to buy proper scoring rather than build it themselves, and one where we would tell them to stay away from it.
- —You have genuine inbound volume from mixed sources and cannot personally work every lead. Past a few hundred leads a month across channels, ranking beats reading. This is the case HubSpot's own copy is describing, and the reason it mentions needing time to gather insights.
- —You sell something expensive enough that one saved sales hour is worth more than the tool costs. At a high contract value, prioritisation has a return that can be measured, and the add-on pricing we read on 6 November 2026 stops looking like a tax.
- —You have a dedicated person whose job is qualification. If someone owns the funnel transition as a full-time role, they need the tooling for it, and our Keap coverage and ActiveCampaign coverage are the ones to read.
And the case to stay away from it: if you have fewer leads than you can work, ranking is solving a problem you do not have, while adding a field nobody updates and a number sales learns to ignore.
We also have to say something about our own product, because this post is otherwise a bit one-sided. BeigeCRM does not ship lead scoring, and we checked our own features page on 6 November 2026 to be sure: no scoring engine, no predictive model, no MQL concept anywhere in the product. That is a deliberate choice for a team of two to twenty people, not an oversight. We give you forms that create a contact instantly, custom fields, automations with triggers and conditions, campaigns that segment from the same records, roles so you can control who sees what, and reporting on stage conversion and deal velocity. A status field and a routing rule are a day's work. A scoring model is a product decision we have made not to take, and if you need one, HubSpot, Salesforce and Zoho are where to look.
What to set up in week one
If you are going to remove the middle stage and replace it with something a five-person team can actually run, this is the order that works.
- —Write the definition in one sentence. Not a score, not a rubric: what does a lead have to have done for a rep to spend ten minutes on them? If you cannot finish the sentence, the stage is not ready and the definition is the first thing to fix.
- —Delete the MQL stage if the sentence does not need one. If a status field can carry it, removing the stage removes the queue nobody owns and the ageing reports nobody reads.
- —Route on something observable, not on opinion. Country, company size, the specific page they asked about, whether they have already bought from you. Automations can act on those and on custom fields, without anyone maintaining a score.
- —Set the clock. Measure hours from submission to first contact and hold the team to a number you can see in reports. This is the qualification metric that actually changes behaviour at small size.
- —Assign an owner to every lead the moment it arrives. Ownership is what stops a lead becoming a queue, and it is the single cheapest thing in this whole list.
- —Review the source every month. If one channel produces leads worth talking to and another produces noise, that is your qualification signal, and it takes one afternoon rather than a model.
None of that needs a marketing platform, a sync, or a second database. It needs one field, one rule, one clock and one owner, which is a small amount of work for a large amount of recovered selling time.
