MOps Metrics: Data Hygiene, MQL→SQL Conversion, List Health

Last quarter I sat in a QBR where the CMO had one question. "MQLs are up 40 percent. Pipeline is flat. Why."

The MOps lead pulled up a dashboard with 26 tiles. Email open rate. Send volume. Contacts in database. Webinar registrations. Form fill count. Average time on site for blog readers. None of it answered the question. The CMO let him talk for two minutes, then said "I'll come back when you can tell me where the leak is" and moved on.

That dashboard was tracking activity. The CMO was asking about funnel health. Those are two different jobs, and if your dashboard can't answer the funnel-health question, you're going to lose the budget conversation.

This is the playbook for what belongs on a QBR slide instead. Six metrics. Each with a B2B SaaS benchmark, a named diagnosis when the number is bad, and an owner. One slide, not six.

Why activity dashboards fail MOps

Activity metrics tell you the team is busy. They don't tell you whether the work is paying off.

Open rate, send volume, campaign count, contacts added, webinar registrations: these are inputs. The CMO is being asked by the CFO for cost-per-pipeline-dollar. If you hand her open rate, she has to translate that into something she can defend, and she'll resent you for making her do it.

The dashboard's job is to answer the CMO's next question before she asks it. That means showing funnel health: are leads real, do they convert, where are they leaking. If you can't trace from "MQL volume" to "pipeline dollars" with named breakpoints in between, the dashboard is decorative.

The fix isn't more metrics. It's fewer, sharper ones, each with a diagnosis attached.

The 6 metrics that belong on a QBR slide

1. Data completeness %

The percentage of records with the five fields routing depends on: email, company, country, employee count, and role.

Healthy: 90% or above. Below 75% means your routing engine is guessing.

This is the metric nobody wants to look at and everybody should. If you can't enrich a record on those five fields, you can't route it to the right SDR, you can't score it, and you can't segment it. The lead exists in your database but it's invisible to every downstream process.

The diagnosis when this number is low: either your forms aren't capturing the fields (progressive profiling missing the basics), or your enrichment vendor is dropping records (Clearbit, ZoomInfo, Apollo silently failing on small companies), or you imported a list and never enriched it.

The first time I rebuilt a scoring model, I found 38% of our "MQLs" had no employee count. The scoring model was treating them as if they were small business when half were actually enterprise. Fix the data first, then touch the model.

2. MQL→SQL conversion rate

The percentage of MQLs that sales accepts within 14 days.

Healthy B2B SaaS: 12-15%. Under 8% means your scoring model is rewarding the wrong behavior. Over 25% means you're under-marketing. The MQL bar is too high and you're hand-delivering only the obvious deals.

This is the one metric the CMO will ask about every QBR. It's the cleanest single signal that marketing and sales agree on what a "lead" is. If it's drifting down quarter over quarter, something upstream broke.

The trap with this metric is treating it as a sales problem. Sometimes it is. More often it's a marketing-source-mix problem: a webinar campaign delivered 800 MQLs, but webinar audiences convert at 4% and demo-request audiences convert at 22%, and the blended number tanked because the mix shifted.

Always look at this metric by source, not just blended. The blend hides everything.

3. List churn rate

The percentage of contacts going hard-bounce, unsubscribe, or marked-spam per send.

Healthy: under 0.5% combined. Over 1% is a sender reputation risk and you should pause the campaign.

This is a hygiene metric, not a vanity one. It tells you whether your list is decaying faster than you're adding to it, and whether mailbox providers are starting to lose trust in your sender domain. Once Gmail or Outlook flags you, the recovery is painful (weeks of warmup, sometimes a domain swap).

Most MOps teams ignore list churn until a deliverability crisis lands. By then it's too late. Watch it weekly.

4. Dedupe rate

Duplicate records created per 1,000 inbound.

Healthy: under 2%. Above 3% and your sales team is calling the same person twice with two different reps.

The diagnosis is almost always "form fills aren't matching on fuzzy email or company." Someone fills a form as jane@acmecorp.com, six months later as jane.smith@acme.com after a rebrand, and the dedupe rule only matched on exact email. Two records, two routing assignments, two SDR touches, and Jane gets annoyed.

Name the root cause on the dashboard. "Dedupe rate 3.4%, fuzzy match on company domain disabled in Marketo, recommend re-enabling." Don't just show the number. Show what's broken.

5. Routing SLA breach %

The percentage of MQLs not assigned to an SDR within the SLA window.

Healthy: under 5%. Above 10% means round-robin is broken or reps are capped.

The SLA windows that matter:

• Inbound demo request: 15 minutes during business hours
• Content MQL (gated asset, webinar): 24 hours
• Tier-1 account match: 5 minutes, alert the AE directly

The reason this metric matters more than people realize: every minute past the SLA window, conversion drops. The Lead Response Management research is old but still directionally true. Contact within 5 minutes is roughly 8x more likely to convert than contact within an hour. If your dashboard says you're breaching SLA on 12% of MQLs, that's not a routing engine problem, that's a pipeline problem.

The diagnoses to check, in order: round-robin is unbalanced (one rep capped, others under-loaded), territory rules have stale ZIP-to-rep mapping, after-hours coverage missing (everything inbound at 9pm Friday breaches by Monday), or the rep count is wrong for the volume.

6. Form-to-lead conversion

The percentage of form submissions that become a usable lead after dedupe, enrichment, and spam filter.

Healthy: 70%+. Below 50% means your forms are catching bots, or your enrichment vendor is dropping records, or your spam filter is too aggressive.

This is the cleanest single signal of "is the top of funnel actually working." A form fills, but does it survive the pipeline that turns it into a routable, scorable, enrichable lead? If half the submissions get junked, your reported "MQL volume" is fiction.

Most teams don't measure this because it requires instrumenting the gap between "form submitted" and "MQL created." It's worth the work, because when this number is healthy, the rest of the funnel math gets honest.

The "high MQL but low SQL" diagnostic

This is the question that tanks QBRs. MQLs are up. SQLs are flat. The CMO wants to know why, and the answer determines whether she gets her budget.

Walk the dashboard in this order:

Step 1: Has the scoring threshold drifted? If the threshold for "MQL" got lowered last quarter (someone wanted volume), you're letting in lower-intent leads. Compare the current threshold against the version 90 days ago. If it dropped, that's your answer.

Step 2: Has the campaign source mix shifted? Pull MQL volume by source for last quarter and this quarter. If webinar grew from 20% of MQLs to 50%, and webinars convert at 4% vs demo requests at 22%, the blended SQL rate has to drop. The marketing team didn't do worse. The mix got cheaper and lower-intent.

Step 3: Is sales capacity the constraint, not marketing? Check the SDR capacity. If MQLs went up 40% but the SDR team headcount is flat and they were already at 90% utilization, they physically can't accept the new volume. That's a sales ops problem, not a scoring problem.

Step 4: Is the MQL definition stale vs ICP? ICP shifts. Maybe last year you sold mostly to 50-200 employee SaaS, and this year the ICP moved upmarket to 500+. The scoring model still rewards 50-200. You're generating MQLs that don't match what sales is being told to close.

Run those four checks in order. The answer is almost always one of them. Tell the CMO which one.

Vanity-metric traps to kill on the dashboard

These belong in a working file, not on the QBR slide.

Open rate alone. Apple Mail Privacy Protection inflates opens by 30-60% because it pre-fetches images for opted-in users. The number you're seeing isn't engagement. It's Apple's privacy proxy. Pair open rate with click rate, reply rate, and meeting-booked rate. If you have to pick one to report, pick reply rate.

Total contacts in database. Bigger isn't better. A healthy database shrinks every quarter as you suppress non-responders, hard bounces, and unsubs. If the contact count only goes up, your hygiene is broken and your deliverability is degrading.

Cost per MQL without cost per SQL alongside. Cost per MQL is meaningless without the conversion rate to SQL. A $50 MQL that converts at 4% costs you $1,250 per SQL. A $200 MQL that converts at 18% costs you $1,111 per SQL. The cheap MQL is the expensive one. Show both numbers, always.

Single-touch attribution. First-touch and last-touch each tell you a sliver of the story and lie about the rest. If you're going to report attribution, run a multi-touch sanity check (W-shaped, time-decay, or data-driven) alongside the single-touch view. When they disagree, the disagreement is the insight.

The QBR slide layout

One slide. Six rows. Six columns.

Everything on this slide answers a question. The CMO can read it left to right and know: where are we, where should we be, what's wrong, who fixes it, what happens next. That's the entire job of the slide.

Operating cadence

The dashboard isn't a one-time build. It's a living instrument with three review rhythms.

Weekly hygiene check (15 min, MOps IC alone): Data completeness, dedupe rate, list churn. If any of these breach, fix this week. Don't let hygiene rot. The longer it sits, the more downstream metrics get corrupted.

Monthly funnel review (60 min, MOps IC + Demand Gen lead): MQL→SQL conversion (by source), form-to-lead conversion, routing SLA breach. Look at trends, name diagnoses, decide on adjustments. Bring receipts to the CMO if conversion is moving the wrong way.

Quarterly to CMO (the QBR slide): The six metrics, with named diagnoses and owners and next actions. No story-time, no apologies, no walls of context. The slide answers the question; you're there to explain the diagnoses if asked.

What "good" looks like in 90 days

If you inherit a dashboard with 26 tiles and start applying this playbook, here's what the numbers should look like in a quarter:

• Data completeness: moves from 70% to 90%. The single biggest unlock — every downstream metric improves once routing isn't guessing.
• MQL→SQL conversion: stable inside 12-15%, with named source-level visibility. Drift gets caught in the monthly review, not the QBR.
• Routing SLA breaches: under 5%, with after-hours coverage handled. The SDR team stops complaining about cold MQLs because they're not cold anymore.
• List churn and dedupe: under benchmarks, weekly hygiene held. Deliverability stays strong, sales stops calling the same person twice.
• Form-to-lead conversion: 70%+, with the gap between "form submitted" and "MQL created" instrumented and visible.
• The dashboard itself: six numbers, not 26. Activity metrics moved to a working file. Every QBR-slide number has a benchmark, a trend, a diagnosis, and an owner.

The CMO walks into the QBR and sees the slide and knows the answer to her next question before she asks it. That's the whole job.

Learn More

• The MOps Stack: Tools, Integrations, and Why You Already Have Too Many
• Lead Routing and SLA Enforcement
• Attribution the CFO Can Defend
• Marketing Operations Manager — JD