Weekly Marketing Ops KPIs for B2B SaaS Teams

Weekly reporting is where marketing operations earns its keep. If you can’t see what moved in the last seven days - by channel, by stage, by dollars - you’re not running growth, you’re guessing. This playbook shows B2B SaaS leaders exactly which KPIs to report every week, why they matter, how to calculate them, and what to do when they drift.

Quarterly views hide execution drift. Monthly views slow feedback. A weekly cadence gives you just enough data to see signal and react before the quarter is lost. It’s the operating heartbeat for demand gen, SDR, and pipeline creation.

Weekly reporting is not about boasting vanity stats. It’s about a short set of leading and lagging indicators that help you answer three questions: Did we create qualified demand? Did Sales touch it on time? Did we turn it into pipeline at an acceptable cost?

Get the discipline right, and you’ll ship more precise campaigns, fix routing issues before they clog, and keep CAC in check. If your dashboard can’t tell you within five minutes where you’re pacing versus target and what to fix by Wednesday, rebuild it.

Think in layers. Layer 1: attention and intent (visitors, impressions, CTR, demo/trial requests, replies). Layer 2: qualification and conversion (MQL, SQL, SQO/opportunity, stage conversion, SLAs). Layer 3: dollars (new pipeline value, win rate, CAC, payback). Layer 4: reliability (data health, tracking uptime, experiment throughput).

Each layer must connect to the next through clear definitions. A “demo request” should not sit in limbo - it becomes an MQL (if ICP fit rules pass), then SQL when an AE accepts it, then an opportunity when there’s a defined next step and value. Your weekly report should show movement across these handshakes.

Set targets per layer. If targets for upstream inputs change (e.g., paid budget cut), translate that down-funnel immediately into revised SQL and pipeline commits. Weekly KPIs are not static; they are contracts with reality.

Top-of-funnel KPIs tell you whether you’re creating qualified attention, not just traffic. Track them by channel and by campaign so you can reallocate budget mid-week if needed. Do not collapse dissimilar channels into a single “leads” number.

At this layer, you’re looking for directional movement and quality signals. A spike in raw leads without an increase in SQLs means targeting or forms need work. A flat lead count with rising SQLs could mean you’re filtering better - double down on that targeting.

Report the last 7 days absolute numbers, the 4-week rolling average, and variance versus last week. This smooths noise without waiting a month.

Organic and SEO Signals

Organic should be the compounding engine. Weekly, track: non-brand organic sessions, demo/trial form submissions from organic, and organic-attributed SQLs. Segment by page type (product pages, comparison pages, pain pages) to spot intent shifts.

Quality control metrics: click-through rate for key pages, indexation status for new content, and % of organic leads that meet ICP criteria. Search is seasonal and lagging, so use a 4-week moving average to avoid whiplash from algorithmic wobble.

If non-brand organic visits rise but organic SQLs don’t, inspect content-to-offer match. Are you ranking for queries you don’t actually serve? Tighten internal linking to product pages and insert higher-intent CTAs on high-traffic informational pages.

Paid Media Efficiency Signals

For paid, weekly precision pays. Report: cost, CPM/CPC, CTR, conversion rate to lead, cost per SQL (not just CPL), and pipeline per dollar. If you’re early or low volume, track “cost per qualified booking” as a proxy while you accumulate revenue data.

Don’t let platform-reported ROAS fool you in SaaS. Instead, map ad click → first-touch contact → SQL → opportunity value. Your weekly table should show: spend, SQLs, SQO rate, new pipeline, and blended CAC trend by campaign.

When cost per SQL jumps week-over-week, first check creative fatigue and audience saturation. If the pipeline per dollar holds while costs rise, your CVR downstream improved - consider keeping the spend but refresh creative. If both degrade, pause and redeploy to proven campaigns.

Outbound and Partner Signals

Outbound is sensitive to domain health and offer quality. Weekly, track: delivered rate, open rate, reply rate, opt-out/complaint rate, positive reply rate (interest or meeting), meetings set, and SQLs. Segment by list source and messaging angle.

For partners, track referrals sourced, SQL acceptance rate, and partner-sourced pipeline. A low acceptance rate usually signals mismatch on ICP or unclear qualification criteria on the partner side.

Outreach velocity matters. Show “new accounts added to sequence” and “active accounts in sequence” to avoid pipeline cliffs two weeks from now. If volume drops this week, next week’s meetings will suffer.

This is where marketing and sales alignment is proven or disproven. Report MQLs created, MQL-to-first-touch time (SLA), MQL-to-SQL acceptance rate, and the reasons for rejection. Then track SQL-to-opportunity creation rate and median days from SQL to opportunity.

Formulas to align on: MQL definition (fit + intent rules), SQL definition (sales accepted and scheduled), and Opportunity definition (value, close date, defined next step). If these are fuzzy, your weekly report becomes a debate instead of a decision tool.

Two non-negotiables for the weekly: SLA adherence and disposition coverage. SLA adherence is the % of MQLs touched within X hours (e.g., 2 business hours). Disposition coverage is the % of MQLs with a clear outcome code (accepted, recycled, disqualified) within 7 days. Aim for 90%+ coverage to keep your dataset trustworthy.

Dollars convert opinions into decisions. Weekly, track new pipeline created (sum of new opportunities’ amounts), weighted pipeline (amount × stage probability), and pipeline coverage versus next-period target. If you can’t see next month’s coverage relative to New ARR goal, you’re driving blind.

Efficiency: blended CAC and CAC payback. CAC (weekly view) = total marketing + sales program costs for the period divided by new customers from that cohort. If weekly customers are too few, show trailing 4-week CAC trend. Payback (months) = CAC ÷ (ARPU × gross margin). If you sell annually upfront, use average first-year gross profit and divide by CAC to get payback in years, then convert to months.

Report win rate and sales cycle (median days) for the last 90 days alongside this week’s pipeline creation to understand whether today’s inputs can realistically hit this quarter’s number. A drop in win rate should trigger qualification or pricing reviews, not just “more leads.”

If you run a PLG or hybrid motion, add activation KPIs: sign-ups, % of sign-ups reaching Activation Event (e.g., first project created), PQLs (product-qualified leads), PQL-to-SQL rate, and expansion/conversion from free to paid.

Weekly, show time-to-activation and the correlation between activation and SQL creation. If PQL-to-SQL is strong but sign-ups stall, fix acquisition. If sign-ups rise but PQLs don’t, your onboarding is weak - run experiments on first-run experience and in-app prompts.

Don’t let PLG sit in a different reporting universe. Integrate product events into CRM so the team reviews one funnel, one language, one WBR.

Great KPIs are useless if the data is rotten. Each week, report: tracking uptime (% sessions with source/medium), UTM coverage for paid and outbound, duplicate rate on contacts/companies, % of contacts with lifecycle stage, % of deals with owner and amount, and routing error rate.

Include a short list of broken forms, integrations in error, and playbooks missed. Your team should know within the first five minutes what must be fixed today to save the week.

If your CRM can’t support this view, fix the system first. HubSpot or an equivalent CRM with standardized lifecycle stages, lead statuses, and routing rules is non-negotiable. If you need an implementation or cleanup partner, see our CRM implementation service.

Timebox it. The weekly is not a brainstorm. It’s a decision checkpoint. Keep the deck or dashboard to a single page with drill-down links for owners to fix issues afterward.

Flow for 30 minutes: (1) 3-minute headline: are we pacing to goal? (2) 10 minutes funnel walk: TOFU → SQL → pipeline → revenue; call out variances >10%. (3) 10 minutes for red flags and owner-assigned fixes. (4) 5 minutes to confirm next week’s experiments and budget shifts.

Every owner leaves with one fix and one action. Document decisions inline on the dashboard so the team sees deltas the following week - no mystery meetings, no second decks.

Reliable weekly reporting needs clean plumbing. Standardize UTMs for every paid click, outbound link, and partner referral. Enforce naming conventions (channel_source_campaign_adset_creative) so you can roll up cleanly without spreadsheets.

Automate cost ingestion into your data hub or CRM. If you operate in HubSpot, pipe in Google/Meta/LinkedIn costs daily, and normalize currencies. Tie costs to campaigns so cost per SQL and pipeline per dollar are native metrics, not exports.

Build automations to guardrail data health: auto-enrich ICP fields, auto-assign leads by territory or segment, and auto-close stale leads with a recycle reason after X days of no response. If you’re short on ops bandwidth, our marketing automation practice sets up Make/n8n workflows, and our performance marketing service ensures platform spend and CRM outcomes stay in sync.

Benchmarks are helpful, but context beats averages. A mid-market ACV with outbound will tolerate higher CAC than a self-serve motion. Use internal history plus peer guidance to set ranges, then iterate monthly.

Good alert design for weekly ops: set thresholds where small deviations matter. Examples: SLA adherence below 80% triggers a same-day routing audit; cost per SQL up 25% week-over-week triggers creative rotation; organic SQLs down 20% versus 4-week average triggers content/technical review.

Publish these rules so nobody argues about whether to act. The data says “fix,” you fix. Then you confirm on next week’s dashboard whether the fix worked.

Agencies tend to report what they own - impressions, clicks, MQLs. Operators report the full chain - SQLs, pipeline, CAC, payback - and fix what breaks across tools, teams, and handoffs. If you prefer one senior owner over a parade of account managers, the operator model will feel lighter and faster. At Anton Chisliuk · AI Marketing Ops, we run the full stack - HubSpot, SEO, paid, outbound, and automation - with proof like taking Cleverence from zero to 80K organic visits/month and ~2K inbound leads/month. If you want an outside view on your weekly KPI setup, book a short Growth Audit - we’ll map your gaps and the fastest fixes.

Keep it to one screen with drill-downs. If you can’t see it all at once, you will miss cause and effect. Build it in your CRM’s reporting tool or in Looker Studio, but anchor the data in the CRM to avoid CSV chaos.

Recommended layout: top row shows Goal vs Current (SQLs, Pipeline Created, CAC, Payback). Middle rows show channel tiles (Organic, Paid, Outbound, Partners) with spend, SQLs, pipeline/dollar, and week-over-week deltas. Bottom rows show Ops Quality (SLA, routing errors, tracking uptime) and Experiment Throughput (planned vs shipped).

Add owner initials to each tile. A number without an owner won’t move. Add a small note field next to any metric in alert state so the team sees the fix path without opening another doc.

Failure mode: “Leads are up, pipeline is flat.” Diagnosis: poor qualification or handoff. Fix: tighten MQL rules, enforce SDR SLA, audit auto-routing and meeting links, and review offers that draw non-ICP traffic.

Failure mode: “Paid looks great in-platform, ugly in CRM.” Diagnosis: attribution misalignment or missing UTMs. Fix: standardize UTM templates, pass GCLID/Click IDs, and reconcile costs daily into CRM campaigns so SQL and pipeline are ground truth.

Failure mode: “Great dashboard, wrong data.” Diagnosis: inconsistent lifecycle stages and statuses. Fix: lock picklists, add validation, require owners on every deal, and run weekly audits for records missing key fields - then automate corrections where possible.

SQLs (weekly): count of leads accepted by Sales within the last 7 days. Target by segment. Action if down >15%: check MQL volume and acceptance reasons, then shift spend to proven channels.

New Pipeline Created: sum of amounts of opportunities created in last 7 days. Action if short: review opportunity creation discipline (are AEs delaying creation?), adjust offers to pull forward intent, run a mid-week outbound sprint on high-fit accounts.

Blended CAC (trailing 4 weeks): (Paid media + content + tool costs + SDR/marketing program costs) ÷ new customers. Action if rising: audit channel mix, trim low-efficiency campaigns, test higher-intent offers, and review pricing/discounting policy if win rate softened.

Definitions drive decisions. Lock your lifecycle: Subscriber → Lead → MQL (fit + intent) → SQL (sales accepted) → Opportunity → Closed/Won-Lost. Lock lead statuses: New, Open, In Progress, Recycled, Disqualified - with reasons. Lock opportunity stages with clear exit criteria and stage probabilities.

Create a single source of truth doc. Include examples of acceptable and unacceptable records at each stage. Include screenshots. If someone new can’t understand how to classify a lead in five minutes, it’s too complex.

Revisit definitions quarterly. When your ICP shifts or your motion changes, definitions must evolve. Don’t let legacy rules sabotage current growth.

Data without action is decoration. Set “If X then Y” plays. If SLA drops under 80%, managers run a same-day blitz to touch all aging MQLs and postmortem routing errors. If organic SQLs dip, content ops pulls the top 10 landing pages to update offers and internal links, and SEO checks indexation/coverage.

Create a standing backlog of fixes and experiments. Each week, pick the top three with the highest expected impact and lowest effort. Mark them as “shipped” or “slipped” on the next report. Momentum compounds.

Close the loop. Every alert should link to the owner and the fix steps. Every fix should show its impact in two weeks. This is how a dashboard becomes an operating system, not a poster.

Weekly KPIs are not a dashboard vanity project; they are the control panel of your go-to-market. Track inputs (attention and intent), handoffs (SLAs and acceptances), and outcomes (pipeline, CAC, payback). When a number tilts red, fix it within the week, not the quarter.

If your current reporting can’t answer “where are we pacing and what to do next” in under five minutes, it’s time for a rebuild. Start with clear definitions, instrument your CRM, and publish a one-page WBR that owners trust.

Want a second set of senior eyes on your setup? Book a short Growth Audit and get a prioritized plan to fix tracking, SLAs, and the KPI stack that drives pipeline.

Written by Anton Chisliuk - operator, not agency. Leads, sales, content, and systems - one senior operator.

- What’s the minimum viable weekly dashboard for a B2B SaaS with a sales-led motion?

Five tiles: SQLs, new pipeline created, MQL→SQL acceptance rate, SLA adherence, and blended CAC trend. Add channel tiles for paid and organic if you spend meaningfully there. Keep it to one screen and link to drill-downs.

- Should we include revenue closed this week, or is that too lagging?

Include it, but don’t over-index. The purpose of the weekly is to steer inputs and handoffs. Closed-won is a health signal, but pipeline creation and SLA adherence are what you can still influence this week.

- How do we handle small sample sizes week-to-week?

Pair weekly actuals with a 4-week moving average and variance. For CAC and win rate, also show a 90-day view. You still want the weekly to trigger fixes, but use the longer window to confirm trends before making big bets.

- Our paid channel reports great CPL, but SQLs are weak. What’s the first fix?

Audit audience and offer match. CPL can be low if you attract the wrong profiles. Raise intent (use demo/trial offers or comparison pages), tighten geo/firmographic filters, and ensure your form/routing qualifies out non-ICP leads early.

- How strict should our MQL definition be?

Strict enough that Sales trusts it. If acceptance rate is below 70%, you’re wasting AE time. Use fit (firmographic/technographic) plus intent (behavioral signals) to score. It’s better to pass fewer, higher-quality MQLs weekly than flood the queue.

- What tool should we build the weekly report in?

Where your data is most reliable. If your CRM is clean, build it there for immediate actionability. If you have a warehouse and BI muscle, centralize there but keep ownership and definitions aligned to CRM objects.

- How do we quantify experiment throughput on a weekly report?

Track experiments planned vs shipped, with projected impact and owner. Keep a simple “ICE” or expected pipeline impact tag. If you ship fewer than two meaningful experiments a week, you’re not iterating fast enough.

- Can weekly KPIs work for a hybrid PLG and sales-led model?

Yes. Add PQLs and activation metrics next to MQL/SQL, and show cross-over rates (PQL→SQL, PQL→Paid). Keep one dashboard so the team sees the full motion - don’t split PLG and sales into silos.