Brightline Robotics: Traffic Tripled, Pipeline Fell
Brightline builds warehouse-robotics software and sells to two very different buyers: operations directors at 200–2,000-person logistics firms (38% of closed-won revenue) and IT infrastructure leads at enterprise 3PLs (62%). Over twelve months the blog grew from 18,000 to 55,000 monthly sessions, but new contacts stayed flat at roughly 1,900 per month and marketing-sourced pipeline fell 9%. Sales complains that inbound leads 'just want the free checklist.' The demand generation manager proposes hiring two more writers to double blog output, arguing that traffic is the proven leading indicator. Current numbers: 12 posts produce 71% of organic sessions; blog-to-contact conversion is 0.6%; the demo request page converts 1.8% of visitors; MQL-to-SQL is 11% against a 24% internal target; cost per MQL is $310. There is no lifecycle reporting beyond a monthly MQL count, and the lead scoring model was built two years ago and has never been reviewed.
1. Sales blames lead quality, demand generation blames lead volume. Using the numbers in the setup, locate the actual constraint and decide on the two extra writers. Show your arithmetic.
Read the worked answer
Arithmetic first. 55,000 sessions × 0.6% = about 330 blog-sourced contacts a month out of 1,900 total. Doubling output might add 300–400 contacts, i.e. roughly 20% more volume into a stage that already rejects 89% of its intake. That is buying more of the wrong thing.
Where the gap actually is. MQL→SQL at 11% versus a 24% target means about 54% of the distance to target sits at sales acceptance, not at the top of the funnel. Second signal: traffic tripled while contacts stayed flat, so blog conversion fell by roughly two-thirds. The added sessions are low-intent, and they are not coming from the persona that closes - one buyer is over-represented in traffic and under-represented in revenue.
Decision. Decline the two writers. Fund one writer plus a marketing-ops half-FTE, and spend the next six weeks on: rebuilding scoring against closed-won contact attributes, adding a mid-funnel offer for the revenue-heavy persona, and a landing page test on the demo page (0.6% to contact, 1.8% visitor-to-demo is the shortest lever).
What would change my mind: if MQL→SQL were at or above 24% and demo page conversion were healthy, then volume would be the constraint and headcount would be justified.
2. Propose a defensible MQL definition for Brightline. Which signals would you include, which would you deliberately exclude, and what would make you revisit it?
Read the worked answer
Treat MQL as a handoff agreement, not a marketing metric. It should predict that sales will accept and work the record, so I would build it from two factors multiplied, not one score summing everything.
Fit: company size in the 200–2,000 or enterprise band, industry in logistics/3PL, role in the two buying groups (or a known influencer adjacent to them). These are stable properties set from enrichment and form fields.
Intent: two or more high-cost actions inside a window - pricing page, demo page viewed twice, ROI calculator completed, live webinar attendance, trial signup. Explicitly exclude: a single blog view, newsletter subscription, one checklist download, and email opens (unreliable post-privacy changes).
Negative signals I would add: student or free-mail domains with no company match, known competitor domains, and job-title patterns like 'student' or 'looking for opportunities'.
Threshold, not a single event. A contact should need fit plus a minimum intent score to cross, and the handoff needs an SLA - if sales does not touch an MQL in 24 hours, that is a routing failure, not a lead-quality failure.
Revisit trigger: a monthly review with sales of SQL acceptance by MQL source and by persona. The current model is two years stale, so I would deliberately re-derive thresholds from closed-won contact attributes rather than from opinion.
3. The CMO wants a single dashboard number for 'marketing's contribution' next quarter. What would you put on the dashboard and what would you refuse to include?
Read the worked answer
Refuse first. I would refuse a bare MQL count as the headline, a blended cost-per-lead across two personas with different economics, and any single-touch attribution figure presented as the contribution number. Each of those will be gamed or argued with within a quarter.
Build in two waves. Wave one, shippable now with existing data: stage-to-stage conversion rates by source and by persona - visitor to contact, contact to MQL, MQL to SQL, SQL to closed-won - plus cost per SQL, not cost per MQL. This is honest because the data already exists in contact and deal properties even though nobody reports on it.
Wave two, roughly 60 days: pipeline influenced by content, with the attribution model written down on the dashboard itself (first touch, last touch, or linear) and a note on its known bias. Pair it with a cohort view - contacts created in month X, pipeline generated within 90 days - because it survives reporting-date arguments.
Why not just revenue? With a 90-day-ish B2B cycle and small deal counts, revenue attribution swings wildly month to month. Stage rates and cohort pipeline are the numbers a CMO can act on weekly, and they show whether the fix is working before revenue does.