Why Paid Search and Organic Disagree on MQLs
In Q3, a B2B company's website sessions rose 30%, but demo form submissions fell 8% and MQLs fell 12%. The paid search manager says paid search drives 60% of MQLs based on HubSpot's 'Original Source' contact property. The content lead says the blog drives 45% of first touches based on a multi-touch attribution report. The marketing ops manager sees 'Latest Source' telling a third story. UTM tagging is inconsistent: some links use utm_source=LinkedIn, others linkedin, some paid social links have no UTMs, and the webinar campaign was never set up as a HubSpot campaign. Offline field events are tracked by manually editing contact properties. Leadership wants one dashboard for next quarter. The company has Marketing Hub Enterprise, CMS Hub Professional, and Salesforce integrated with HubSpot.
1. Explain why the three reports disagree. Distinguish what each HubSpot data source actually measures.
Read the worked answer
They measure different things. Original Source is a contact property set from the first known source; it does not change when the contact later clicks an ad. Latest Source reflects the most recent source, so a blog reader who later fills a demo form from paid search appears as paid search. Multi-touch attribution distributes credit across several touches and may count touches that never set source properties.
Traffic analytics is session-based; form submissions are event-based; contacts are people. One person can create five sessions and two submissions, so a 30% session increase with fewer submissions can still happen if traffic is lower-intent.
The missing webinar campaign and inconsistent UTMs make it worse: untagged links fall into referral/direct, and LinkedIn and linkedin become separate values. Manual edits to offline contacts create a source no report can audit.
Recommendation: do not declare one report right. Define the question: acquisition (original source), last interaction (latest source), or journey (multi-touch). Then fix tracking before comparing.
2. Design a tracking governance plan for Q4. Include UTM conventions, campaign setup, and how to handle offline events.
Read the worked answer
UTM convention: lowercase everything; standardize utm_source (linkedin, google, newsletter), utm_medium (paid_social, cpc, email), and utm_campaign (q4-webinar-security). Use one shared builder and a naming doc. Never put UTMs on internal links.
Campaign setup: create a HubSpot campaign for every initiative before launch. Associate the landing page, emails, ads, blog posts, and list. Add campaign parameters to all external links. For paid social links that currently lack UTMs, fix the ad URLs at the source, not with redirects.
Offline events: create a campaign and track attendance with a list or custom event; import attendees with the campaign association. Do not manually change Original Source. Add a required 'how did you hear about us' form field only if sales will use it, and keep values controlled.
QA: spot-check URLs weekly and audit the Sources report for untagged traffic, uppercase variants, and direct traffic spikes.
3. What should the leadership dashboard show, and what caveats should it include?
Read the worked answer
Show four blocks: traffic (sessions by source/medium, new versus returning), conversion (form submissions, contacts created, MQLs), pipeline (deals, pipeline value, revenue from Salesforce), and campaign (spend, MQLs, pipeline per campaign). Separate first-touch and multi-touch views rather than blending them.
Include definitions and caveats on the dashboard: date window, attribution model, lookback window, and data completeness (share of untagged sessions). Mark small-sample campaigns as directional. Show form submissions and contacts separately so deduplication is visible.
For leadership, lead with one question: which campaigns create pipeline? Then show paid search first-touch and multi-touch side by side, with the blog as an assist. Add a monthly data-quality tile: untagged traffic percentage, missing campaign assets, and manual source edits. That prevents false precision and gives marketing ops a fix list.