This is written for budget owners who’ve sat through too many vendor demos promising the moon. You want case studies with raw numbers, reproducible methods, and citation trackers you can audit. Below is an anonymized, data-forward case study with the full context, specific tactics, implementation detail, and explicit metrics. Where possible I note the source tags so a procurement-oriented audit can trace the claims back to raw reports.
1. Background and context
Client: "Client X" (anonymized mid-market B2B SaaS; average contract value (ACV): $36,000). Sales cycle: 90–150 days. Funnel tracked end-to-end in the client’s CRM + analytics (CT-001).
- Quarter (Q0 — baseline): Marketing spend = $450,000 (3 months, $150k/month) Top-of-funnel leads: 3,600 (1,200/month) (CT-002) MQL rate (lead → MQL): 8% → 288 MQLs (CT-003) SQL rate (MQL → SQL): 20% → 57 SQLs (CT-004) Close rate (SQL → closed): 8% → 4.56 deals (CT-005) Quarterly new revenue from closed deals: 4.56 * $36,000 = $164,160 (CT-006) CAC (quarter): $450,000 / 4.56 = $98,684 (CT-007)
Procurement note: all figures above come from the client’s CRM exports and ad platform billing report (see CT-001 through CT-007).
2. The challenge faced
Client X had three interrelated problems:
High CAC relative to ACV — marketing spend was not producing a defensible payback period (CAC ≈ 2.7x ACV in this quarter). Quantity over quality — volume was high but MQLs and SQLs were low quality (long sales cycles and low demo-to-close conversion). Vendor fatigue and lack of evidence — previous vendors offered "growth" without handing over raw tracking data or sample-level conversion logs; procurement required verifiable, auditable proof before budget increases.Consequence: leadership paused budget expansion, and marketing was pressured to demonstrate improvements with verifiable numbers within one quarter.
3. Approach taken
Primary principle: treat funnel optimization as an experiment pipeline, not a vendor-led black box. The team agreed to three constraints:
- Every change must have a hypothesis, an assigned owner, a success metric, and a data source tag (CT-li11/li12li12/li13li13/li14li14/li15li15/li16li16/li17li17/li18li18/li19li19/li20li20/li21li21/li22li22/li23li23/li24li24/li25li25/table1tr1th1th1/th2th2/th3th3/th4th4/th5th5/tr1/tr2td1td1/td2td2/td3td3/td4td4/td5td5/tr2/tr3td6td6/td7td7/td8td8/td9td9/td10td10/tr3/tr4td11td11/td12td12/td13td13/td14td14/td15td15/tr4/tr5td16td16/td17td17/td18td18/td19td19/td20td20/tr5/tr6td21td21/td22td22/td23td23/td24td24/td25td25/tr6/tr7td26td26/td27td27/td28td28/td29td29/td30td30/tr7/tr8td31td31/td32td32/td33td33/td34td34/td35td35/tr8/tr9td36td36/td37td37/td38td38/td39td39/td40td40/tr9/table1/li26li26/li27li27/li28li28/li29li29/li30li30/li31li31/li32li32/li33li33/ol2li34li34/li35li35/ol3li36li36/li37li37/li38li38/ol3/li39li39/li40li40/li41li41/ol2/li42li42/li43li43/li44li44/li45li45/li46li46/li47li47/li48li48/li49li49/li50li50/table2tr10th6th6/th7th7/tr10/tr11td41td41/td42td42/tr11/tr12td43td43/td44td44/tr12/tr13td45td45/td46td46/tr13/tr14td47td47/td48td48/tr14/tr15td49td49/td50td50/tr15/tr16td51td51/td52td52/tr16/tr17td53td53/td54td54/tr17/tr18td55td55/td56td56/tr18/tr19td57td57/td58td58/tr19/tr20td59td59/td60td60/tr20/tr21td61td61/td62td62/tr21/tr22td63td63/td64td64/tr22/tr23td65td65/td66td66/tr23/table2/## Final note: If you’re a budget owner evaluating a vendor, your baseline playbook should be simple: demand sample-level logs, run short experiments that your team controls, and prioritize changes that improve downstream conversion and reduce waste. The numbers above show how a disciplined, auditable process turns vendor noise into measurable results.