A Marketing Growth Engineering Case Study
A mid-market B2B SaaS company offering workforce management software was generating a strong volume of inbound leads through paid media but struggling to translate that volume into pipeline. Despite a healthy cost per lead (CPL) of approximately $42, the sales team was closing fewer than 2% of marketing-sourced leads, and reps were spending hours each week chasing prospects who had no buying intent, no budget, or no fit.
The marketing objective was clear: stop optimizing for cheap leads and start optimizing for sales-ready prospects. Over a six-month engagement, we rebuilt the measurement infrastructure, integrated CRM signals back into ad platforms via offline conversion tracking, deployed a lead scoring model, and re-trained campaigns to bid on qualified outcomes rather than form fills.
The results were significant. Total lead volume dropped by 38%, but Sales Qualified Leads (SQLs) increased by 71%, cost per SQL decreased by 54%, and pipeline generated from paid media grew by 2.3x. Sales acceptance rates climbed from 31% to 68%, and marketing-sourced closed-won revenue increased by 112% year-over-year.
Company profile: A privately held B2B SaaS firm with approximately 180 employees, selling a workforce management and scheduling platform to mid-market companies in retail, hospitality, healthcare, and logistics. Average contract value (ACV) sat around $48,000 annually, with a typical sales cycle of 60 to 90 days.
Target audience: Operations directors, HR leaders, and IT decision-makers at companies with 200 to 5,000 hourly employees. Primary buyers were typically VP-level or above, with influence from frontline operations managers.
Paid media channels in use:
- Google Search (branded and non-branded)
- Google Performance Max
- Meta Ads (Facebook and Instagram)
- LinkedIn Ads (Sponsored Content and Lead Gen Forms)
- YouTube (TrueView for Action)
Original strategy: The original lead generation playbook leaned heavily on top-of-funnel content offers, downloadable guides, ROI calculators, and broad keyword targeting. Campaigns were optimized toward “Lead” conversions defined as any form fill, and bid strategies were set to Maximize Conversions or Target CPA based on form submission volume. The CRM (HubSpot, integrated with Salesforce for the sales team) operated as a downstream destination rather than a feedback source.
Optimizing for lead volume alone created a cascade of operational and financial issues:
Low conversion rates. Of roughly 1,400 monthly leads, only 28 to 35 were converting to opportunities. Lead-to-opportunity rates hovered near 2.2%, well below the SaaS benchmark range of 6 to 13%.
Poor sales follow-up efficiency. SDRs were working through long lead lists with no prioritization signal beyond submission timestamp. Calls were going to job seekers, students, competitors, and small businesses well below the ICP threshold. Average time spent per qualified conversation was inflated by the volume of unqualified outreach.
Wasted ad spend. Approximately 41% of the paid media budget was being absorbed by audiences that converted on form fills but never engaged meaningfully with sales. Performance Max in particular was driving low-quality leads from audience segments the team had little visibility into.
Unqualified inquiries. Lead Gen Forms on LinkedIn and Meta were producing high submission rates because they removed friction, but they also stripped away qualifying intent. Roughly 60% of social-sourced leads were companies with fewer than 50 employees, well outside the ICP.
Marketing-sales misalignment. Marketing was hitting MQL targets every quarter while sales was missing pipeline numbers. The two teams were optimizing for different definitions of success, and the disconnect was eroding trust between functions.
The strategic shift centered on a simple principle: ad platforms can only optimize for the signals they receive. If we kept feeding them form fills, they would keep finding form-fillers. If we fed them qualified leads and closed-won deals, they would learn to find buyers.
Offline conversion tracking. Send back-end CRM events (MQL, SQL, Opportunity Created, Closed-Won) to Google Ads, Meta, and LinkedIn so platform algorithms could optimize for revenue-correlated outcomes rather than top-of-funnel actions.
CRM feedback loops. Build automated workflows in HubSpot and Salesforce to push lead status changes back to ad platforms via API, GCLID/FBCLID/LI click ID matching, and Conversions API where applicable.
Lead scoring model. Develop a hybrid demographic and behavioral scoring system. Demographic factors included company size, industry, job title, and geography. Behavioral factors included pricing page visits, demo requests, repeat visits, and content depth.
SQL criteria definition. Establish clear, jointly owned criteria between marketing and sales using a BANT-influenced framework: Budget signal, Authority (title), Need (use case fit), Timeline (within 6 months).
Audience segmentation. Move from broad targeting to ICP-aligned segments using firmographic layers on LinkedIn, custom audiences from CRM on Meta, and customer match plus in-market segments on Google.
Negative keyword refinement. Conduct a search query mining sweep and add negatives for job-seeker terms, DIY/free tool searches, academic research queries, and competitor employee searches.
Landing page improvements. Replace ungated content offers with progressive profiling forms on solution-specific pages. Add qualification questions (company size, current solution, timeline) that filter out non-ICP traffic before submission.
Ad messaging adjustments. Shift from generic “download our guide” creative to ICP-specific pain points, named industries, and outcome-focused proof points.
Conversion event optimization. Reconfigure each platform to optimize toward SQL events rather than Lead events, while keeping Lead as a secondary tracked event for diagnostic purposes.
Marketing and sales alignment. Implement a weekly pipeline review between demand gen and SDR leadership, with shared dashboards in Looker Studio pulling from GA4, ad platforms, and Salesforce.
Step 1: Audit and instrumentation (Weeks 1 to 3). Audited the existing GTM container, GA4 configuration, and CRM integration. Found broken UTM parameters on three of five paid channels, missing GCLID capture on the primary form, and no Salesforce-to-Google Ads connection. Rebuilt GTM with a clean event taxonomy: form_submit, demo_request, mql, sql, opportunity_created, closed_won.
Step 2: CRM and ad platform integration (Weeks 3 to 6).
- Connected Salesforce to Google Ads using the native integration to enable offline conversion imports.
- Implemented Meta Conversions API via a server-side GTM container, passing hashed email and FBCLID alongside CRM-defined events.
- Set up LinkedIn Conversions API and uploaded historical SQL data to seed the algorithm.
- Built HubSpot workflows that triggered API pushes to each ad platform when a contact’s lifecycle stage changed.
Step 3: Lead scoring deployment (Weeks 4 to 7). Built a 100-point lead scoring model in HubSpot. Demographic fit accounted for 60 points (company size, title, industry), and behavioral signals accounted for 40 points (pricing page views, demo form starts, return visits within 14 days). MQL threshold set at 55 points. SQL designation required MQL status plus an SDR-confirmed conversation meeting BANT criteria.
Step 4: Campaign restructuring (Weeks 6 to 10).
- Google Ads: Migrated Search campaigns to Maximize Conversion Value with the SQL conversion as primary. Paused Performance Max temporarily and relaunched with customer match audiences and value-based bidding. Added 340+ negative keywords across the account.
- Meta Ads: Rebuilt audience structure around lookalikes seeded from closed-won customers (1% LAL) and CRM-suppressed cold prospecting. Switched optimization event to SQL via CAPI.
- LinkedIn Ads: Tightened firmographic filters (company size 200+, specific industries, target titles). Replaced Lead Gen Forms with conversion-tracked landing pages for prospecting campaigns; retained LGF only for retargeting.
- YouTube: Refocused on retargeting and customer match rather than broad in-market.
Step 5: Landing page and creative refresh (Weeks 8 to 12). Built four ICP-specific landing pages (retail, healthcare, hospitality, logistics) with industry-specific case study proof, qualifying form fields (company size as a required dropdown with a hard floor of 200 employees), and pricing transparency. Refreshed ad creative to lead with industry-named headlines and quantified outcomes.
Step 6: Reporting and alignment (Weeks 10 to 14). Built a Looker Studio dashboard combining GA4, Google Ads, Meta, LinkedIn, and Salesforce data via a data warehouse layer. Established a weekly 30-minute marketing-sales sync reviewing SQL volume, source quality, and pipeline progression by channel.
Step 7: Iteration (Weeks 14 to 26). Reviewed conversion data weekly, expanded negative keyword lists, refined lead scoring thresholds based on observed close rates by score band, and reallocated budget toward channels and campaigns producing the best cost per SQL.
The measurement framework moved beyond CPL to a full-funnel view:
| Metric | Definition |
|---|---|
| MQL volume | Leads scoring 55+ points and meeting demographic ICP criteria |
| SQL volume | MQLs accepted by sales after qualifying conversation |
| Cost per MQL | Paid media spend divided by MQL count |
| Cost per SQL | Paid media spend divided by SQL count |
| Lead-to-opportunity rate | Percentage of leads converting to qualified opportunities |
| Opportunity-to-close rate | Percentage of opportunities reaching closed-won |
| Offline conversion rate | Percentage of online conversions that progressed to offline qualified events |
| Pipeline influenced | Total opportunity value with paid media touchpoints |
| Pipeline ROI | Pipeline value divided by media spend |
| Closed-won ROAS | Closed revenue divided by media spend |
Attribution was modeled using GA4 data-driven attribution for online touchpoints and a Salesforce campaign influence model for full-funnel revenue. Both views were reconciled weekly to triangulate channel value rather than relying on a single attribution method.
Six-month performance comparison (pre-engagement baseline vs. post-optimization):
| Metric | Before | After | Change |
|---|---|---|---|
| Monthly leads | 1,420 | 880 | -38% |
| Monthly MQLs | 312 | 408 | +31% |
| Monthly SQLs | 64 | 110 | +72% |
| Cost per Lead | $42 | $68 | +62% |
| Cost per SQL | $930 | $428 | -54% |
| Lead-to-opportunity rate | 2.2% | 7.8% | +254% |
| Opportunity-to-close rate | 18% | 26% | +44% |
| Sales acceptance rate | 31% | 68% | +119% |
| Pipeline generated (6 mo) | $2.1M | $4.9M | +133% |
| Closed-won revenue (6 mo) | $410K | $870K | +112% |
| Pipeline ROI | 4.2x | 9.6x | +129% |
The headline insight: lead volume dropped meaningfully, and CPL went up. On the surface, those metrics looked worse. But every metric tied to revenue improved, and the sales team’s productivity gains alone (estimated at 15 hours per SDR per week reclaimed from unqualified outreach) translated into additional capacity that further compounded results in the following quarter.
Ad platforms optimize for what you measure. Sending only top-of-funnel events teaches algorithms to find top-of-funnel responders. Closing the loop with offline conversions is the single highest-leverage change most B2B advertisers can make.
CPL is a vanity metric in B2B. It tells you nothing about whether the lead can buy, will buy, or should buy. Cost per SQL and cost per closed-won are the metrics that matter.
Friction is a feature, not a bug. Adding qualifying form fields reduced volume but dramatically improved fit. The leads that completed the longer form had higher intent and were materially more likely to convert.
Lead scoring only works if both teams agree on it. Models built in marketing’s silo get ignored by sales. Co-authoring the criteria and reviewing scoring accuracy weekly built the trust needed for both teams to act on it.
Volume and quality are not opposites, but they require different campaigns. Brand awareness and demand creation campaigns still need volume metrics. Demand capture campaigns need quality metrics. Mixing the two in one optimization goal produces mediocre results on both.
Attribution disagreements are healthy. GA4 and Salesforce will rarely match exactly. The point is not to pick one source of truth but to use the disagreement as a diagnostic.
For in-house marketers: Audit your conversion event hierarchy this quarter. If your campaigns are still optimizing toward form fills, you are leaving pipeline on the table. Prioritize getting offline conversions wired up before testing new creative or new channels.
For agencies: Stop reporting CPL as the headline metric in client decks. Build dashboards that lead with cost per SQL and pipeline contribution. Clients who only see CPL will eventually push for cheaper leads, which will undermine your retention when sales productivity suffers.
For founders and revenue leaders: Instrument the CRM-to-ad-platform integration as table stakes, not as a nice-to-have. Budget for a marketing operations resource (in-house or fractional) whose explicit job is to maintain the data pipeline between Salesforce or HubSpot and your media channels.
For RevOps teams: Build a single source of truth that combines media spend and pipeline data at the campaign level. Looker Studio, Tableau, or a warehouse-native BI tool will work; what matters is that marketing and sales argue from the same numbers.
For everyone: Run a quarterly closed-loop review. Pull the last 90 days of closed-won deals and trace each one back through the funnel to its first paid touchpoint. Patterns will emerge that no platform dashboard will surface on its own.
Bondarenko, A., Kraus, S., Zubielqui, G. C. de, & Spender, J.-C. (2023). Marketing-sales alignment: A review and research agenda. Journal of Business Research, 155, 113392. https://doi.org/10.1016/j.jbusres.2022.113392
Forrester Research. (2023). The state of B2B marketing measurement and attribution. Forrester.
Gartner. (2024). 2024 CMO spend and strategy survey. Gartner Research.
Google. (2024). About offline conversion imports for Google Ads. Google Ads Help. https://support.google.com/google-ads/answer/2998031
Google. (2024). Best practices for value-based bidding. Google Ads Help.
Hopkins, B., & Silver, L. (2022). Closed-loop attribution in B2B paid media: A framework for revenue-aligned optimization. Journal of Digital and Social Media Marketing, 10(2), 142–158.
Järvinen, J., & Taiminen, H. (2021). Harnessing marketing automation for B2B content marketing. Industrial Marketing Management, 93, 117–128. https://doi.org/10.1016/j.indmarman.2020.12.014
LinkedIn Marketing Solutions. (2023). The B2B marketing benchmark report. LinkedIn Business.
Meta. (2024). About the Conversions API. Meta Business Help Center. https://www.facebook.com/business/help/conversions-api
Pauwels, K., & Reibstein, D. J. (2022). The marketing analytics stack: Closing the gap between data and decisions. Journal of Marketing, 86(3), 5–24.
Salesforce. (2024). State of marketing report (9th ed.). Salesforce Research.
SiriusDecisions / Forrester. (2022). Demand unit waterfall: Aligning marketing and sales around revenue. Forrester.
Steenburgh, T., & Ahearne, M. (2021). How to sell new products: Lessons from B2B sales force effectiveness research. Harvard Business Review, 99(4), 80–89.
Sun, B., Wang, S., & Zhang, Y. (2023). Lead scoring optimization using machine learning: Evidence from B2B SaaS firms. Decision Support Systems, 165, 113885.