Testing Audience Ads for B2B Pipeline Without Last-Click Bias
Adventure Works Cloud is a B2B software company spending $15,000 per month entirely on Microsoft Advertising search, with a $180 CPA and a 4.2% demo-request conversion rate. The team is considering Microsoft Audience Ads using LinkedIn profile targeting: job function IT, company size 1,000+, and seniority manager or above. Forecasts suggest a $9.20 CPM, 0.32% CTR, and 0.9% landing-page conversion rate, implying a roughly $320 last-click CPA. A 30-day site visitor remarketing list has 38,000 users. The sales cycle is 90 days, and CRM data shows 60% of closed-won deals touched at least three paid channels. The CFO will only approve a test if reported CPA stays below $200. The team has never run native ads.
1. Should Adventure Works shift 20% of search budget to Audience Ads? Address the CFO's CPA constraint and propose a measurement design.
Read the worked answer
Do not shift 20% of search budget yet. At the forecast $320 last-click CPA, the CFO's $200 constraint will be violated by definition, and last-click is the wrong scorecard for a 90-day B2B cycle where 60% of closed-won deals touch three or more paid channels. Audience Ads is upper-funnel; it will look expensive last-click even if it creates pipeline. Propose a fixed test budget of $2,500-$3,000 per month for 90 days, funded from a test reserve rather than by cutting proven search. Hold search budgets steady so the test is not confounded. Measure: incremental demo requests via a geo or audience holdout, CRM-matched leads and pipeline, assisted conversions, view-through conversions with a stated window, and cost per qualified opportunity. Report both last-click CPA and pipeline-based cost. If the CFO requires a CPA below $200, agree in advance that the metric is cost per qualified lead or cost per pipeline dollar, not last-click. Shift budget only if the test shows incremental pipeline at an acceptable cost.
2. How would you structure the Audience Ads campaign and targeting? Identify two platform-specific pitfalls.
Read the worked answer
Structure two campaigns: prospecting with LinkedIn profile targeting (job function IT, company size 1,000+, seniority manager+) and remarketing to the 38,000 site visitors with a frequency cap. Split ad groups by job function or solution so creative and landing pages match. Use native ad copy and a landing page built for the audience, not a search landing page. Two platform-specific pitfalls: (1) LinkedIn profile targeting segments can be too small or unavailable in some markets, so check audience size and have in-market or remarketing fallbacks; (2) audience overlap between LinkedIn targeting and remarketing can cause over-frequency and inflated view-through conversions. Exclude remarketing from prospecting or cap frequency. Also watch view-through attribution: set a defensible window and report it separately from click-through. Do not judge the test on CTR alone; 0.32% is plausible for native, but pipeline is the outcome.