Run a quarterly B2B paid-media audit

Before any export runs, write down the exact quarter being audited (start date, end date, time zone), the full list of channels in scope, and the attribution model that will govern every number in the deck. The audit falls apart the moment two cuts of the same data use different windows or different attribution rules, because every disagreement about a channel verdict traces back to a disagreement about one of those two inputs. Pick first-touch for a top-of-funnel honesty check, last-touch for a bottom-of-funnel ROI check, or a weighted multi-touch model if the CRM supports it, and lock the choice for the whole audit. Document the scope, window, and model in the first slide of the deck so every table that follows reads against the same frame.

• State the exact audit window with start date, end date, and time zone in writing.
• List every paid channel in scope and every channel explicitly out of scope so nothing drifts in later.
• Pick one attribution model, document the lookback window (30, 60, or 90 days), and freeze it for the audit.
• Decide whether sourced pipeline uses created-date or close-date cohorts and record the choice.

Pull a clean export per channel into one normalized spreadsheet with the same columns in the same order: channel, campaign, ad group or ad set, creative, impressions, clicks, cost, leads, MQLs, SQLs, opportunities, and sourced pipeline. Google Ads, LinkedIn Campaign Manager, and Meta Ads Manager all expose this at the campaign level with a date-range filter. G2 and Capterra report traffic, buyer intent signals, and category leads from their vendor dashboards and typically require a manual CSV. Normalize campaign names on the way in so a Google Search campaign and a LinkedIn Sponsored Content campaign line up as peer rows. Platform-reported conversions are useful as a sanity check but never as the primary number, because platform conversion counting double-counts across channels and ignores sales acceptance.

• Export spend, impressions, clicks, and platform conversions per campaign from each ad platform.
• Pull a parallel CRM export of leads, MQLs, SQLs, and opportunities joined on UTM campaign for the same window.
• Add G2 and Capterra traffic, buyer intent, and category lead reports as separate rows on the same sheet.
• Normalize channel labels (paid-search, paid-social-linkedin, paid-social-meta, review-site-g2, review-site-capterra) so cuts work later.

With the normalized sheet in place, compute cost per MQL, cost per SQL, cost per opportunity, and sourced-pipeline-to-spend ratio for every channel. CAC is total channel spend (media plus agency plus platform fees) divided by closed-won customers sourced by that channel in the window, and payback is CAC divided by average monthly recurring revenue per account. Rank channels by sourced-pipeline-to-spend ratio first, then by payback months, because a channel can score well on pipeline multiple but still show a long payback if the deal size is small. Flag any channel where sourced pipeline is below 2 times spend as a candidate for reduction, and any channel above 5 times as a candidate for reinvestment. The ratios are the honest headline the deck gets built around.

• Compute cost per MQL, cost per SQL, cost per opportunity, and CAC per channel in the same table.
• Compute sourced-pipeline-to-spend ratio as sourced pipeline dollars divided by total channel spend.
• Compute payback months as CAC divided by average new MRR per customer sourced from the channel.
• Rank channels by pipeline ratio, flag anything below 2x red and above 5x green.

Pull every active creative (Google Ads responsive search assets, LinkedIn sponsored posts and conversation ads, Meta static and video creatives) and rank them by sourced qualified pipeline, not by click-through rate or platform-reported conversions. CTR tells you whether a creative is interesting, but qualified pipeline tells you whether the audience it attracted is the audience the sales team wants to work. Separate top-of-funnel creatives (educational, broad) from bottom-of-funnel creatives (demo, pricing, comparison) and judge them against the goal of their stage. A great top-of-funnel creative drives MQLs at a reasonable cost per MQL; a great bottom-of-funnel creative drives SQLs and opportunities at a reasonable cost per SQL. Confusing the two is how marketing teams end up with a top-of-funnel wall and a dry bottom.

• List every active creative with impressions, clicks, cost, leads, MQLs, SQLs, and sourced pipeline joined on UTM content.
• Tag each creative as top-of-funnel or bottom-of-funnel based on offer and landing page.
• Rank top-of-funnel creatives by cost per MQL and bottom-of-funnel creatives by cost per SQL.
• Mark the top three winners per channel for scale-up and the bottom three for retirement.

Open the search-term report for Google Ads, the audience and company-size breakdown for LinkedIn, and the placement and demographic report for Meta, and look for the long tail of spend that produced no leads or produced the wrong leads. On Google, flag any search term with more than a defined cost threshold (choose a number appropriate for your average deal size) and zero conversions, and add it to the negatives list. On LinkedIn, inspect spend by job function, seniority, and company size and cut any audience segment outside the ICP. On Meta, review placement spend and cut automatic placements that are soaking impressions on audience network or Reels when the campaign goal is lead gen. The hidden quarter-over-quarter savings almost always live in placement and audience waste, not in a single bad campaign.

• Pull search terms with meaningful cost and zero conversions on Google Ads and add them as negatives.
• Review LinkedIn spend by job function, seniority, company size, and industry and cut non-ICP segments.
• Review Meta placement spend and switch high-waste placements off automatic.
• Document the negative-keyword and audience changes with the dollar amount of expected recaptured spend.

Review-site paid programs behave differently from search and social and need to be judged on their own terms. On G2, pull category traffic, buyer intent signals, profile visits, and category leads from the Seller dashboard and tie them back to CRM opportunities whose first-touch was a G2 referral or whose UTM carried a G2 source. On Capterra, pull PPC clicks, leads, and the per-lead cost from the Vendor Portal and reconcile against the same CRM window. The question that matters is not platform-reported leads but closed-won deals that first touched a G2 or Capterra property, because buyer-intent signals are a top-of-funnel indicator and the real test is what percentage of them turn into pipeline. Benchmark cost per sourced opportunity against your paid-search comparable, since those two channels compete for the same bottom-of-funnel budget.

• Pull G2 traffic, buyer intent, profile visits, and category leads for the window.
• Pull Capterra PPC clicks, leads, and cost per lead for the window.
• Join review-site leads to CRM opportunities on UTM source and first-touch to get sourced pipeline.
• Compare cost per sourced opportunity on G2 and Capterra against paid search as the apples-to-apples peer.

The audit is only useful when it ends in a one-page cut list and a one-page reallocation plan. The cut list names every campaign, ad set, keyword theme, creative, placement, or audience that will be paused or stopped, with the dollar amount of recaptured spend next to each line. The reallocation plan names where that recaptured budget plus any net new budget will go for the following quarter, grounded in the winners identified in steps three through six. Keep both pages concrete: campaign names, dollar amounts, start dates, and owners. A paid-media audit with no cut list is a book report. A paid-media audit with a cut list and a reallocation plan is a budget decision.

• List every item being paused or killed with the quarterly dollar savings next to it.
• List every item being scaled up with the proposed new monthly budget and the expected pipeline contribution.
• Sum the net budget delta and compare it against the quarterly plan so finance sees the full picture.
• Name an owner and a start date next to every line on both pages.

Walk the audit through a single 45-minute readout with the marketing lead, sales lead, revops lead, and the finance partner in the room. Lead with the sourced-pipeline-to-spend ratio by channel, show the top three creative winners and the top three losers, present the cut list and the reallocation plan, and close with the four numbers that will be tracked between now and next quarter: total paid spend, sourced MQLs, sourced SQLs, and sourced pipeline. Set a mid-quarter 20-minute check-in on the calendar before the meeting ends so the reallocation plan is actually reviewed six weeks in, not three months in. The audit becomes a budgeting ritual rather than a one-off slide deck only when the follow-up cadence is on the calendar before the room clears.

• Deliver the readout in 45 minutes with the four GTM leaders present and nobody else.
• Walk the pipeline ratio table, the creative winners and losers, and the cut-and-reallocate pages in that order.
• Lock the four follow-up metrics (spend, MQL, SQL, sourced pipeline) and name an owner for each.
• Put the mid-quarter 20-minute review on the calendar before the room clears.