Nacho Hernández
A good B2B paid media report fits on one page and answers four questions: what you spent, what pipeline that spend created, what changed since last period, and what you will do next. Everything else (platform screenshots, 40-row keyword tables, impression charts) is appendix. If your weekly report cannot be read in five minutes by someone who owns revenue, it is documentation, not a report.
This is the structure I use when I build reporting for B2B teams running Google Ads, LinkedIn Ads and Meta: which sections to include, which metrics earn page one, how to tie platform numbers to pipeline in the CRM, and how to write commentary that leads to a decision. Benchmarks are deliberately absent; they vary too much by deal size and sales cycle to be worth quoting.
What should a B2B paid media report include?
Five blocks, in this order. If a block does not change a decision, cut it.
| Block | What it shows | Why it earns its place |
|---|---|---|
| 1. Headline | Spend, qualified leads (or SQLs), pipeline created, cost per qualified lead, vs. last period and vs. target | Lets an executive stop reading after ten seconds |
| 2. Channel split | The same four numbers per channel and per campaign group (brand, non-brand, retargeting, ABM) | Shows where budget is working and where it only looks busy |
| 3. What changed | The two or three biggest movements, each with a cause | Replaces a wall of charts with a story |
| 4. Tests and learnings | What you tested, the result, the decision | Builds a record so nobody repeats a failed test |
| 5. Next actions | Budget moves, tests, fixes, owner and date | Turns the report into a work plan |
Which metrics belong on page one?
Lead with outcomes the sales team recognises. In most B2B accounts that means qualified leads, meetings booked, opportunities created and pipeline value, all taken from your CRM rather than from the ad platform. Platform-reported conversions usually count form fills, including unqualified ones, and different platforms claim the same lead, so the sum of platforms rarely matches the CRM.
| Metric | Use it for | Watch out for |
|---|---|---|
| Cost per qualified lead | Comparing channels on a like-for-like basis | Define “qualified” once with sales and do not change it mid-quarter |
| Pipeline created per $1 spent | Judging channels against the sales cycle | Recent weeks look weak because deals are still maturing |
| Lead-to-opportunity rate by campaign | Spotting campaigns that attract the wrong people | Needs enough volume; do not read it on 12 leads |
| Impression share (non-brand search) | Checking whether budget or bids are capping you | Only matters if the campaign is profitable |
| CTR, CPC, CPM | Diagnosing why cost per lead moved | Never the headline; cheap clicks are not cheap leads |
How do you connect ad data to pipeline?
The platform sees a click. The CRM sees a deal. The report lives in the join between the two, and it only works if the plumbing is in place before you need it. Four things matter: consistent UTM parameters on every ad (source, medium, campaign, and a content tag), click IDs captured in hidden form fields, original source and campaign stored on the contact and the deal, and conversions sent back to Google and LinkedIn from the CRM once a lead is qualified so the platforms optimise for quality rather than form fills.
In HubSpot this means checking that original source drill-down fields and campaign associations are populated for paid traffic; if they are not, fix that first. My piece on marketing attribution in 2026 covers the model question, and the MMM, attribution and incrementality framework explains when platform numbers should not be trusted at face value.
Be explicit about lag. If your typical cycle from first touch to opportunity is 60 to 90 days, last week’s pipeline number is incomplete by definition. Report leads and cost per qualified lead weekly, and pipeline monthly on cohorts old enough to have matured.
How often should you report, and to whom?
- Weekly, for the operators. Spend pacing, anomalies, tests running, actions for the next seven days. One page, written by the person who runs the accounts.
- Monthly, for marketing leadership and sales. Channel economics, lead quality feedback from sales, budget reallocation proposals.
- Quarterly, for executives. Pipeline and revenue contribution, payback, the one or two strategic bets for next quarter.
If you cover US buyers, build the weekly cut on Monday to Sunday in one fixed time zone and say which. Mixing platform time zones is a common source of numbers that never reconcile.
How do you write commentary that drives decisions?
Use the same four-part pattern for every point: what changed, the likely cause, the action, the owner. “Cost per qualified lead on LinkedIn rose from $210 to $290 because the new audience expansion reached lower-seniority titles; we are restricting to director and above from Wednesday; owner: paid media lead” is a decision. “LinkedIn CPL is up” is an observation. Label guesses as guesses, and flag any metric where tracking changed during the period.
What mistakes make paid media reports misleading?
- Reporting platform conversions as results. Two platforms claiming the same lead means the total is inflated. Reconcile to CRM counts.
- Comparing periods with different lag. This week against the same week last year ignores that last year’s deals have had time to close.
- Averaging brand and non-brand together. Brand search flatters the blended cost per lead and hides weak prospecting.
- Hiding the denominator. A 40% lift on 10 leads is noise. Show volumes next to every rate.
- Changing definitions silently. If “qualified” changes, mark the break in the chart and restate history.
Can you automate the report?
Yes, and you should automate the assembly, not the judgement. Pulling spend and conversions from each platform, joining them to CRM outcomes and flagging anomalies is repetitive work a pipeline or script can do reliably. The narrative and the decisions stay with a person. I describe the sequence in my guide to automating marketing operations with AI. Get the manual version right for a month first; automating a confusing report just produces confusion faster.
Frequently asked questions
What should a weekly B2B paid media report include?
Spend against pacing, qualified leads and cost per qualified lead by channel, the two or three biggest changes with their causes, tests running, and a short list of actions with owners and dates. Keep it to one page.
Which KPIs matter most for B2B paid media?
Qualified leads, meetings booked, opportunities created and pipeline value, measured in the CRM. Clicks, CTR and CPC are diagnostic metrics that explain movement but should not be the headline.
Why do ad platform numbers not match my CRM?
Platforms count form fills and attribute conversions with their own windows and models, and several can claim the same lead. The CRM counts people and deals. Treat the CRM as the source of truth and use platform data to explain changes.
How long should I wait before judging B2B campaigns on pipeline?
It depends on your sales cycle. If first touch to opportunity usually takes 60 to 90 days, judge pipeline on cohorts at least that old, and use cost per qualified lead as the weekly leading indicator.
Can I build this in Looker Studio or HubSpot?
Yes. Either works if UTM discipline and CRM source fields are in place. The tool matters less than agreeing the definitions and the join between ad spend and CRM outcomes.
Want paid media reporting that leads to decisions?
I help B2B teams connect ad spend to pipeline and build reporting that sales and finance trust. Start with a short call.
