B2B marketing KPIs: the metrics that hold up at the board

B2B marketing KPIs sort into three layers: activity (what marketing did), pipeline (what marketing produced), and revenue (what marketing impacted). The board cares mostly about layer three. Marketing ops manages layer one. Demand gen runs against layer two. Most KPI listicles repeat the same 10-15 metrics with mild reordering. This piece organizes them by funnel stage and accountability layer, names which KPIs hold up under board scrutiny, and surfaces the underlying data-quality dependency: every KPI is measured against an account universe and contact graph that has to actually represent the addressable market.

For LinkedIn-native enterprise and mid-market SaaS, the standard data graph (Apollo, ZoomInfo, Clay, Cognism, Lusha, Demandbase) supports the metrics. For teams selling into local-business, SMB, trades, restaurants, or franchise operators, the same graph covers decision-maker mobile at 10-20% against a discovery-first benchmark of 60%+. The KPIs are accurately measuring engagement against a fraction of the addressable market.

1. The three layers of B2B marketing KPIs

1.1. Activity KPIs

Volume of leads, content produced, campaigns run, sessions, downloads, MQLs. Useful for managing the engine. Weak as standalone proof of impact.

1.2. Pipeline KPIs

Opportunities created, pipeline value, conversion rates by stage, sales cycle length. The middle layer. Bridges activity to revenue.

1.3. Revenue KPIs

Marketing-sourced revenue, marketing-influenced revenue, CAC, LTV:CAC, payback period. The board layer.

2. The top KPIs by funnel stage

2.1. Top of funnel

Sessions by source, organic keyword rankings, share of voice, brand search volume. Activity-layer metrics. Useful for trend tracking, not for proving impact in isolation.

2.2. Mid funnel

MQLs, MQL-to-SQL conversion rate, content engagement depth, target-account engagement (for ABM motions). Connects to data foundation: target-account engagement only computes correctly if the target-account list is right.

2.3. Late funnel

Opportunities created, pipeline value, sales cycle length, win rate, average deal size. The sales attribution lens. Marketing's role is contribution.

2.4. Revenue and customer KPIs

Marketing-sourced revenue, marketing-influenced revenue, CAC, LTV, LTV:CAC, payback period, NRR (net revenue retention), churn. Board KPIs.

3. KPIs that hold up in a board meeting (and ones that don't)

3.1. Hold up

Wide attribution lens (any marketing touch on a closed deal). Defensible to the CFO. Rule of thumb: target 30%+ of closed-won marketing-influenced for B2B SaaS.

3.2. Hold up

Months for marketing plus sales acquisition cost to be repaid by gross margin. Under 12 months is strong. 12-18 months is acceptable for high-LTV ICPs. Over 18 months requires explanation.

3.3. Hold up

Cycle-time difference for deals with marketing engagement vs. without. Underused. Hard to game. Board-credible when the cohorts are clean.

3.4. Don't hold up alone

Inflatable. Useful only paired with MQL-to-SQL conversion plus downstream contribution.

3.5. Don't hold up alone

Lead cheapness can mean lead worthlessness. Use only with downstream conversion and revenue contribution.

3.6. Don't hold up alone

Engagement metric, not outcome metric. Diagnostic value only.

4. KPI ownership

5. The data foundation that makes B2B marketing KPIs trustworthy

Every KPI above is computed against an account universe and contact graph. Three failure modes drive most KPI drift.

Wrong denominator (universe incomplete). When the account universe in the CRM doesn't include 30-50% of the TAM, "ICP coverage" and "marketing-sourced revenue %" report numbers against the wrong base. Common in non-LinkedIn-native segments. Discovery is upstream of enrichment.

Wrong numerator (mis-attributed engagement). When contact records are wrong (mobile coverage 10-20%), engagement attributed to the named decision-maker is often attributed to the wrong person. The metric looks fine. The meaning isn't.

Stale variable drift. Industry codes, employee counts, and tech-stack fields used in segmentation drift over time. KPIs computed by segment shift silently. Decay is faster in local-business segments because of higher closure rates, ownership transitions, and phone or email turnover.

Database-size claims (300M+ contacts in your provider) don't predict whether your TAM is in there with usable detail. The honest test is your 100 target accounts. The manual enrichment tax (about 45 minutes per account by hand vs. about two minutes on a discovery-first stack) is capacity marketing ops eats to keep KPIs trustworthy. That's capacity that isn't running campaigns.

6. How KPIs look different for non-LinkedIn-native ICPs

For LinkedIn-native B2B (enterprise and mid-market SaaS, professional services), the standard KPI set is interpretable because the underlying data graph (LinkedIn plus corporate web, surfaced via Apollo, ZoomInfo, Clay, Cognism, Lusha, Demandbase, 6sense) covers the TAM at usable accuracy.

For local-business, SMB, trades, restaurant, or franchise operators, the same data graph hits an architectural ceiling. About 50% of decision-makers have no LinkedIn presence. Mobile direct-dial coverage runs at 10-20%. The result: KPIs report cleanly, but they're measuring engagement against the LinkedIn-visible slice of the TAM. The other half of the addressable market is invisible to the metrics.

The fix is a discovery-first data layer underneath the existing stack. DataLane complements the LinkedIn-dependent providers, builds the account universe with non-LinkedIn-sourced data (license records, permits, franchise filings, POS detection), and brings the unmeasurable half of the TAM into the KPI denominator. The vendor-churn pattern (a VP cycling through Apollo, ZoomInfo, and Clay annually) doesn't fix the underlying coverage architecture. The source graph is the same across all three.

7. A recommended starter KPI dashboard

Headline KPIs (six): marketing-influenced revenue, CAC, LTV:CAC, pipeline contribution, MQL-to-SQL conversion rate, and ICP account coverage. Supporting diagnostics: MQL volume by source, content engagement by stage, target-account engagement, and sales cycle length with vs. without marketing touch.

ICP account coverage is the hidden KPI that surfaces data-foundation health. Rarely tracked. High-stakes. If your "ICP penetration" denominator is wrong, every metric above it is wrong too.

8. How DataLane fits in B2B marketing KPIs

B2B marketing KPIs assume an account universe the data layer can supply. For LinkedIn-native ICPs, the standard inbound and content KPIs map cleanly because horizontal contact databases cover the segment well. For local-business segments where the funnel begins at outbound discovery, the same KPIs distort because the underlying account universe is partial. DataLane is a discovery-first data layer indexing 17M+ U.S. local business locations from non-LinkedIn sources (licensing boards, permit filings, franchise registries, POS detection, NPI registry). It delivers 60%+ DM mobile coverage at 80%+ accuracy on segments where horizontal providers return 10-20%.

In a KPI design exercise, DataLane changes the upstream inputs more than the downstream metrics: a CAC, ROAS, or pipeline-velocity number measured against a partial TAM is worse than the same metric measured against a complete one. The KPIs don't need to change. The universe they're measured against does. For LinkedIn-native motions, horizontal providers carry the segment cleanly and DataLane isn't needed.

Frequently asked questions

B2B marketing KPIs depend on where the funnel actually starts. For LinkedIn-native ICPs, the inbound and content KPIs map cleanly. For non-LinkedIn segments, the same metrics distort because the funnel begins at outbound discovery, not inbound research. Pick KPIs that match the channel mix the segment actually responds to.