The current trade-press reframe is a good one: stop measuring lead volume, start measuring relationships, win rates, and long-term pipeline quality. Anyone who has watched a slide of "leads generated" get one skeptical question about pipeline knows the old KPI stack is bankrupt.

The problem is what happens six months after the reframe. Teams adopt "pipeline is the metric" as a slogan and still cannot answer the questions the CFO actually asks. What did marketing contribute this quarter? How much of that pipeline started with a marketing origin? How much did marketing accelerate on the way through the funnel? How much of it comes back next year as expansion? The plumbing underneath the reframe is where teams get stuck.

In every enterprise measurement audit we have run in recent years, the same picture shows up. The dashboard is clean. The stage names look right. The reports render on time. Underneath, four inputs are broken: stage definitions that drift by rep, opportunity ownership that flips mid-quarter, self-reported source that a third of the sales team ignores, and activity capture that misses most meetings. The attribution debate happens on top of that data, in place of fixing it.

How do you measure pipeline contribution from marketing?

A defensible pipeline contribution report has four columns on the same row: sourced (opportunity created from a marketing origin), influenced (touched by marketing before close), accelerated (velocity delta on marketing-touched deals), and expansion (marketing-touched renewals and upsells). Teams that report a single number are flattering themselves; a CFO who asks a follow-up question exposes the gap within a minute.

The four-column report forces honesty in a way a single-number report cannot. Sourced tells the room whether marketing is bringing in new pipeline. Influenced answers whether marketing is showing up in the deals sales closes. Accelerated captures whether marketing is compressing the cycle. Expansion is the read on marketing's role post-close. The four columns give a CFO a real answer where a single number produces a talking point. Only 38% of global marketers say they even evaluate holistic ROI across traditional and digital together, while 84% say they are confident in their ROI measurement (Nielsen, 2024). The confidence has been rising faster than the practice for years. Four columns is what a CFO's follow-up questions require.

The confusion between the first two columns is where every measurement conversation gets stuck, and it is an implementation problem underneath a definitional debate.

Marketing sourced pipeline vs marketing influenced pipeline: what is the actual difference?

The forensic answer is that sourced is a subset of influenced, and enterprise reporting shows both columns side by side using the same stage logic underneath. Sourced pipeline is what marketing originated (a lead form, a demo request, a marketing-attributed contact); influenced pipeline is every opportunity marketing touched before close, whether marketing originated it or not.

Every team we have audited has been arguing about which is correct. Neither is correct in isolation. Sourced is defensible; it maps cleanly to marketing origin. Influenced is honest about how B2B buying actually works: 6 to 10 people on a buying committee, each doing 4 to 5 pieces of independent research (Gartner, 2023), and none of them raise a hand for the ad they saw six weeks ago. Reporting only sourced pipeline understates marketing's role in every enterprise deal. Reporting only influenced pipeline lets marketing take credit for pipeline it did not create. The forensic answer is both columns, side by side, using the same stage definitions, opportunity-owner logic, and source-capture rules underneath. Which is what breaks in most enterprise measurement setups: the underlying inputs behind the stage definitions and the ownership rules.

The definitional debate is a mask over an implementation debate. What most teams call "sourced versus influenced" is closer to "our CRM stage names drifted, and now nothing rolls up cleanly."

How do you measure demand generation ROI?

Two disciplines separate the teams that can defend a demand-gen ROI number from those who guess: source-data governance, and a cost model that includes labor, tools, and content, not just media spend. Given both, the formula is closed-won revenue attributable to demand programs, divided by fully loaded demand cost, over a rolling window that matches the sales cycle.

The trap most teams walk into is a "ROI" number that divides one number (pipeline or revenue) by another number (media spend). That number is a media-efficiency ratio dressed as an ROI number: fine as a supporting metric, disastrous as a headline number for a CFO conversation. Eight in ten advertisers now run marketing mix modeling and brand-lift studies (Ebiquity/WFA, 2026); only 15% say that effectiveness evidence is the primary driver of how budgets get set (Ebiquity/WFA, 2026), and 46% of organizations sit at the lowest maturity levels for integrating the data sources those models depend on (Ebiquity/WFA, 2026). Models exist. The connection between models and budget decisions does not, because the source data is not integrated enough to defend a real ROI number under scrutiny. A defensible demand-gen ROI report has three lines the current version does not: fully loaded cost including headcount, a rolling window that matches the sales cycle rather than a calendar quarter, and a source-of-truth for opportunity origin that both sales and marketing signed off on before the number got calculated.

The ROI conversation is inseparable from a velocity conversation. The teams that can defend an ROI number can also tell a CFO exactly how long each stage takes, because it is the same measurement architecture underneath.

What is pipeline velocity and how do you calculate it?

A team measuring velocity honestly can tell a CFO whether marketing is creating volume, improving quality, compressing cycle, or lifting deal size: four different bets, four different budgets. The formula is (opportunities × win rate × average deal size) divided by sales cycle length, calculated per segment, per source, per quarter. The trendline is more informative than the single number.

The velocity formula is not the hard part. The hard part is that "opportunities" means one thing to the SDR team, another thing to the account executive, a third thing to marketing operations, and a fourth thing to finance. Fix that first. In practice: write the stage definitions on a shared page, get every stage-transition trigger documented, put opportunity ownership on a rule and not a preference, and require self-reported source on every opportunity as a mandatory field. Then calculate velocity. A velocity number calculated on top of drift is a number that looks smooth quarter to quarter while the underlying reality moves. That is the pattern behind every "surprisingly bad Q4." Velocity was moving all year, and the report did not catch it because stage definitions kept sliding. The corollary rule: velocity trendlines are only readable if the source-of-truth for each variable is stable across the window. If someone renamed "SQL" to "Opportunity" three months in, the trendline the CMO is staring at is measurement drift, not business change.

Velocity, ROI, sourced, influenced: the debates that fill quarterly reviews all sit on top of the same four broken inputs.

The four inputs that break before the dashboard does

Every enterprise measurement audit surfaces the same four broken inputs: stage definitions that drift by rep, opportunity ownership that flips mid-quarter, self-reported source that half of sales ignores, activity capture that misses two-thirds of touches. Sixty to seventy-five percent of buy-side leaders say measurement falls short on rigor, timeliness, trust, and efficiency (IAB, 2026). The four inputs are why.

The pattern is a specific one we keep seeing play out across enterprise accounts. Two accounts in the same quarter, same audit method. Both had clean weekly reports going to the C-suite for years. Both dashboards passed a first look. Underneath, the SDR team's definition of "SQL" had diverged from the account executive team's definition. Opportunity ownership was being reassigned inside the pipeline based on who followed up, not who created. Marketing's self-reported source field was blank on roughly 30% of records, because sales had learned to skip it. Activity capture was on for calls but off for most meeting types. The reports looked right and rendered on time. They were also incapable of answering the question the CFO started asking that quarter: how much of this quarter's pipeline is marketing actually responsible for?

Every attribution debate that lands on the wrong answer is downstream of the same failure. When sourced versus influenced does not reconcile between the marketing dashboard and the sales dashboard, the CRM has drift underneath. A demand-gen ROI model that cannot be replicated when the CFO's team runs it points at patchy activity capture. Pipeline velocity that swings quarter to quarter without an explanation from the sales leader is stage definitions moving under the report.

Nielsen's 2024 confidence-versus-practice gap tells the same story from the other direction. 84% of global marketers say they are confident in their ROI measurement, up from 69% two years earlier; only 38% say they evaluate holistic ROI across traditional and digital (Nielsen, 2024). The gap between those two self-reports is the measurement architecture that has not been built.

In every enterprise account we have walked into, the fix has been the same set of moves: a data-governance intervention that shows up as a measurement improvement six weeks later. Picking a new attribution model changes none of it.

What does a clean pipeline measurement architecture look like?

A working measurement architecture has four fixed inputs and one flexible layer. Fixed: stage definitions locked in a shared doc, opportunity ownership assigned by rule, self-reported source as a required field, activity capture on for every meeting type. Flexible: whatever attribution model, MMM, or dashboard the team wants, because they finally sit on data those tools can use.

The order matters. Attribution work done before the inputs are fixed produces a report that looks credible for two quarters and collapses on the third when sales leadership rotates and stage definitions get renamed again. Fix the inputs first. Then whatever model the CMO wants (first-touch, last-touch, W-shaped, MMM, self-reported) works on top. This is the work Moving Parade's Foundations engagement produces before any media plan ships. It is also the reason our client work on measurement architecture holds up under CFO scrutiny six months later: the four inputs are already in place.

One move: Print your CRM's stage definitions this week, walk them to your top three account executives, and ask each one to describe the trigger for moving an opportunity from Stage 2 to Stage 3. If you get three different answers, you have found the reason your pipeline reports drift. Fix that first. Every other measurement conversation is downstream of it.

Which pipeline measurement approach fits which team?

Frequently asked questions

What is the difference between marketing sourced pipeline and marketing influenced pipeline?

Sourced pipeline is opportunities marketing originated: the lead form, the demo request, the first campaign touch. Influenced pipeline is every opportunity marketing touched before close, whether or not marketing originated it. Sourced is a subset of influenced. Enterprise teams report both, side by side, because a CFO asks both questions.

How is pipeline velocity calculated?

Pipeline velocity is (open opportunities × win rate × average deal size) divided by sales cycle length, calculated per segment and per source, ideally per quarter. The formula is straightforward. What breaks the calculation is unstable inputs: stage definitions that drift, opportunity ownership that flips, and inconsistent source capture. Fix those first, then run the formula.

What is a good demand generation ROI benchmark?

There is no universal benchmark that will survive a CFO's follow-up questions. A defensible demand-gen ROI number is closed-won revenue attributable to demand programs divided by fully loaded cost (labor, tools, content, media), over a rolling window matching the sales cycle. Compare your number to your own last four quarters, not to a public benchmark.

Should B2B teams pick sourced or influenced pipeline as their headline metric?

Neither on its own. Report both, side by side, with the same stage and ownership logic underneath. A single-metric headline invites the exact debate that never resolves. Sourced answers whether marketing is creating new pipeline; influenced answers whether marketing shows up in the deals sales is closing. Both answers matter, and a headline needs both columns.

What are the most common measurement failures in enterprise B2B marketing?

Four inputs break before any dashboard does. Stage definitions drift by rep, so "SQL" means one thing to sales and another thing to marketing. Opportunity ownership flips mid-quarter, which prevents sourced and influenced from ever reconciling. Roughly 30% of records have a blank self-reported source field because sales learned to skip it. And activity capture misses most meeting types, understating influenced pipeline by two-thirds.