Demand Gen Metrics That Actually Predict Pipeline: A B2B Framework

Most B2B dashboards are built to make marketing look busy, not to help the business make decisions. You've got thirty metrics on yours. Leads generated, cost per lead, impressions, click-through rate, email open rate, engagement rate. All of it moves. None of it predicts whether you're going to hit pipeline.

This is not a new problem. It's just an expensive one that keeps getting more expensive.

Why Vanity Metrics Still Own Your Dashboard

The short answer: they're easy to collect and easy to report. MQLs feel like progress because the number goes up. Clicks feel like traction. Impressions feel like reach. Nobody gets fired for showing a slide where all the bars are trending upward, even if nothing downstream is moving.

The deeper problem is that most marketing teams inherited their reporting structure from a world where the funnel was linear and attribution was impossible. You generated leads, sales worked them, and marketing took credit for whatever happened at the top. That model rewarded volume over quality, and dashboards were built to reflect it.

Here's the compounding issue: leadership started asking harder questions. CFOs got involved. Boards wanted ROI, not reach. And rather than rebuild the measurement system, most teams just added more metrics on top of the ones they already had, turning dashboards into walls of numbers that answer nothing.

As we've written before, the MQL is a comfortable lie. It feels like accountability. It isn't. What follows are the five metrics that actually hold marketing accountable to the outcome everyone actually cares about: pipeline.

The 5 Metrics That Actually Predict Pipeline

1. Pipeline Velocity

Pipeline velocity measures how fast qualified deals are moving through your funnel, expressed in dollars per unit of time. The formula is: (number of qualified opportunities × average deal size × win rate) divided by average sales cycle length.

Why it predicts pipeline: a drop in pipeline velocity tells you something is wrong before you miss the quarter. Either opportunities are stalling, deal size is shrinking, win rate is slipping, or cycles are lengthening. It's an early warning system, not a lagging indicator.

Track it weekly. If velocity is declining for three weeks in a row, you have a problem. If it's accelerating, you're doing something right, and the job is to figure out what and double down on it.

2. Cost Per Pipeline Dollar

Not cost per lead. Cost per pipeline dollar: how much did you spend in marketing to generate one dollar of pipeline? If you spent $100,000 last quarter and generated $2M in pipeline, your cost per pipeline dollar is $0.05. That's a number a CFO can work with.

This metric forces you to connect spend to pipeline, not to activity. It also immediately surfaces channel efficiency in a way CPL never can. Paid LinkedIn might have a lower CPL than paid search, but if LinkedIn-sourced deals convert to pipeline at half the rate, your cost per pipeline dollar from LinkedIn is twice as high. The headline metric was lying to you.

3. Win Rate by Source

Win rate is almost always reported in aggregate. That's not useful. A 22% overall win rate tells you nothing. Win rate broken down by lead source tells you everything.

If your content-sourced deals close at 31% and your paid social deals close at 14%, you have a clear reallocation signal. If inbound close rate is 3x outbound, your demand gen strategy should reflect that. If one ABM cohort is closing at 40% and everything else is at 18%, that cohort is your proof of concept for a bigger bet.

According to LinkedIn's B2B Institute, at any given moment only about 5% of your total addressable market is actively in-market. That makes close rate by source one of the sharpest signals you have for identifying which channels are reaching in-market buyers versus people who fill out a form and disappear.

4. Marketing-Sourced Pipeline Percentage

What share of your total pipeline originated from a marketing touch? This is the single most important number for justifying the marketing budget, and most teams report it inconsistently or not at all.

The challenge is attribution. First-touch, last-touch, and multi-touch models will all produce different numbers, and each one is wrong in its own way. What matters is that you pick a model, stick to it, and use it consistently enough that quarter-over-quarter comparisons are meaningful.

Marketing-sourced pipeline percentage also gives you a defensible answer to the CFO question: "what did marketing contribute this quarter?" The answer should not be impressions or MQLs. It should be a dollar amount with an attribution model behind it.

5. Time-to-Close by Campaign Cohort

This one gets missed almost everywhere. Average sales cycle length is a useful baseline, but the real signal is how campaign-sourced deals compare to baseline. If deals sourced from your Q1 content campaign take 45 days to close versus a 70-day company average, that campaign is pulling better-qualified buyers. If a specific webinar cohort takes 90 days when the baseline is 70, that event is generating curiosity, not intent.

Time-to-close by cohort also helps you set realistic pipeline forecasts. A deal sourced from a branded search campaign is likely to close faster than one sourced from a top-of-funnel display campaign. Treating them identically in your forecast is the reason forecasts are wrong.

How to Instrument These in HubSpot and Salesforce

The good news: both platforms can surface all five of these metrics without a custom data warehouse. The not-so-good news: it requires some intentional setup that most teams skip.

Pipeline Velocity

In HubSpot, build a deal-based report using the Deal Amount, Close Date, and Create Date properties. You can calculate cycle length manually or use the "Days to Close" property. For win rate, filter by "Deal Stage = Closed Won" versus "Closed Lost." Salesforce has a native Opportunity pipeline report with similar fields. You'll want to track this as a time-series report so you can spot trend changes week over week.

Cost Per Pipeline Dollar

This requires connecting your ad spend data to your pipeline data. In HubSpot, pull total campaign spend from the Campaigns tool (or import from a spreadsheet if your paid platforms aren't integrated), then divide against pipeline generated by contacts tied to that campaign. In Salesforce, the Campaign Influence model handles this natively if campaigns are set up correctly. If you're using UTM parameters consistently, this becomes much cleaner regardless of platform.

Win Rate by Source

In HubSpot: use the "Original Source" or "Latest Source" property on the contact level, then join to deal records. Build a deal report filtered by Closed Won vs. Closed Lost, grouped by contact source. In Salesforce: the "Lead Source" field on the Opportunity object does this natively. The challenge is data hygiene. If 40% of your deals have "Other" or blank as the lead source, your win rate by source report is meaningless. Fix the data before you trust the report.

Marketing-Sourced Pipeline Percentage

In HubSpot: create a pipeline report filtered to deals where the contact's original source matches your marketing channels (organic search, paid search, paid social, email, etc.). This gives you marketing-sourced pipeline in dollars, which you divide by total pipeline. In Salesforce: Campaign Influence is the right model here, though you'll need to decide on first-touch vs. multi-touch attribution before you run the report.

Time-to-Close by Cohort

In both platforms, this means tagging deals by the campaign or source that originated them, then reporting on average days to close, segmented by that tag. In HubSpot, use the "Original Source Drill-Down" properties to get to the campaign level. In Salesforce, you'll use Campaign Member records tied to Opportunities. The key is consistent campaign tagging upstream, which is a UTM discipline problem as much as a CRM problem.

The Board-Ready Reporting Template

Your quarterly marketing slide to leadership should answer four questions and nothing else. If it answers more than four questions, it answers none of them.

The four questions:

• What did marketing contribute to pipeline this quarter, in dollars? State the marketing-sourced pipeline total and what percentage of total company pipeline it represents;
• Are we generating pipeline more or less efficiently than last quarter? Show cost per pipeline dollar, current quarter versus previous two quarters, so the trend is visible;
• Which channels are producing pipeline worth closing? Win rate by top three or four sources. Not all channels, not aggregate. The ones that matter;
• What are we doing differently next quarter based on this data? One or two bullets. Reallocation decisions, tests being turned off, bets being increased. If you can't answer this, the report is just history, not strategy.

This is a one-page view. It doesn't require an appendix with 30 supporting metrics unless someone asks. If leadership is asking questions you can't answer with these four data points, the answer is to add those data points to your measurement system, not to add more slides.

The organizations that get this right don't just run better campaigns. They make better decisions, because they've replaced the activity theater with a shared understanding of what moving the business actually looks like. As we've covered in depth, closed-won is the only vanity metric worth chasing. Everything on this list gets you there faster.

The Bottom Line

You don't have a metrics problem. You have a prioritization problem. The tools to measure pipeline velocity, cost per pipeline dollar, win rate by source, marketing-sourced pipeline percentage, and time-to-close by cohort all exist inside HubSpot and Salesforce today. The reason most teams aren't using them is that nobody stopped to ask which numbers actually predict whether the quarter closes.

Start there. Build the four-question board report. Drop three metrics for every one you add. Your CFO will notice, and so will your pipeline.

If you want to surface these metrics across your paid channels automatically, including cross-platform pipeline attribution by campaign, Yirla was built to do exactly that.