B2B Demand Generation: What the Published Numbers Measure

What B2B demand generation covers

B2B demand generation is the work that happens before anyone enters your pipeline: awareness, interest and intent across a market that is mostly not buying yet. The arguments are not about that definition but about the statistic each guide picks to explain why the work is worth funding.

The published wordings differ in one place that matters. One stops at the handoff: "creating awareness, interest, and intent across your target market before prospects enter your sales pipeline."

Another takes in the far side too, calling it "the entire range of B2B marketing and sales initiatives that generate interest." Four more sit between the two.

So the category has a settled definition. What it does not have is a settled reason, and the reason is always a claim about the B2B buyer journey.

One neighbouring category settles its own reason with arithmetic instead of a statistic. The fit of product-led growth is decided by dividing a published sales quota by your price.

Where the case gets made with a journey statistic, the figure changes from page to page. One page credits Gartner with 61 percent, a second credits 6sense with 70 percent, and a third carries no percentage at all.

One agency service page keeps its numbers inside a picture of its own product, where nothing can be checked.

Read them in sequence and you get three numbers for one idea, and none of the three cites another. If you are building a case for a budget, you have to pick one, so know which one you picked.

What each of the eight numbers counts

Put those numbers side by side with what each one asked, and the spread stops looking like a trend. A number that counts digital research and a number that counts a stated preference cannot be placed on the same axis, however similar the sentences around them sound.

Six of those descriptions are different questions. A buyer who prefers a rep-free experience may still take a call. A buyer who finds online information sufficient for a shortlist has not finished buying.

Say you are deciding whether to staff a sales development team, the group that also drives B2B lead generation. The 61 percent preference figure argues for fewer callers. The 70 percent journey figure argues for the same thing. The 2024 finding that buyers start 80 percent of first contacts argues for someone being there to answer.

Eight figures, and three come from one 2011 study, and only one readable page prints how it worked. It surveyed "1,500 decision makers and influencers in 22 major B2B companies who had recently been involved in purchasing decisions," asking how far through the process they were before engaging a supplier sales rep.

That is the whole reason 57 and 65 both circulate. They come out of one study, quoted as if they were two findings from two places.

You can use a small study. Say it is small when you do, and your reader can weigh it for themselves.

Where the 67 percent statistic came from

The most quoted figure in this category came with a correction from its own author, in the same year it started circulating. The correction targets exactly the sentence those guides still write today.

Megan Heuer, then VP Research at SiriusDecisions, published "Three Myths of the 67 Percent Statistic" on 3 July 2013. Her first line sets the approved wording: "67 percent of the buyer's journey is now done digitally."

Then she names the misreading. "The 67 percent statistic in no way says that no one talks to a salesperson before getting halfway through the buying cycle, but this is how some have interpreted it."

The correction, dated the same year as the statistic

1. 1The wording the firm approved is about digital research, not about avoiding salespeople.
2. 2The first myth listed is the reading that most pages still print.

The three myths read as instructions for a demand generation programme. That is why they sit here in the body and not in a footnote.

1. Sales is not locked out

   The first myth is that sales does not get involved until more than halfway. Heuer writes that the standard is higher for sales to add value, not that the conversation is closed. Her instruction to marketers is to help sellers engage early with messages, content and tools.

2. Waiting is not a strategy

   The second myth is that it is best to let buyers find you when they are ready. The post lists what to do instead: account profiling, propensity modelling, segmentation and targeting, plus search and activity patterns that show where interest is forming.

3. Inbound is not finished at the top

   The third myth is that your current inbound strategy needs no updating. The instruction is to stop treating inbound as an awareness tool: "Do not ignore the role of inbound marketing at later stages of the buyer's journey," including for buyers looking for validation instead of education.

Compare that with how the figure travels now. One guide writes that "70% or more of the buyer's journey occurs before a prospect ever speaks to a sales representative." That is the sentence Heuer identified as myth one, carried thirteen years and one number later.

Why the same 70 percent has three sources

The figure travels in a 70 percent version too, on three separate pages, and each of the three credits it somewhere else. One of them credits it to two different organisations within a few paragraphs of itself.

None of that makes the underlying figure wrong. It makes the figure unusable as evidence until you open the source and see what it counted.

The attribution is printed, which is more than most pages do

1. 1The bracket names a publisher and a year, so the claim can be traced.
2. 2The paragraph above it borrows a different statistic from a different publisher.

Printing the source is the minimum, and that page does it. The test that follows is whether the source measured your question.

One current measurement sits beside all eight of them. Forrester released its State of Business Buying research in December 2024.

Thirteen people inside the organisation are involved in the buying decision, and 89 percent of purchases cross two or more departments. That release does not publish its sample, so read the figures as the shape of your buyer’s room and not its size.

Notice what it does to your question. Every number above asks how far along one of your buyers is, and this one counts how many of them there are.

Forrester answers that by dropping the individual. Its B2B Revenue Waterfall asks you to move beyond marketing-qualified leads and to advance opportunities with connected buying groups instead.

For example, thirteen people from one account filling in thirteen forms is one opportunity on that reading, and thirteen leads on yours. Which of the two your report counts is a decision, and the 2024 report has a second finding about it.

Who makes the first contact

Read past the headline figure in the trade coverage of the 2024 report and a second finding appears, one that almost never travels with the first. Everything below is quoted from that coverage, and Heuer had written it in 2013.

A trade title covered the research on 10 October 2024: "B2B buyers are nearly 70% through their purchasing process before engaging with sellers, and 80% of the time, it's the buyers who initiate the first contact."

The same report, the finding that rarely travels

1. 1Contact happens. The report says the buyer usually starts it.
2. 2It also reports 81 percent of buyers having a preferred vendor at first contact.

The same coverage lists three more figures from that report. Eighty one percent of buyers had a preferred vendor at the time of first contact, and 85 percent had already established purchase requirements. The average buying cycle ran 11.3 months, across a buying group averaging 11 people. The coverage prints no sample size for any of them.

Those numbers change your job. If most buyers arrive with a preferred vendor already chosen, your question is not how to capture a form fill faster.

Your question is what put a vendor on that list months earlier, and whether you were doing it then. That is a different job from lead nurturing, which starts after someone is already known to you.

What to build when the buyer arrives already decided

Four instructions survive once buyers arrive with a vendor already chosen. Each rests on a sentence a named source printed, not on a figure whose unit is unclear. You can act on all four without first deciding which percentage is true.

1. Be on the list before the research starts

   The 2024 coverage reports 81 percent of buyers holding a preferred vendor at first contact and 85 percent having set requirements already, without printing a sample size for either. The work that matters is what your buyer read months earlier, so plan the awareness programme against the 11.3 month cycle that coverage gives, not against a quarter.

2. Write for the late stages too

   Heuer's third myth is the one most programmes still fail. Look at what you publish for a buyer who is validating a decision instead of learning a category: implementation detail, pricing logic, security answers, comparisons. One publisher reports 87 percent of tech buyers saying credible non-biased sources are essential when building a shortlist.

3. Give sales a reason to be early

   Nothing in these figures says the conversation is closed until 70 percent. Heuer asks marketers to equip sellers with messages, content and tools so they can be useful early, and names the standard that raises: sellers now have to add value the buyer could not find online.

4. Reach a buying group, not a champion

   The average buying group in that 2024 report is 11 people, so a single persona and a single asset will not cover it, whatever stage of the B2B marketing funnel they sit in. The alternative is to route an account to sales with the roles engaged and the content consumed already attached.

None of the four is a lead nurturing sequence, and none of them requires you to settle which percentage is right. They follow from the three parts nobody disputes: the research is long, the group is wide, and contact happens.

What a demand generation programme costs

Read all nine and you will find pipeline results, a budget share and product pricing menus. What none of them prints is what a demand generation programme costs to run, so the answer below is built from the three costs they do name.

Time is the first one, and one vendor states it plainly: "a full engine typically takes 6 to 12 months to show significant impact on the sales pipeline."

The cycle underneath may be longer still. The 2024 trade coverage puts the average B2B buying cycle at 11.3 months without printing a base for it. If that average holds anywhere near your category, a programme judged at six months is being judged before its first cohort has finished buying.

People are the second cost. One agency ran its own first year with two full time staff. Its wording on handing the work off: lead gen "can be easily outsourced," while demand gen "can be outsourced only partially and requires close collaboration with your marketing team."

Media is the third, and it is the one you can price today. Media spend is also what sets B2B lead generation cost per lead in this category. The head term for this category carries an average cost per click of 86.32 dollars, which is what Google Ads reported for United States English search in September 2026. A hundred clicks is 8,632 dollars, before anyone fills in a form.

That number moves with your category, not with your skill. A term this expensive means your buyers are worth a lot and your competitors know it. That is the argument for earning the position instead of renting it.

Your timing question sits one step behind your budget question, and it has its own evidence. For example, when you are told six to twelve months, ask what that number was measured on. B2B content marketing sets out what each published figure counts.

One guide prints a budget figure and credits it to a 2025 survey: "26% of enterprise organizations across all industry segments now invest over 70% of their marketing budgets on demand generation." That is a sentence on a page with a survey behind it, so open the survey before you treat it as a target for your own split.

For instance, spending yours against a named threshold and not a share is what our B2B PPC service sets first.

Five mistakes to check in two minutes

Every mistake below is visible on a page cited in this piece, which makes each of them cheap to check and cheap to avoid. Four are habits of writing, not failures of budget. The fifth is a reading habit that costs nothing to change.

The first is quoting a percentage without its unit, and it costs you a strategy built on the wrong question. Sixty one percent of buyers preferring a rep-free experience and 67 percent of the journey being digital point at different work.

Write the unit beside the number in your own decks and the mistake stops travelling through your team. For example: "61 percent say they prefer a rep-free experience, Gartner via a vendor blog" tells your reader what they may do with it.

The second is waiting for the buyer to arrive. That is Heuer's second myth, and it costs you every account that never raises a hand, whatever the B2B buyer journey looks like in your category. She names the alternative: account profiling, propensity modelling, segmentation and targeting, plus reading search and activity patterns for interest that has not surfaced. Run that list against a named set of companies and you are doing account-based marketing. That is a different budget shape, and the fit test comes before the commitment.

The third is publishing only top of funnel material, so the later stages of the B2B marketing funnel have nothing to read. A buyer validating a shortlist wants implementation detail, security answers and comparisons, and meets a category explainer instead. That same 87 percent of tech buyers say credible non-biased sources are essential at exactly that moment.

The fourth is reporting a metric your reader files under something else. MQL volume sent up as demand generation performance reads as lead generation performance to anyone who learned the split the other way. The cost is a budget conversation about the wrong number.

The fifth is quoting a case study from its introduction. One published case offers two versions of its own first year, and the larger one sits in the summary. Open the body before you repeat a result, including your own.

None of these five needs a tool or a budget to fix. Four are edits to how you write a number down.

Decide which side MQL sits on before you report anything

Writing the number down exposes a second disagreement, and this one costs money instead of credibility. The metric most teams report is placed by published guidance on both sides of the line it draws between demand generation and lead generation.

None of these pages defines MQL operationally, so nothing shows they count it differently. One name ends up on both sides of a line each of them drew. The definitions that do exist sit in what a marketing qualified lead is and what a sales qualified lead is.

That still costs you. Report MQL volume as demand generation performance to a board that learned the split from one of those pages, and you are answering a question nobody asked.

Pick a side, write your definition down where your reader can see it, and report against that. The definitional half of the same argument, demand generation vs lead generation, splits five ways on its own. If the two labels are the argument, the difference between an MQL and an SQL is the place to settle it, and lead scoring is how the label gets applied without a meeting.

One agency states the harder problem directly: demand generation campaigns "require a mix of self-attribution, customer interviews, and digital tracking which most companies don't have." Its practical fix is to ask buyers how they heard about you and put that answer next to the analytics.

One more habit is worth building, and the same agency page supplies the example. Its introduction claims 72 inbound opportunities and over a million dollars of pipeline in year one with two staff, and further down the same page says 63 opportunities and $600k.

Both sentences sit on the page and cover the same first year. Neither is hidden. A reader who quotes the introduction inherits a number the body of the page does not support, and nothing in the quotation shows it.

So read the body before you repeat the headline. That applies to your own case studies as much as to anyone else's, and it is the cheapest quality check in this category.