A campaign that produces a thousand cheap leads and a campaign that produces a hundred and fifty expensive ones can look very different on a dashboard and identical in the bank. Whether marketing contributes to revenue depends less on any single campaign than on five capabilities working together: knowing which accounts are worth winning, reaching the right people inside them, saying something useful at the stage they are actually at, keeping demand creation and demand capture running at the same time, and measuring the result against pipeline rather than against form fills. This article sets out those five, what each one fails to do on its own, and the questions to ask before approving the next campaign.
The lead-volume trap
Compare two campaigns honestly. Campaign A brings 200 leads, of which 10 become qualified opportunities. Campaign B brings 80 leads, of which 16 become qualified opportunities. Campaign A wins on volume and on cost per lead. Campaign B produces more actual business.
Even that comparison is incomplete, because it says nothing about deal size, win rate or sales effort consumed. But it is enough to show the mechanism that causes the damage: advertising platforms optimise towards whatever you tell them is a conversion. Tell a platform that a form fill is the goal and it will become very good at finding people who fill in forms. Some of those people were never going to buy anything.
Google’s own direction of travel points the same way. In mid-October 2023 it retired four attribution models from Google Ads and Analytics, leaving data-driven attribution and last-click, on the grounds that fewer than 3% of conversions used the retired ones. The industry has been moving away from simple proxies for years. The harder change is not which model you choose but which event you call a conversion.
The attribution side of this, including what connecting the CRM actually involves and why AI answers make measurement harder, is covered separately in full-funnel attribution for B2B. This article is about the system that sits upstream of the measurement.
Capability 1: knowing who is worth winning
Everything downstream inherits the quality of this decision. If the account list is wrong, precise targeting and excellent creative will efficiently reach the wrong companies.
The inputs are already in the business. Which customers renewed and which churned. Which deals closed quickly and which consumed six months of sales time for nothing. Which industries produced referenceable work. That history is more useful than any purchased intent list, because it reflects your actual delivery rather than a vendor’s model of your market.
The question: do we know which customers are most valuable to us, and can we say why, in terms other than revenue size?
Capability 2: reaching the whole buying group
A B2B purchase of any size involves several people with different jobs, and reaching one of them well is not the same as convincing the group. Targeting usefully splits into three layers.
| Layer | What it asks | How it fails |
|---|---|---|
| Account relevance | Are these the companies that fit how we actually win? | A list built from revenue potential alone, ignoring whether the work suits us |
| Stakeholder relevance | Are we reaching the people who evaluate, approve, use and pay? | Speaking only to the economic buyer, or only to practitioners |
| Intent relevance | Is there evidence this account is researching the problem now? | Treating intent signals as certainty, or waiting for them before doing anything |
The usual mistake is over-narrowing. Layering every filter at once can leave an audience too small to deliver reach, which starves the campaign of the exposure it needs before anyone recognises the company name. A workable split is to run focused campaigns against priority accounts while keeping a broader educational campaign alive for buyers who are not yet identifiable at all.
The question: are we reaching the people who will decide, or only the people who are easiest to target?
Capability 3: saying something useful at the right stage
One advertisement cannot answer every question a buying group has, because the questions are different at each stage and often held by different people. This is where most B2B creative underperforms: it treats a first encounter and a final shortlist as the same conversation.
| Stage | The buyer’s actual question | What works | What to stop doing |
|---|---|---|---|
| Awareness | Is this a problem I should care about? | Expert commentary, original analysis, a distinctive point of view on the problem | Leading with a demo request |
| Consideration | What are my options and how do they differ? | Comparison pages, frameworks, methodology notes, webinars | Claiming to be best without showing the criteria |
| Decision | What will this cost me, in money and in risk? | Pricing clarity, implementation detail, customer evidence, named references | Another broad awareness message |
| Retention and expansion | Am I getting value, and what else should I use? | Onboarding material, customer education, product updates | Treating the signature as the end of the journey |
Two practical notes. The asset with the highest click-through rate is often not the one that appears in closed-won journeys, because the formats that create demand and the formats that close it do different work. And the decision-stage questions are the ones most companies answer least well: a page with no price and no implementation detail cannot be used by the person building an internal business case.
For the page-level craft, see writing comparison pages, service pages that get cited and conversion-ready pages. For what makes a point of view worth publishing at all, this piece goes further.
The question: does our content help the buyer make progress, or does it mainly describe us?
Capability 4: creating demand while capturing it
Search campaigns, retargeting and conversion advertising all work on demand that already exists. They are efficient precisely because the buyer has already decided they have a problem. What they cannot do is produce that decision in someone who has not made it yet.
A business running only capture activity will see good cost per lead and a slowly shrinking pool, because nothing is replenishing the set of people who know the company exists. A business running only demand creation will build awareness it fails to convert. The two need to run together, and they need to be judged differently.
| Demand creation | Demand capture | |
|---|---|---|
| Job | Make the problem legible and the company familiar | Convert existing intent into a conversation |
| Typical channels | Thought leadership, paid social, original research, AI search visibility | Branded and non-branded search, retargeting, direct outreach |
| Fair measure | Reach among target accounts, branded search volume, share of answer in assistants | Qualified pipeline, cost per opportunity, win rate |
| Unfair measure | Cost per lead this month | Credit for demand it did not create |
The last row is where budgets usually go wrong. Capture channels look efficient partly because they harvest demand that other activity created, and the reporting hands them the credit because they were closest to the conversion. Judging an awareness campaign on immediate lead volume, and a conversion campaign on nothing but cost per lead, reliably produces a decision to cut the thing that was feeding the thing that works.
AI search sits mostly on the creation side now. An assistant that names your company when a buyer asks who the credible providers are has done demand creation work, and it leaves no click behind. The Pew Research Center, tracking 900 US adults through March 2025, found that when a Google AI summary appeared, people clicked a source inside the summary on just 1% of visits. For the mechanics of being named at all, see how AI search engines choose what to cite, and for a baseline you can run yourself, the 15-minute visibility audit. On how to think about the budget split, see GEO or performance marketing.
The question: if we stopped all demand creation this quarter, how long before capture performance fell, and would we be able to tell why?
Capability 5: measuring against pipeline
The fifth capability is what makes the other four improvable. Without it, the team is guessing which of the first four is weak.
The short version: pass a source identifier into the CRM and keep it on the record, mark opportunities and closed revenue against it, and report pipeline by source instead of leads by source. That single change alters most budget decisions and needs no new tooling. One distinction is worth keeping clean: marketing-sourced revenue means opportunities attributed to marketing under a stated rule, while marketing-influenced revenue means opportunities that had qualifying marketing contact. They overlap, so they cannot be added together.
The fuller treatment, including conversion import, offline conversions and the honest limits of any attribution model, is in full-funnel attribution, with the arithmetic in calculating SEO and GEO ROI and the reporting format in reporting to leadership.
The question: can we explain what is working, why, and what we would change next, without using the word impressions?
Why the five have to be connected
Describing these as five connected capabilities is Brilad’s framing rather than an industry standard, and the reason for framing it that way is that improving one in isolation tends not to show up in results.
| If this is weak | The symptom you will see | What teams usually blame instead |
|---|---|---|
| Intelligence | Leads arrive and sales rejects most of them as a poor fit | Creative, or the platform’s targeting |
| Targeting | Good accounts engage, but deals stall with no clear reason | Lead quality |
| Creative | Reach is fine, engagement is weak, nothing progresses | Budget |
| Demand balance | Capture costs rise quarter on quarter with no obvious cause | Platform auction prices |
| Measurement | Everyone has a different number and the argument repeats monthly | Each other |
That last row is the most common and the most expensive. A marketing and sales disagreement about lead quality is usually a measurement disagreement in costume, and it stays unresolved as long as the two teams are counting different objects at different moments.
The checklist before approving a campaign
- What business outcome is this campaign expected to move, stated before launch?
- Which accounts and profiles does it target, and what evidence says those are the valuable ones?
- Which members of the buying group will it actually reach?
- Does the creative match the stage those people are at, or only the stage we wish they were at?
- Is this creating demand, capturing it, or quietly doing neither?
- Will marketing contact end up on the CRM record, and can we find it later?
- Are we measuring qualified pipeline, or leads?
- If this works, will we be able to say why, and if it fails, will we know which of the five capabilities was at fault?
- Have marketing and sales agreed what counts as a qualified lead, in writing?
If several answers are no, the useful next move is usually not a larger budget. It is to connect what is already being spent.
What this framework does not do
It does not predict revenue. The five capabilities describe where a marketing system tends to break, which is diagnostic rather than forecasting. Nothing here supports a claim that fixing capability three produces a given percentage increase.
It also does not shorten a sales cycle. If the real cycle is nine months, the feedback on any change arrives in nine months, and compressing the review window simply means judging the change before the evidence exists. Teams that take measurement seriously have to accept slower learning as part of the deal.
And it is not a substitute for a good product and credible delivery. Marketing that successfully puts an unready business in front of serious buyers accelerates the discovery that it is unready.
Frequently asked questions
Which of the five should we fix first?
Measurement, almost always, because until pipeline is reported by source the other four are being judged on the wrong evidence. The exception is when sales is rejecting most leads as a poor fit, which points at intelligence and targeting and does not need better measurement to diagnose.
Is cost per lead a useless metric?
No, it is a useful efficiency check within a channel over time. It becomes misleading when used to compare channels that do different jobs, or when it is the number that decides budgets. Efficiency at acquiring the wrong leads is not a result.
How much budget should go to demand creation?
There is no defensible universal split, and anyone quoting one is guessing about your market. The practical approach is to establish what capture performance looks like today, change the balance deliberately in one direction, and watch branded search and qualified pipeline over a period at least as long as your sales cycle.
Does this apply to a small company with one marketer?
Yes, with the scope reduced. One person can hold a short list of target accounts, one stage-appropriate asset per stage, and a source field on the CRM record. The capabilities are responsibilities, not headcount.
How does AI search change the five?
It mainly changes capability four, by adding a demand creation channel that influences buyers without producing a visit, and capability five, by adding something that has to be measured indirectly. The first three are unaffected in substance: knowing who to win, reaching the group and being useful at the right stage were already the job.
Should we hire an agency for this?
It depends which capability is weak. Measurement and technical connection are often faster to buy in, while intelligence about which customers are valuable usually cannot be outsourced because it lives in your delivery history. If you do brief outside help, this RFP template covers what to ask for.
Sources
- Google announcements on the retirement of the first-click, linear, time decay and position-based attribution models in Google Ads and Google Analytics 4, mid-October 2023
- Pew Research Center, “Google users are less likely to click on links when an AI summary appears in the results”, 22 July 2025, based on the tracked browsing of 900 US adults during March 2025
- Google Ads and Google Analytics Help documentation on conversion import and offline conversion tracking