The board meeting question every B2B marketer dreads: "So what did we get for the marketing spend?"
You know the work is compounding. Rankings are climbing, the founder's LinkedIn is growing, demo quality is improving. But the deals closing this month started nine months ago, and the deals your current work will close have not happened yet. The spreadsheet looks like an act of faith.
This is the measurement problem of long sales cycles: by the time revenue proves a channel works, you have either doubled down blind or cut a channel that was working.
The answer is not better software. It is a handful of practical habits: measuring leading indicators, capturing attribution honestly, and reporting in a way a board can trust. This guide covers all three.
1. Why Last-Click Attribution Lies in B2B
Most analytics tools default to last-click attribution: the final touchpoint before conversion gets all the credit. In B2B, this produces confidently wrong conclusions.
A typical enterprise buyer's real journey: hears your founder on a podcast, reads three blog posts over two months, sees the founder's LinkedIn content weekly, asks a peer about you in a WhatsApp group, and finally Googles your brand name to book a demo.
Last-click attribution records that journey as: "branded search converted a lead". The podcast, the content, the LinkedIn presence and the referral, the things that actually created the demand, get zero credit. Follow the data literally and you would cut everything except bidding on your own brand name.
Three structural realities make click-based attribution unreliable in B2B:
- Committees, not individuals. Six people influence a typical deal. Your tracking sees one of them.
- Dark social. Slack messages, WhatsApp forwards, podcast listens and word of mouth leave no click trail at all.
- Time decay. Cookies expire and devices change long before a six-month cycle completes. The first touches are usually gone from the data by the time the deal closes.
The conclusion is not to abandon measurement. It is to stop expecting click paths to tell the whole story, and to build the fuller picture deliberately.
2. Leading Indicators Worth Tracking
If revenue lags by two quarters, you need indicators that move within weeks. Three tiers, in order of how tightly they connect to money:
Tier 1: qualified pipeline signals
- Qualified demos booked. The single best leading indicator. Count only demos matching your ideal customer profile; raw demo volume rewards noise.
- Sales-qualified leads and their velocity. How many MQLs become SQLs, and how fast. Improving conversion and shortening lag signal improving lead quality.
- Pipeline value created per month. The currency the board actually cares about, visible months before revenue.
Tier 2: demand signals
- Branded search volume. People Googling your name were influenced somewhere. A rising trend is demand creation working, whatever last-click says.
- High-intent page traffic. Visits to pricing and comparison pages predict pipeline far better than total traffic.
- Engaged target accounts. For narrow ICPs: how many named target accounts visited key pages this month.
Tier 3: activity quality
Rankings for money keywords, ICP-fit LinkedIn engagement, email reply rates. Useful for diagnosing channels, but never report tier 3 as success on its own. It is evidence the engine is turning, not that the car is moving.
3. Self-Reported Attribution: The Highest-ROI Fix
The cheapest, most revealing measurement improvement available: add one required field to your demo form. "How did you hear about us?" As free text, not a dropdown.
Free text matters because dropdowns force answers into your categories. The magic is in verbatim answers: "my CFO forwarded your LinkedIn post", "you came up in a Telegram group", "ChatGPT recommended you". These are the invisible journeys click tracking cannot see.
Setup takes an afternoon:
- Add the required free-text field to every conversion form, demo, audit and contact alike.
- Pipe the answer into a CRM field on the contact record, next to the click-based source field. The two together, first click and buyer's own memory, bracket the truth.
- Tag the answers monthly into channel groups: search, LinkedIn, referral, AI assistant, podcast, event. Thirty minutes of work.
- Report the tagged distribution alongside pipeline numbers every month.
4. The CRM Hygiene That Makes Any of This Possible
Every measurement approach dies without basic CRM discipline. Four prerequisites, none optional:
- Defined lifecycle stages. Written definitions for MQL, SQL, opportunity and customer, agreed between sales and marketing. If the demo threshold is fuzzy, every downstream number is fuzzy.
- Source fields that persist. Original source, captured once, never overwritten, plus the self-reported answer. A CRM that silently overwrites source on every new touch is rewriting history.
- Timestamps at stage changes. Velocity metrics need to know when each transition happened. Most CRMs record this automatically; check yours actually does.
- A closed loop with sales. When deals close or die, the outcome must flow back against the original source. This single connection is what lets you say "content-sourced deals close at twice the rate of paid-sourced deals", the sentence that justifies budgets.
5. A Monthly Report a Board Can Trust
One page, four sections, the same structure every month. Consistency is what builds trust; a new format each month reads as hiding something.
- Pipeline. Qualified demos, SQLs, pipeline value created, and closed-won revenue with its original sources. Each with a trend against the previous three months.
- Demand. Branded search trend, high-intent page traffic, self-reported attribution distribution.
- Channel notes. Two or three sentences per active channel: what changed, what was learnt, what happens next. Sentences, not screenshots of dashboards.
- Cohort view. The section almost everyone omits, and the one that solves the lag problem: show each month's leads as a cohort and track their progression to SQL and opportunity over subsequent months. This is how January's content work visibly becomes July's pipeline, and how you buy patience for channels that compound.
6. When to Kill a Channel and When to Wait
The hardest budget call with long cycles: is this channel failing, or just early? Two questions cut through:
Has it had one full sales cycle plus ramp time? A channel started in January cannot be judged on revenue before autumn. Compounding channels such as SEO and founder-led content need six to twelve months; paid capture shows signals faster. Judging a compounding channel at ninety days guarantees you kill it exactly when it was about to pay.
Are the leading indicators moving in sequence? Healthy channels progress: tier 3 first (rankings, engagement), then tier 2 (branded search, intent traffic), then tier 1 (demos, pipeline). If tier 3 has moved for two quarters but tier 2 has not followed, something real is wrong: targeting, message or offer. If the sequence is progressing on schedule, hold your nerve, and show the board the cohort view while you do.
Kill decisions deserve the same rigour as investment decisions. Write down, in advance, what evidence would justify stopping. It keeps a slow month from executing a channel that six months of data supports.
The Bottom Line
You cannot make a six-month sales cycle report like an e-commerce funnel, and pretending otherwise produces decisions worse than no data at all. Measure leading indicators in tiers, add self-reported attribution this week, keep the CRM honest, and report cohorts so the lag becomes visible instead of invisible.
Pipeline, not vanity metrics, is the standard we hold our own work to. If you want a measurement framework like this wired into your marketing, book a discovery call and we will walk through your funnel together.
