From Ad Click to Revenue: Attribution for Lead Gen

Why "attribution" usually means nothing in B2B lead gen

Most B2B teams think they have attribution. They don't.

What they have is a Google Ads dashboard showing cost per click, a Meta dashboard showing cost per lead, and a HubSpot report showing how many MQLs came in last month. None of those numbers connect to the closed-won deal sitting in Salesforce. None of them tell you whether the $40,000 you spent on LinkedIn last quarter produced $200,000 in revenue or $20,000.

The reason is mechanical. Ad platforms track the click and the form fill. Your CRM tracks the opportunity. The bridge between them is usually broken: half-assed UTMs, no source field on closed deals, and a sales team that books meetings in Calendly, not in the CRM.

You don't need a fancy attribution platform to fix this. You need a few specific things wired up correctly, and one honest model.

This is what we build for clients in our CRM and funnel audits and follow-up systems. Here is the operating version.

The four pieces you actually need

Before you talk about models, you need data. Here are the four pieces, in order of how much pain they cause when missing.

1. UTM discipline on every link.

Not "sometimes." Not "on the big campaigns." Every paid link, every email, every LinkedIn post that drives traffic to a tracked page. Standard taxonomy: utm_source, utm_medium, utm_campaign, utm_content, and utm_term where it applies. Bake the format into a template your team copies from, not a doc they may or may not read.

If your UTMs are inconsistent, stop reading and fix that first. The rest of attribution is built on this.

2. Hidden source field on every form.

Most B2B forms capture name, email, company, phone. Almost none capture the actual referrer or first-touch source. Add hidden fields that pull from UTM params, document.referrer, and a first-touch cookie set on first visit. Push these into the CRM as a Lead Source and Original Source field, two separate fields, because last-touch and first-touch usually point to different things.

This is one of the most common gaps we find when we segment a messy CRM into a revenue-ready database.

3. CRM stages that map to dollar values.

If your stages are "New," "Working," "Qualified," "Won," your attribution will be useless. Stages need to mean something to a salesperson, usually tied to deal size, decision-maker engagement, or a specific buying milestone. Every stage transition should fire an event into your reporting layer with the deal value attached.

4. A source field on closed deals, filled by sales.

This is the one nobody wants to do. Sales reps hate filling out "How did this customer find us?" in the CRM. Make it required at the Closed-Won stage. Not optional. Either a single dropdown of your actual channels, or a free-text "primary reason this deal closed" field. Review it weekly.

Without this, you cannot compute revenue by source. Period.

The model: first-touch, last-touch, and a weighted middle

Once you have the data, you have to decide how to give credit. There are three models worth knowing, and one you actually want to run.

First-touch attribution gives 100% of the credit to the source that first brought the contact into your world. It's useful for understanding awareness, specifically which channels actually introduce you to new people.

Last-touch attribution gives 100% of the credit to whatever source the deal was associated with when it closed. It's what your ad platforms report by default, which is why it tends to over-credit Google branded search and direct traffic.

Linear attribution splits credit evenly across every touch. It's honest, but it over-credits low-intent touches like display impressions.

The one we recommend for B2B lead gen: a weighted multi-touch model, but a simple one. Something like 40% first-touch, 40% last-touch, and 20% spread across the middle touches. Run it monthly. Don't get fancy with Shapley values until you have two years of clean data and a data team.

If you only have time for one model, use last-touch mapped to closed revenue, not closed leads. That's the bare-minimum upgrade most teams need. Pair it with lead scoring so the lead volume metric stops being the thing everyone optimizes for.

What good attribution actually changes

Once attribution is wired up, the decisions shift fast.

You stop asking "how many leads did LinkedIn get us" and start asking "what's the cost per closed deal from LinkedIn versus Google versus the trade show." You might discover that your cheapest leads close at the lowest rate and your most expensive leads close at the highest rate, meaning your current budget is allocating against the wrong metric. We see this constantly in Meta lead ads audits.

You also start catching real problems earlier. If a channel used to produce $4 in revenue per $1 spent and is now producing $1.50, you'll see it in the monthly attribution view before you see it in pipeline reports. By the time MQLs drop, you've already lost three months.

And it changes how you budget. Most B2B teams underinvest in channels that close deals slowly but consistently (events, partnerships, organic) and overinvest in channels that produce quick but low-quality leads. Real attribution makes that visible.

The honest limits

Attribution will not give you certainty. Even a clean multi-touch model is an estimate. Offline channels (referrals, trade shows, word of mouth) will always be partially invisible. Long B2B sales cycles mean a campaign you run this quarter may not produce closed revenue for six to twelve months, so you need a rolling view, not a monthly snapshot.

What attribution gives you is a much smaller margin of error than you're operating with today. If you can confidently say "this channel produced between $1.50 and $2.50 per dollar spent this quarter," you can make better budget calls than the team running on "we got 200 leads this month."

That's the bar. Build to that bar before you buy any attribution software.

If your CRM has revenue sitting in it that you can't tie back to the campaigns that produced it, we'll audit your funnel for free and show you exactly which pieces are missing.