First-click vs last-click is a false binary

I've sat through too many meetings where marketing teams argue about attribution models like they're choosing between two correct answers.

"Last-click credits the conversion. That's what matters." "First-click credits the discovery. That's what started it."

Both camps are wrong. Both models are broken. You have to pick the one that does the least damage.

Why last-click is particularly bad in B2B

Last-click attribution in a 9-month sales cycle almost always credits brand search or direct traffic.

Here's the typical journey: A procurement engineer sees a LinkedIn post. They click through, browse specs, leave. Three weeks later they Google your brand name because they remembered you. They submit a quote request. Last-click attribution credits "Brady marking systems" brand keyword.

Your brand campaign gets full credit. Your demand generation budget gets nothing.

Over time, this creates a predictable doom loop:

1. Demand gen campaigns get no credit because they're rarely the last touch
2. Demand gen budgets get cut because they "don't perform"
3. Brand search metrics stay strong because brand intent is already there
4. New pipeline dries up 6-9 months later
5. Everyone blames the market

You didn't have a market problem. You had an attribution problem that caused a budget allocation problem.

Why first-click overcorrects

First-click solves the demand gen problem but creates a new one: it over-credits generic discovery terms and upper-funnel content.

"Industrial safety equipment" gets a first-click credit for a deal that was really closed by a technical spec landing page visit three months later. Now you're over-investing in generic awareness keywords and underinvesting in the specific terms where engineers make final decisions.

The practical fix

Neither model is right. But you have to work within the constraints you have.

If forced to choose: weight toward first-click for top-of-funnel budget decisions and last-click for optimization decisions on high-intent terms.

The better option: run a hybrid view in BigQuery. Pull all touchpoints for closed-won opportunities in the past 12 months. Count how many unique campaigns appeared in each deal's touch history. Allocate budget weight proportionally to touch frequency, not position.

It's not perfect multi-touch attribution. But it's directionally more honest than either first-click or last-click alone.

And "directionally more honest" is what you actually need when making $50K/month budget decisions.