Multi-Touch Attribution & Marketing Mix Modeling (MMM)Playbook3 min readUpdated September 2026

A Worked Example of Finding Wasted Ad Spend

"Cut the weakest-performing channel" is a common instinct when ad spend efficiency needs to improve, but the channel that looks weakest in a simple last-click report is sometimes doing real work the report just isn't capturing. Here's a worked example of how a closer look with multi-touch data changes the picture.

The setup: three channels, one report says cut one of them

Say a company runs three paid channels: search, social retargeting, and a display prospecting campaign. A last-click report shows search converting well, retargeting converting well, and display prospecting converting poorly, with almost no direct conversions attributed to it. The obvious read is to cut display prospecting and shift that budget to search.

This is a genuinely reasonable conclusion from the report as presented, which is exactly what makes the mistake so easy to make. Nothing about a standard last-click dashboard signals that the report itself might be missing something important.

What the touch history reveals once you look closer

Tracing the full touch history of customers who converted through search or retargeting reveals that a meaningful share of them had an earlier touch from the display prospecting campaign, sometimes weeks before their eventual conversion. Display was never meant to close deals directly, it was introducing prospects who then got captured and closed by search or retargeting later. A last-click report has no way to show this, because it only credits the final touch.

This pattern is easy to miss unless you specifically go looking for it, since the default report every platform shows you is built around last-click credit by default, and nothing about that default view hints that an earlier, uncredited touch might be doing real work.

Recalculating what display prospecting is actually worth

Once those assisted conversions are counted, even at a partial credit rather than full credit, display prospecting's real contribution looks very different from the near-zero the last-click report showed. Compare your blended CPC against the $5.26 all-industry average on Google Ads1 before deciding a channel is inefficient purely on its own cost, since a channel feeding conversions into other channels can be worth running even at a higher CPC than that benchmark, if its downstream contribution is real.

What actually would have been wasted spend here

The waste in this scenario isn't the display prospecting budget, it's the decision that almost got made to cut it based on an incomplete view. Real waste tends to show up differently: a campaign with no assisted conversions either, one that neither closes deals nor introduces prospects who convert elsewhere, is a much stronger candidate for cutting than a channel that looks weak only because of how it's being measured.

That second kind of campaign, the one contributing nowhere in the journey, is worth actively looking for once you've cleared the channels that turned out to have a hidden introducing role. It's usually a smaller and more specific list than the original weak-looking group started out as.

How to check your own channels for this pattern

  • Pull assisted-conversion data for any channel that looks weak under last-click attribution before deciding to cut it
  • Check whether that channel's touches tend to appear early in converting customers' journeys, which would suggest an introducing role rather than a closing one
  • Compare the channel's cost against what it would cost to replace its introducing function elsewhere if you did cut it
  • Only cut a channel once you've confirmed it isn't quietly feeding conversions that get credited elsewhere

A short decision rule helps here. Before cutting any channel, ask two questions: does it close deals directly, and does it appear early in the journeys of customers who close through another channel? If the answer to both is no, cut it and move the budget to a channel with a proven result. If the answer to the second is yes, keep it and judge it on assisted conversions rather than last-click numbers. Revisit the decision after a full sales cycle, since introducing channels often show their value only weeks after the first touch, and an early cut can hide that.

The broader lesson for spend efficiency reviews

Efficiency reviews built entirely on last-click data will systematically undervalue channels that play an introducing role in the funnel, and systematically overvalue channels that happen to sit at the end of a journey regardless of how much work they actually did to get there. Building multi-touch or assisted-conversion data into every spend efficiency review, not just when a channel already looks suspicious, catches this before a genuinely useful channel gets cut by mistake.

Executive Capability Standard

What Good Looks Like

The standard is checking assisted-conversion and touch-history data before cutting any channel that looks weak under last-click attribution alone.

Building The Capability (5-Stage Skill Ladder)

1. Learn:Pull assisted-conversion data for your current channels to see whether any weak-looking channel is quietly feeding conversions credited elsewhere.
2. Do Manually:Manually trace the touch history of your best-converting customers to check how often other channels appear earlier in their journey.
3. Delegate:Have a growth or performance marketer own the quarterly spend efficiency review, including assisted-conversion analysis, before any channel gets cut.
4. Automate:Build assisted-conversion reporting into your standard dashboard so it's visible by default, not something someone has to remember to pull separately.
5. Buy:Bring in analytics help to set up proper multi-touch tracking if your current setup can't reliably trace touch history across channels yet.

How to Get Started

Frequently Asked Questions

How do we know if a weak-looking channel is actually introducing customers who convert elsewhere?

Trace the full touch history of customers who converted through your strongest channels and check how often a touch from the weak-looking channel appears earlier in that same journey. A meaningful overlap suggests an introducing role the last-click view is missing.

Should we give a channel full credit for an assisted conversion?

Most teams use partial credit rather than full credit for assisted touches, since the closing channel still did real work too. The exact split matters less than simply counting the assist at all instead of giving it zero credit the way last-click attribution does.

What does genuinely wasted ad spend actually look like?

A channel with no direct conversions and no meaningful presence in the touch history of customers who converted elsewhere. That combination, contributing nothing to a close and nothing to an earlier touch that led to one, is a much stronger case for cutting than weak last-click numbers alone.

Sources

Where we quote a benchmark, we show its source. Other figures in this guide are estimates or general guidance, so check them against your own numbers.

  1. Google Ads Average CPC (All Industries). WordStream 2025 Google Ads Benchmarks, 2024-2025.

Related Guides