Why last-touch attribution undervalues your best growth channels (and how Snapchat thinks about it)

If you’ve been in UA long enough, you’ve had some version of this conversation: a channel looks expensive in your dashboard, you cut it, and installs drop somewhere else. Last-touch attribution handed you bad intelligence. You acted on it.

It’s not a rare edge case. It’s one of the most common ways mobile growth gets mismeasured.

Most teams optimize for what shows up in their reports, not for what actually drives conversions. And the gap between those two things is getting harder to ignore as app growth becomes more complex: more competition across UA channels, more pressure on ROAS, more signal loss from privacy changes. Simply scaling spend on the same channels no longer guarantees incremental growth.

Snapchat’s Liam Clack, Product Marketing Manager for small and medium customers at Snap, addressed this directly in Session 8 of Singular’s Zero to Hero Workshop: Beyond Last-Touch: Scaling Incremental App Growth. Here’s what he covered, and what it means for how UA teams should be thinking.

Zero to Hero is a free, on-demand workshop series featuring expert sessions, 10 minutes each. It’s built to help mobile growth teams scale from first install to global impact. Build. Launch. Scale. Win.

Register for free access → lp.singular.net/zero-to-hero-workshop 

The problem with last-touch

Last-touch attribution gives 100% of the conversion credit to the final interaction before a user installs. It’s simple. It’s fast to act on. And it’s been the default model for most UA teams since the category began.

The problem is that app users don’t discover, consider, and convert in a single click. They encounter your app across multiple channels, across multiple sessions, often over several days. The channel that creates initial awareness rarely gets the credit. The channel that closes the install does.

As Liam put it:

“We end up optimizing for what gets credit, not what truly drives growth.”

This creates a structural bias. Channels that capture demand at the bottom of the funnel look better than they are. Channels that create demand at the top look worse. And over time, budget flows toward closers and away from builders, until there’s less demand to capture.

Demand capture vs. demand creation: the distinction most dashboards blur

Not every channel plays the same role in a user’s journey. Understanding the difference is the starting point for better measurement.

Demand capture channels reach users who are already searching and ready to act. Search ads, branded retargeting, app store placements. They’re efficient at closing, and last-touch attribution rewards them heavily because they’re usually the final click.

Demand creation channels do something harder to measure: they introduce users to your app before those users knew they wanted it. They build awareness, seed intent, and move cold audiences into the consideration phase. Without them, the pool of users ready to be captured by demand capture channels gradually shrinks.

This is the part of the journey where Snapchat is most deliberate. It’s built to reach users before they’re searching, influence intent before it crystallizes, and convert that early attention into installs. Most platforms optimize for the bottom of the funnel. Snapchat plays the full journey, and that’s a strategic choice worth understanding.

What the data actually shows

Liam didn’t just make the case conceptually. He brought the numbers. Singular’s multi-touch attribution analysis of Snapchat gives the data its foundation, and three figures from the session make the point clearly:

75% of 13-34-year-olds across 25+ countries is the size of Snapchat’s daily audience. And importantly, a substantial share of that audience isn’t being reached elsewhere. In the US, more than half of daily Snapchatters are not on YouTube. Large portions aren’t on TikTok, Facebook, or Instagram.

Liam’s point: incremental reach, reaching users your other channels can’t, is itself a driver of incremental growth. If 84% of Snapchat-driven installs come from users not reached on any other platform, then a large share of its impact is inherently additive. That impact doesn’t show up clearly in last-touch numbers because last-touch doesn’t measure whose journey Snapchat started, only whose conversion it closed.

51% assist rate. Snapchat influences 51% more installs than it receives last-touch credit for. Those are real conversions, driven in part by Snapchat exposure, that never get attributed back to the platform in standard reporting.

~20% ROAS improvement when measured under multi-touch attribution versus last-touch alone. That’s not a data artifact. That’s a meaningful chunk of real performance that last-click reporting quietly drops.

For context, the Singular ROI Index 2026, which introduced MTA leaderboards for the first time alongside traditional last-touch rankings, found Snapchat appearing across 25 leaderboards. It does not underperform. It’s being underreported.

How to scale your app on Snapchat, straight from Snapchat

The framework (hack!) for scaling on Snapchat comes down to three levers working together:

1. Signals: better outcomes start with better data going in. That means passing high-quality signals from your MMP and CAPI so Snapchat’s systems can optimize toward the post-install events that matter, not just the install itself. Snapchat supports 90-day post-install action pairing and privacy-safe measurement with or without SKAN supplemental reporting. If you’re on Singular, you can connect your MMP data to Snapchat directly through Singular’s cross-platform attribution, keeping your measurement unified and your signal clean.

2. Snapchat’s automation tools: once signals are in place, let the system do more. Snapchat’s Smart Targeting uses machine learning to go beyond predefined audience segments and surface users more likely to convert, at scale, without constant manual intervention. Snapchat’s Smart Budget shifts spend dynamically toward the best-performing ad sets. This is how incremental reach actually scales: not by manually expanding who you target, but by letting Smart Targeting find users you’d never have explicitly targeted in the first place, and Smart Budget follow the performance.

3. Formats: two stood out in the session. Sponsored Snaps appear directly in the chat feed, one of the highest-attention environments on the platform, and they’re built for lower-funnel conversion events like installs and in-app purchases. Advertisers have seen around an 18% increase in unique app ads converters. App End Cards add App Store screenshots and reviews to Snap ads, creating a one-tap path to install at the exact moment someone is ready to commit, with around a 19% boost in SKAN installs for advertisers using them. Together, these are part of Snapchat’s App Power Pack, which has produced around a 25% lift in app installs since launch.

5 practical steps to scale incremental app growth

Liam closed with five:

1. Pass high-quality signals from your MMP and CAPI. The system optimizes toward what you feed it.
2. Lean into automation. Smart Targeting and Smart Budget perform better when you’re not over-constraining them.
3. Use Target Cost to balance volume and efficiency, rather than chasing one at the expense of the other.
4. Refresh creative frequently. Discovery formats need new angles. Stale creative kills reach.
5. Measure incremental impact, not just last-click. Set up lift-based tests, allow time for learning, and compare across channels, not just within one platform. Singular’s incrementality testing guide covers the setup.

The discipline that ties all five together: clear success metrics before the test, time to generate signal before a decision, and where possible, lift-based approaches that show true causal impact rather than attributed credit.

“Winning teams measure differently. They prioritize incremental impact across the full user journey.”

— Liam Clack, Snapchat

A thought on what’s coming next in measurement

The mobile measurement conversation isn’t stopping at Snapchat and last-touch attribution. It’s already moving into territory that would have sounded speculative two years ago.

Singular’s own data, published in the Q1 2025 Quarterly Trends Report, showed an 8,400% increase in customers receiving traffic from ChatGPT in a single quarter. Now ChatGPT is running ads, and it comes with a measurement problem that makes the last-touch attribution problem look simple: conversational ad environments don’t preserve referrer data cleanly, campaign-level attribution is limited, and the usual toolkit doesn’t work.

Singular is among the first MMPs to support ChatGPT Ads, with automated cost reporting and a Conversions API in Extract by Singular. Read what that actually means for measurement.

The teams already building clean measurement infrastructure today, passing good signals, testing incrementally, thinking across the full journey, are the ones who’ll be positioned to move quickly when new channels like ChatGPT Ads mature enough to scale. Measurement infrastructure isn’t just a reporting function. It’s a competitive advantage.

The bottom line

Not many UA teams get a direct playbook from the platform itself on how to measure and scale their app campaigns correctly. That’s what Liam laid out in 10 minutes: how last-touch attribution misses the channels that create demand, why incremental reach is a performance lever in its own right, and how Snapchat’s own performance framework, signals, Smart Targeting, Smart Budget, formats, is designed to drive growth that holds up beyond the dashboard.

The session Liam delivered in Zero to Hero isn’t just about Snapchat. It’s a sharper way of thinking about how you measure any channel that operates earlier in the user journey. And it’s free to watch.

Watch Session 8 and all 9 ‘Zero to Hero’ sessions here → lp.singular.net/zero-to-hero-workshop

Already measuring? See how Singular handles attribution, incrementality, and new channels like ChatGPT Ads, all in one place.

Frequently asked questions

What is last-touch attribution in mobile marketing? 

Last-touch attribution gives 100% of conversion credit to the final ad interaction before a user installs or completes an in-app action. It’s the default model in most UA dashboards because it’s simple and fast to act on. Its core limitation is that it can’t capture the influence of earlier touchpoints that created the intent the final click converted.

Why does last-touch attribution undervalue discovery channels? 

Last-touch attribution gives 100% credit to the final ad interaction before a user converts. Discovery channels, including social platforms that operate earlier in the funnel, rarely get that final click. They introduce users, build awareness, and drive consideration, but the conversion often happens elsewhere. So last-touch overvalues the closer and undercounts the channel that created the intent in the first place. Singular’s MTA analysis shows Snapchat assists 51% more installs than it receives direct last-touch credit for, a gap that reflects how much early-funnel influence gets lost in standard reporting.

What is incremental reach in app advertising? 

Incremental reach means reaching users who aren’t already being reached by your other active channels. It’s a direct driver of incremental growth because it expands the pool of potential converters rather than targeting audiences your campaigns are already touching. Snapchat’s audience in many markets is largely distinct from other major platforms, meaning a large share of its install impact is additive.

How do you measure incrementality for app campaigns?

The most reliable method is a lift-based test: split your target audience into an exposed group and a holdout group, run your campaign, and compare install rates between the two. The difference is your incremental lift. For cross-channel incrementality, Singular’s Incrementality & A/B Testing gives you a unified view across all your channels rather than relying on any single platform’s self-reported lift study.