How To Use 3 Pillars of Mobile Growth: Attribution, Measurement, & Testing

Mobile growth is often more an art than a science, but there are three pillars that every marketer should embrace to grow their businesses. Take it from an expert. Anthony Scarpaci, VP of Growth at Acorns, stopped by our podcast “How I Grew This” to give us the rundown on those pillars: attribution, measurement, and testing.


Understanding the health and purpose of each marketing channel in your portfolio helps your team know when and where to invest.

Different channels produce varying levels of ROI and diminishing returns. Having your thumb on the pulse of what is and isn’t producing helps the marketing team know where to spend more time and effort.

Attributing conversions to a specific channel, especially with a unified model, is sometimes complicated. However, diving deep to try to find which channels provide the most traction helps a brand attract and convert new customers more effectively than a wide-net approach.

The difference between attribution and incrementality

When measuring marketing performance, it’s important to distinguish between attribution and incrementality since they measure different aspects of your marketing’s effectiveness.

Attribution is: ‘I can associate this marketing impression to this customer conversion.’ Incrementality is: ‘did that impression drive the behavior of that customer?’” — Anthony Scarpaci, VP of Growth at Acorns

Attribution describes the process in which companies match the correlation between two data points — such as clicks on an ad to online purchases. It’s important to pay close attention to events that cause spikes in attribution — such as TV commercials to website visits — because these events can alter the ROI associated with these channels.
Incrementality measures the effectiveness of each advertising method and its ability to drive conversions. Incrementality measurements consider the direct effect of each touchpoint’s ability to drive conversions but does not include sales that would’ve happened incidentally.

Multi-touch attribution

Building a multi-touch attribution (MTA) solution to track all clicks and views through platforms takes time and lots of experimentation to perfect.

In reality, achieving a unified MTA solution is outdated and consumes more resources than are justified.

“You really have to stitch together more than one methodology to get to what the true performance of a given marketing campaign is.”  — Anthony Scarpaci, VP of Growth at Acorns

Considering the many costs and variables involved in building unified MTA solutions, they aren’t always the most practical framework to implement for every brand.

“I tell my team now [that] multi-touch attribution is dead,” says Scarpaci. “It’s dead, and it’s probably for the best from a consumer first standpoint… What I tell them now is to think about any bottoms-up solutions that we build as fractional attribution.”

That being said, there are still tools in your arsenal, such as ad hoc experiments and third-party tools which can help you establish frameworks for giving partial credit to different aspects of your marketing mix for conversions.


Fractional attribution is at the heart of measuring each channel’s success.

Deciding the weight, given to how a company measures success, depends on many factors: the size of a business, the stage of growth, and the flexibility in a company’s marketing and advertising campaigns.

Successfully measuring each channel’s ROI means leveraging statistics, experiments, and common sense.

Business significance vs. statistical significance

Statistics can be a bedrock for larger companies because of the scale at which each channel has already been developed, their budgets, and other important factors. The sheer volume of attributions and the rate at which a larger company grows makes relying solely on statistical modeling a viable option.

However, for smaller-scale businesses and startups focused on hyper-growth, statistical modeling can be problematic.

According to Scarpaci, it can hinder “your ability to move fast on things when you need to move fast [and], when you need to constantly find the next best thing to go improve conversion rate, drive scale and find winners.”

“If you allow three months for an experiment to understand if a 2% lift is statistically significant, you’re not spending your time on the right things.” — Anthony Scarpaci, VP of Growth at Acorns

Businesses need to balance experiments and velocity and know how to get the most out of their frameworks while continuing to drive conversions.

“We are balancing the level of analytical rigor with … some common sense,” says Scarpaci. This means, for smaller businesses, you should pause an experiment that hasn’t reached statistical significance if you know its hypothesis is not a home run.

You have to strike a middle ground there so that you can balance velocity and confidence in what you’re experimenting.”


Making big bets and taking calculated risks is essential to driving growth at scale.

Creating experiments (especially of the small and audacious variety) is key to finding a winning formula in the long term.

“Trust in your principles and your frameworks of what you know works within marketing as you make big bets.” — Anthony Scarpaci, VP of Growth at Acorns


Without experimenting with new methods of outreach on a regular basis, it’s difficult to understand the next steps as your company continues to develop. There are a variety of experiment types to select from, including:

As Scarpaci experienced, unleashing a creative team’s new ideas on an ad experiment not only excites the team, but it can also beat other ad results — in Scarpaci’s case, by 20%. They achieved this result by combining best practices like a strong call to action, urgency, and clear value proposition with their exciting new creative.

Remember: The goal is to make sure you’re leveraging both creative and analytical teams to drive results. This is more important now than ever before as the life of a creative ad, especially in digital channels, gets shorter and shorter.

The importance of testing

Testing is critical to growth because audiences can be fickle and tastes can change quickly across channels. For example, while landing pages and video ads have traditionally been the gold standard for conversions, they don’t always perform in the way you’d expect.

As social media and other online content platforms become more popular, it’s important for marketers to test different methods across different channels to capture the divided attention of today’s consumers.

Attribution the right way

It’s time to change how we frame attribution. Scarpaci’s advice? Recognize that each marketing channel interacts with each other and assists with last-touch conversion. To learn more, listen to our full conversation with Scarpaci on our podcast “How I Grew This.”