PhotoSì Finds Small Tweaks That Deliver Big Revenue Wins

A photo printing software company uses Amplitude Analytics to tweak the user experience for big revenue gains.

Customer Stories
January 31, 2024
Luca Photosi Headshot
Luca Longobardi
Data Scientist at PhotoSì
Photosi Feature Image

Insight/Action/Outcome: Looking at user data in Amplitude, the PhotoSì team saw many users buying multiple frames, but having to repeat the selection process for each frame they wanted to add. That led to a question: Could they create a feature allowing customers to add more frames in a single step? The answer was yes: Giving users the option to buy multiple frames in one step, the average order value for the category increased by €6 and the number of frames sold shot up 30%. That led to a 16% revenue increase (€330,000).

We live in a world that’s dominated by data. That means all teams behind an app or platform have the ability to learn everything about their customers and improve their experiences. We should all use this power to make something beautiful.

As a data scientist at PhotoSì, I’m responsible for everything concerning the experience of our products, whether on the web or our app. PhotoSì makes beautiful personalized objects from people’s photos: photo books and framed prints, but also calendars, puzzles, phone covers, and clothing. To improve our products and digital strategy, my goal is to glean as much knowledge as I can about our customers through our rich product data.

For any data scientist, our most important responsibility is to enable data-driven decisions. The challenge is to understand the margin for improvement in whatever you analyze. I have a process I follow. If we are studying a step in our current experience, I will:

  1. Learn more about what users do during that step and what our current performance is.
  2. Set up a plan, asking, “Do I know enough, or do I want to know more?” Needing to know more likely means you add more ways to track user behavior.
  3. I collect the data and start to analyze the behavior of this user or cohort.
  4. From this analysis I determine how much room for improvement there is in this step.
  5. The final step is the creative part of the job: working with others to identify the solutions that will achieve that margin of improvement. The development team and the UX team usually find the correct combination of solutions to improve upon the current experience. Then, we monitor the result and create an iterative process to ask whether we’re satisfied with the outcome or if we want more improvement. The process is then restarted from the beginning.

A huge part of this process for us at PhotoSì is using Amplitude to gather data and analyze behavior.

A small change can open up new markets

When I joined the company three years ago, PhotoSì had just chosen Amplitude as our analytics platform. Though I wasn’t part of the selection process, I understand the team’s reasoning. It started by seeing Amplitude was highly ranked in an analyst report, having the most capabilities of any product analytics tool on the market. Testing it against several competitors, the team fell in love with Amplitude.

Before Amplitude Analytics, PhotoSì did have some tracking, but it was very crude. As I’ve said, proper tracking is the first step to improving the customer experience. A lot of my initial work was putting this tracking in place—first in the app, later on the web—because we were essentially starting from nothing. That work paid off as we incrementally learned more and more about customers’ journeys, from basic conversion charts to taking full advantage of the deep analysis Amplitude offers.

Since adopting Amplitude three years ago, we’ve made countless improvements to our app and website and have had some significant wins.

One of my favorite improvements in the app was to the PhotoSì image picker. One of the main issues in our overall experience is the number of interactions required for the user to compose a project, especially for more complex products. It’s even more painful when a user tries to select photos from a shuffled bucket, as is the case with many phones’ classic photo pickers. The flipside to such a large pain point is a large margin for improvement.

Our solution was to group photos in an easier way for people to understand: The “picker by period” feature allows users to select photos by day, week, month, or year. The introduction of this feature alone led to a 4 percentage point increase in conversions. This was a significant win for this improved experience and drove meaningful growth.

We also improved the promo code experience in the app. Users are of course very motivated by discounts, but they encountered a frustrating number of errors when inputting promo codes into our app. Often these errors resulted from simple typos or not understanding the code’s terms, such as the minimum purchase amount. We know that users who have a good promo code experience are more likely to convert, so reducing these errors would be an easy win. We addressed this by implementing a feature so users are no longer responsible for typing in a promo code at all. Instead, we provide a list of promos automatically activated when the user becomes eligible, and the user simply taps to select a code at checkout. By implementing this feature and measuring the results in Amplitude Analytics, we reduced errors by half, which significantly improved conversion.

Conversion is a great metric because it ties directly to revenue. Every time we can increase revenue by improving the customer experience is a win, and sometimes even a small change gives revenue a big boost, as we saw in our move to allow users to buy multiple frames.

My colleague, Manuel Mainetti, whose team works on user acquisition and retention, conducted a market analysis showing that some customers envision decorating with a wall of frames. Looking at the user data, we found many users were buying multiple frames, but they had to go through the same process again and again for each frame they wanted to add.

Manuel’s team asked: Could we add a feature allowing customers to add more frames in a single step? Worst-case scenario, we would sell some extra frames, but in the best-case scenario, we would open up a whole new market for ourselves. And that’s what we did: Giving users the option to buy multiple frames in one step, the average order value for the category increased by €6 and the number of frames sold shot up 30%. That led to a 16% revenue increase (€330,000). This example demonstrates how collaborating across teams enables us to find new opportunities in the data.

Amplitude features help us find the impact

None of this would have been possible without Amplitude, which delivers the quantitative answers I need to inform the team. Imagine trying to achieve these successes without an understanding of how users move through our customer experience or a way to measure our improvements. It would be impossible. It’s often easy to see where users struggle in a process, but not always so easy to understand why.

Amplitude allows me to see the impact a particular event has on the product, and from that, the opportunity for improvement. From understanding to measuring to improvement scoring, we rely on Amplitude.

From understanding to measuring to improvement scoring, we rely on Amplitude.

Our two favorite features are funnel charts and a testing module with Amplitude. We start with conversions as a quantitative measure of user behavior at any given step in a process. We then ask if we’re satisfied with that rate. If we think there’s room for improvement, we look at other charts in Amplitude to identify possible solutions. Does this step require too many interactions of the user? Could it be simplified? The differences between two possible solutions often look ambiguous, but the testing module brings clarity to those differences.

The insights that settled our login questions

Our best aha moment was an insight that settled a long-running and hotly contested debate: Should app users log in at the beginning or in the middle of the experience? It felt like this debate had been going on since before I joined PhotoSì. This debate started with the news that iOS14 would kill the IDFA, meaning the only way to retarget users would be through email.

Putting the login at the beginning of the app process would allow us to retarget users who left a product unfinished, as well as optimize our campaigns generally. On the other hand, many people strongly argued that making users log in before they started building a product would be a barrier to engagement.

We finally laid the debate to rest by testing both solutions. The results surprised us all, because even those who argued for login at the beginning didn’t expect to see such a great improvement. When the login at the beginning of the pap process, 10% of users don’t engage at all, but this eclipsed by the positives: the overall conversion rate is greater for both purchases and installs, because users are more engaged, and PhotoSì captures their email address for retargeting.

Amplitude Analytics insights also helped us improve login on our web platform, which had much lower login conversions than our app. It’s a funny story, because it tells you we can all take the basics for granted. We couldn’t understand what was wrong with our user experience to account for this difference in logins. Finally, we looked at a detailed view of our individual users in Analytics and discovered the problem had been staring us in the face all along: We actually had a bug in our login process that stopped some users from proceeding. Fixing the bug accounted for a 7-point improvement in conversions (from 75% to 82%), and further improvements to the login experience pushed conversions up another six points (to 88%).

Fixing the bug accounted for a 7-point improvement in conversions (from 75% to 82%), and other improvements to the login experience pushed conversions up another six points (to 88%).

Amplitude changed how we work

Apart from increasing revenue, Amplitude is a big growth driver and time saver. If you’re not using a digital analytics platform, you’re not analyzing user behavior. That means you’re losing time by debating the best choice and missing opportunities to grow.

The funny thing about my app login example is that the debate stayed on the table for so long, when the answer turned out to be easy. Having quantitative data backed by testing eliminates that debate because it’s simple to identify what matters and to choose what’s best for our customers and the overall business.

Amplitude is invaluable, and it’s not all about the money. More than any dollar figure we could put to it, Amplitude has changed the way we work at PhotoSì, and the way we ask and answer questions. It’s like history started the day Amplitude arrived at the company, and the picture is much clearer now.

About the Author
Luca Photosi Headshot
Luca Longobardi
Data Scientist at PhotoSì
Luca Longobardi is a Data Scientist at PhotoSì. His mission is to enhance business understanding and to optimize marketing plans. He is experienced in Data Platform Building using various ETL tools and database technologies, both relational and non-relational.