Tuesday, May 6, 2014

Optimizing Upsells: How We Drove Millions of App Installs

It’s pretty tough to get people to download your app, and relying on App Store SEO can be unpredictable. However, one of the few levers you do have control over is the web traffic you already own. It’s now common practice to direct mobile web visitors into the App Store via some kind of promotion.

But ... what works the best? How much does the UI matter? How about the copy?

Optimizations like this are great opportunities to use A/B testing and data-driven decisions. Over the past year or so, we’ve run several experiments to see what works the best for us. Read on and see if you can guess the outcomes correctly!

Experiment 1: Straightforward vs. Promo

We had originally launched the Polyvore iPhone app with a straightforward “Get the app!” message and a screenshot of the editor. But, since then, the app has done consistently well in its downloads and ratings, and we wondered whether conveying that would help. We also knew that not all of our users are interested in the editor, so perhaps focusing on a different part of the app experience would be better.

L: Straightforward + editor message / R: Promo + shopping message

Winner? Tie.

Learning: Perhaps our web downloaders are already familiar with the site and don’t need that kind of information, or a 5-star rating is not that impressive to end users. But, since it didn’t make a difference in install rates, we kept the new one.

Experiment 2: UI format

Next, we tried different types of UI. Apple recommends a more subtle smart banner, while we’ve seen stronger approaches with whole screen takeovers.

L: Popover / M: Smart banner / R: Takeover

Winner? The takeover screen, by a mile.
Smart banners actually caused a 15% drop in installs, while takeovers had 4X the CTR of popovers.

Learning: Easily dismissible or ignorable UIs will be dismissed or ignored before visitors have a chance to see the content. Of course, we don’t want to annoy our visitors, so we try to be smart about only showing them on the first visit.

Experiment 3: Custom vs. generic

We picked a few of our main pages to experiment with and see if customizing for a particular referer or piece of content would help.

On our shop and product details pages, the upsell paired shop messages with an assembled custom “screenshot” of that same landing page in the app, using accurate images and text. On our set detail page, we looked specifically at Pinterest referrals and focused on how users could pin to their boards straight from Polyvore.

L: Shop page / R: Thing page

Winner? Mixed.
Custom content worked great on the shop and thing pages (+5-35% install rates) compared to the generic “download our app message”, but actually tanked on the set page (-46%).

L: Pinterest take 1 / R: Pinterest take 2

We then tried adjusting the Pinterest-specific message to be more about the set itself, which improved numbers quite a bit, but still not up to the control at -16% installs. In the end, we reverted to the generic upsell for Pinterest referrals on set pages.

Learning: Customizing the content of an upsell to what people came for usually works well, but be careful to give just enough info to get them wanting more. For visitors who are less focused (ex: social referrals in the browser of a different app) a contentful upsell is enough to fulfill their curiosity without needing to download the app or proceed into the website.


These three experiments are just the beginning of many more things you could try. For example, we have yet to explore delaying the upsell or seeing if other social referrals behave similarly to Pinterest. Is there anything interesting that you’ve discovered while working with app upsells?