One of the challenges facing mobile advertising companies is lowering the price of attracting app downloads, a cost that is constantly increasing for the companies developing the applications.
It’s the holy grail of internet ads – getting users who come across an ad for an app to click on the ad and download the app, known in the industry as the conversion rate.
Application developers generally don’t have access to information about users’ conduct in app stores, making it difficult for them to improve their conversion rates. Until now, that is. An Israeli company called StoreMaven decided to look at user conduct in app stores such as Google Play and the Apple App Store and come up with a way of measuring the effectiveness of particular marketing messages.
StoreMaven was established last year by Gad Maor, Adam Rakib and Eldad Abel, three of the eight founders of Shaker, a social networking software firm that was sold last week to Playstudios for a price estimated at tens of millions of dollars.
Their Shaker experience taught the three founders of StoreMaven that one way of overcoming the lack of transparency in app stores was to create several different store pages for identical apps and measure user responses to the different pages.
In 2013, they began developing mobile-based games involving a large number of players, based on their experience at Shaker. “That was our gamble,” Gad Maor’s brother, Yonatan, told TheMarker.
Gad left the venture in 2014 to develop StoreMaven. Shaker later developed a cooperative relationship with Playstudios and together they developed a game called POP Slots, a virtual casino for a large number of players that integrated the concept developed early on by Shaker.
“Shaker was a data company and the entire time we were developing the tools and capacity to find logic in huge quantities of information,” Yonatan Maor says.
“In 2013, Shaker took a sharp change in direction and became a company developing only for mobile. We found ourselves confronting a challenge: how to get people to find and install the application at a cost that would enable us to grow and continue to market the product. When we started to look at where people make the decision over whether or not to download the app, we understood that it happens in the app store.”
Most users find themselves in an app store by way of advertising, recommendations or because they are looking for an app in a certain category. Ultimately the user comes face-to-face with the app’s page in the app store, which is essentially advertising and includes pictures, a video and text.
“For companies, whether they’re gaming companies or content companies, this page is the product’s home page, but developers don’t have any idea what people do there,” Maor contends. “The developers don’t know how the users behave and what they waste their time on, and most importantly, what motivates them to download an application or to pass it up.”
To increase conversion rates, companies that are developing apps sometimes try several versions and for that purpose also develop several app pages in the app store to see which is better. But the broader the examination becomes, the more advertising and marketing costs grow with it, so companies want the process to take the least amount of time possible.
At Shaker, as with any app, the founders say it was impossible to obtain information about user behavior directly from Apple or Google. In addition, the two stores didn’t allow developers to test different messages to see which one worked best.
“To access the information, as a practical matter, we created our own version of an app store, essentially duplicating them, and we started directing some of our users there instead of to the real store,” Maor says.
“In addition to the fact that this was the first time a company that wasn’t Apple or Google was able to say something about how users behave in the store, we could also check various versions of app pages at the same time. We used different icons, pictures, video, and ultimately we learned which message was best for our application and improved the conversion rate.”
What began as an experiment for Shaker became a separate company that now has 15 employees in Tel Aviv and $400,000 from the 500 Startups venture capital fund and from Shaker itself.
StoreMaven permits companies to direct some of their users to a duplicate page where the testing is carried out. The company’s tool allows app developers to see the volume of traffic to the page, how many people access it per day, who skips over the page and who clicks on a picture, while at the same time analyzing user behavior and drawing conclusions. The system provides information about the leading version of a client’s page. If users actually attempt to download the app, they are directed to the genuine app store.
The heart of StoreMaven’s product is an engine based on a set of algorithms designed simply to predict which page would be the winning version as quickly as possible, and in the process save the client money.
“Companies send a lot of users here with each user costing them money, but they are interested in paying the cost to achieve better [statistical] clarity,” Maor says.
For his part, Adam Rakib says: “The more the companies perceive a winning opportunity from high clarity, instead of through 5,000 clients with 3,000, for example, the more satisfied they will be, because it costs them less money.”
StoreMaven reports, by the way, that it is profitable.
In one example presented by the company, there is a version in which a minute change to the images on the page is shown to have low prospects for success. In such a case, the system will suggest stopping that version, which will also save the client money.
“It’s funny to see how company product people are sometimes certain that a particular example will work and the system shows that they are wrong,” Maor recounts. “We [at StoreMaven] have our own betting system among us on what will win,” he adds.
One of the questions that arises regarding StoreMaven’s operations is the legality of the product. Its founders claim that it is not only legal, but that it is being used by hundreds of clients, including leading companies such as Zynga, Yahoo, Playtika and Saga. They add that Google itself is a client, even though the company has a similar product that it launched in May of last year.
Asked about the prospect that potential clients would be lost to Google’s product, Maor says clients strike a balance between using StoreMaven and Google’s alternative. “Each one has its advantages. We place our emphasis not only on which application page is best, but on how users respond to each message that is tested and in the process enable our clients to improve more and more over time.”
Want to enjoy 'Zen' reading - with no ads and just the article? Subscribe todaySubscribe now