{"id":10555,"date":"2013-09-23T08:30:02","date_gmt":"2013-09-23T15:30:02","guid":{"rendered":"http:\/\/www.dfcint.com\/dossier\/?post_type=companies&p=10555"},"modified":"2018-06-26T18:37:48","modified_gmt":"2018-06-26T18:37:48","slug":"priori-targets-app-data","status":"publish","type":"post","link":"https:\/\/www.dfcint.com\/dossier\/priori-targets-app-data\/","title":{"rendered":"Priori Data Interview: Targeting App Data"},"content":{"rendered":"
\"Priori

Priori Data CEO Patrick Kane.<\/p><\/div>\n

SEPT. 23, 2013 \u2022 Priori Data Interview: A big issue in the mobile and online space is trying to understand the size of the market and the potential for products in the marketplace.\u00a0 This information is very critical for developers and investors looking to make smart decisions.\u00a0 Unfortunately there has been a great amount\u00a0of hype and misinformation when it comes to accurate mobile data despite there being quite a few companies looking to provide analytics on mobile games.\u00a0 After examining most of them DFC Intelligence decided to partner with Priori Data GmbH out of Germany.\u00a0 Priori provides download estimates for individual apps and even more impressive they have developed a detailed taxonomy that can categorize apps into many categories and sub-categories.\u00a0 For example, games alone have 120 different sub-categories.<\/p>\n

This type of organized data is invaluable when looking to estimate market size and market potential.\u00a0 This year DFC Intelligence and Priori jointly published a detailed report The Global Market for Games and Entertainment Applications on Smartphones and Tablets that provided market sizing for 26 leading countries with five-year forecasts for market growth.\u00a0 However, this report is only the tip of the iceberg of the type of analysis that can be done with Priori\u2019s data collection.\u00a0 To get a better understating of the distinct results that Priori Data provides we talked to CEO Patrick Kane.<\/p>\n

DFC: As an independent research house Priori Data is a relative newcomer launched in April of 2013. But your roots go back further as the research side of app discovery and analytics firm Xyologic, founded in 2010. Tell us more about Priori’s early beginnings at Xyologic, and how did you grow into a separate organization. <\/span><\/em><\/p>\n

Patrick: Priori does indeed draw its roots from the analytics arm of Xyologic. For a long time, Xyologic was a hybrid app discovery and analytics company. Obviously to build a discovery solution you need to have a comprehensive database for users to search, so that was their first step. Next they added technologies to help them deliver the \u201cbest\u201d search results for their users, a component of which included developing download estimates for every app.<\/p>\n

The download estimates proved quite valuable, because suddenly they were able to help people \u201cmeasure\u201d the popularity of apps. Download estimates aren\u2019t published by the app stores \u2013 Google Play provides download ranges for a given app, i.e. they display that an app has received between 500,000 and 1,000,000 downloads \u2013 or the developers, and at that time there was remarkably little information on how to quantify the economic opportunity generated by apps. Xyologic\u2019s search technology had inadvertently addressed an industry analyst pain point, and they began sharing a series of free monthly reports with the community. On the basis of these reports, they also generated enough buzz to bootstrap their initial development phase by selling custom datasets to a series of mobile & technology industry clients.<\/p>\n

\"PrioriXyologic raised funding in June 2012 and with the additional investment their focus shifted exclusively to the app discovery side. It was at that time that Patrick Kane, Priori\u2019s CEO, started discussing with the Xyologic co-founders the opportunity to further extend the initial work they had done on the analytics side. After confirming the market opportunity and determining that a full-fledged analytics and research service would require a new infrastructure, investment, and brand, the idea for Priori was born.<\/p>\n

DFC: What is the strategic goal for your firm? Where do you fit in amongst competitors such as App Annie, Distimo, and Flurry? What makes Priori stand out? <\/span><\/em><\/p>\n

Patrick: Our vision is to become the data backbone of the mobile app industry. We want to provide our clients with a definitive source of app metadata, delivered via cutting-edge big data analysis tools and differentiated research products, allowing them to better shape their mobile and digital strategies.<\/p>\n

AppAnnie and Distimo are competitive \u2018app store analytics\u2019 companies, while we consider Flurry\u2019s \u2018app usage analytics\u2019 a very important but adjacent data category. With respect to AppAnnie and Distimo, we see two key points of differentiation with our offering:<\/p>\n

First, we provide a complete top-to-bottom view of the app stores. Most of our competitors use methodologies based heavily on the \u201ctop charts\u201d as displayed by the app stores. If an app is on one of these charts, they have a pretty good sense for its overall scale and ranking. But if the app hasn\u2019t appeared on a top list – say it is an up-and-coming app or appeared briefly and then dropped out – they struggle to provide quantitative information around it. We measure the richness and activity which appears just beneath the layer of the top lists, and we pass that along to our clients. Our solution also allows us to provide a complete \u201cmarket view\u201d of a platform or category, because we are counting information from all apps.<\/p>\n

\"\"Second, we have a novel app taxonomy, which breaks down the apps in our database into more than 750 sub-categories. As former analysts ourselves, we know the value of being able to compare an app\u2019s performance on an apples-to-apples basis. So while a competitor might be able to paint a picture of what goes on in the \u201cCard Games\u201d category on iOS, we can tell whether performance was driven by Blackjack, Solitaire, or Slots. These insights can be really powerful, particularly for developers and investors.<\/p>\n

DFC: You have an exclusive license to use Xyologic technology. What is the relationship with Xyologic moving forward? <\/span><\/em><\/p>\n

Patrick: We remain very closely aligned, and believe there is much to gain for both companies going forward. Priori will obtain insights from Xyologic\u2019s ongoing B2C interactions, some of which we will be able to convert to metrics for inclusion in our data platform. Similarly, Priori\u2019s enhanced analytics capabilities should be beneficial for the Xyologic search algorithms. We see great things as a result of our unique partnership.<\/p>\n

DFC: Xyologic had received external funding. Do those investors also carry on as investors in Priori? <\/span><\/em><\/p>\n

Patrick: Yes, we\u2019re very lucky to have the support of Xyologic\u2019s original investors as shareholders in Priori.<\/p>\n

DFC: How are these sub-categories designed? For example, we can easily see a branch that goes: Sports, Racing, Rally\/F1\/Street\/Truck\/Fantasy. But do you go deeper into sub-categories like licensed versus generic vehicles, or sponsored versus retail titles? <\/span><\/em><\/p>\n

Patrick: At the moment we don\u2019t go further than \u201cracing\u201d per your example. Obviously the challenge is drawing the line at which level of segmentation is valuable, and there is a constant trade-off between detail and complexity. Ultimately, we don\u2019t necessarily need to be the mediator for drawing that line, and we plan to introduce functionality that allows our users to create their own groupings and tags which match their unique interpretations of the app stores.<\/p>\n

DFC: What is your methodology in developing sub-categories? Are you working with developers and publishers, or designing them entirely on your own? <\/span><\/em><\/p>\n

Patrick: The sub-categorization methodology is based on algorithms that mine the descriptive app data and generate clusters of similar apps. The clustering process identifies apps with a strong affinity to one another, and then determines which of those clusters are valuable and which ones aren\u2019t.<\/p>\n

\"Top_10_Gaming_SubCategories_iOS_July2013-S\"Our sub-categorization certainly isn\u2019t perfect, and one thing we\u2019ve learned is that everyone has their own idea for how to categorize apps, so we plan to add features that give publishers and other stakeholders the ability to contribute to the dialogue \u2013 and we hope the final result will be even better.<\/p>\n

DFC: As of July you have identified Water & Fire games as the No. 3 top trending game sub-category on iOS. We can see how an app featuring a fireman putting out fire outbreaks would fit in there. In what sub-category would you put a simple game app in which aircraft napalm guerrilla insurgents? <\/span><\/em><\/p>\n

Patrick: This would be tough. Our system would probably categorize such a game as a \u2018flight game\u2019 or \u2018shooting game\u2019, depending on what the publisher itself emphasized when describing their app.<\/p>\n

DFC: When you say you monitor 50 categories, 550 app sub-categories, and 120 game sub-categories, what percentage of that data is from public sources or developer partners? <\/span><\/em><\/p>\n

Patrick: The 50 categories essentially refers to the \u201cnative categories\u201d as given by the app stores. We count 41 on IOS and 30 on Android, and not all overlap. The 550 app and 120 game sub-categories refer to our taxonomy.<\/p>\n

The question of sourcing is a different one. Our underlying database is sourced exclusively from the app stores. The metadata around apps \u2013 their publishers, pricing, versions, ratings, descriptions, etc. \u2013 is taken directly from the stores themselves. We then add our own metrics \u2013 downloads is the easiest example \u2013 which are based on our own models and give additional context to the metadata.<\/p>\n

Our models rely predominately on publicly available data that are supplemented by information we receive from developer partners. We only publish download estimates generated by our market models, and never display the information sourced from developers directly. Developer data is used exclusively for model calibration.<\/p>\n

DFC: You currently have more than 2 million apps in your database, and are adding a further 100,000 per month. What kinds of big picture observations are you able to reliably produce from this data? <\/span><\/em><\/p>\n

Patrick: Well, at the highest level we observe trends in store growth (number of apps, number of publishers, number of downloads) and can break that down by country, category and sub-category as previously mentioned. We are also able to observe key differences across platforms, for example Android users love free apps (there are more free downloads proportionally than on iOS), and iOS users love games (gaming downloads make up 56% of iOS downloads but only 36% of Android downloads, despite an equal proportion of gaming apps).\"Android<\/p>\n

We can also monitor trends in monetization and pricing strategy that is really fascinating. The shift away from paid downloads towards in-app purchases, for example, has been a key recent phenomenon that we\u2019ve witnessed. We can also make very interesting observations related to publisher activities. We monitor when apps were launched, how frequently they are updated, and whether those updates lead to increased performance. On iOS we also follow how publishers expand geographically over time, and whether that process involves localization or marketing efforts.<\/p>\n

Additionally, we see the consumer reaction to apps, via star ratings and comments. For most publishers it is critically important to have a high rated app, and understanding what consumers are reacting to and why is something we can shed tremendous light on.<\/p>\n

DFC: What are the benefits of your big picture focus to developers, publishers and the industry at large? <\/span><\/em><\/p>\n

Patrick: Our proposition is simple. You can\u2019t truly understand how app stores evolve if you don\u2019t consider the long tail. Top charts will always play a role, but they represent only a fraction of all apps in the stores.<\/p>\n

DFC: How many developers are currently participating in your data acquisition? <\/span><\/em><\/p>\n

Patrick: We are in the beta phase of our developer program, and haven\u2019t disclosed the number of participants. What we can say is that we\u2019ve gotten an overwhelmingly positive response from the developers we have approached.<\/p>\n

DFC: By what measure\/standard do you conclude which categories are saturated and which are untapped? <\/span><\/em><\/p>\n

Patrick: This is a topic we continue to study, as we think it is interesting to expose some of the supply\/demand dynamics in the stores. At the moment, we use the number of apps as a proxy for supply, and the number of monthly downloads (in 000s) as a proxy for demand, computing a \u2018number of apps-per-thousand-downloads\u2019 ratio as our measure of saturation.<\/p>\n

Even at the native category level we see very interesting differences. On Google Play, the Business Category is the most saturated, with 2.53 apps per thousand downloads. Gaming categories tend to be the least saturated, with Racing Games topping the list at 0.02 apps per thousand downloads in July 2013.<\/p>\n

On iOS, the Other Games category is the most saturated, with 15.29 apps per thousand downloads in July 2013. Aside from this exception, gaming categories on iOS tend to be the least saturated, as on Google Play. During this same period we found Role Playing Games to be the least saturated native category, with 0.05 apps per thousand downloads. We suspect the \u2018Other Games\u2019 category figure is more a function of an imperfect categorization system rather than a meaningful insight, although it probably tells developers that it is worth picking a defined category when submitting an app.<\/p>\n

\"Article<\/p>\n

There is a lot more to study on the topic of saturation, and we plan to explore this more deeply over time at the sub-category level. We also think a complementary metric to \u2018saturation\u2019 is \u2018concentration\u2019, in other words what proportion of total [category\/sub-category] downloads are represented by the top 5, 10 or 25 apps, ranked in terms of downloads. Looking at the market share comprised by each of these tiers provides additional insight into how the market is structured, and likely makes the \u2018saturation\u2019 statistic more powerful.<\/p>\n

DFC: How do you track monetization?\u00a0 Are you looking for support for trends you believe are occurring, or are you discovering monetization trends in the data? <\/span><\/em><\/p>\n

Patrick: We obviously see the high-level commentary around this topic in the general press, and then look to see whether our data confirms or denies it. As always, the story gets more nuanced when you go beyond the headlines. One of the major trends being talked about is the shift towards in-app purchases and away from paid downloads, and our data definitely confirms that. We\u2019ve seen a consistent growth in recent months in both the number of apps containing in-app-purchase, and the share of downloads of apps with in-app-purchase functionality. On another note, iOS remains well ahead of Android in this respect, likely because of its more developed gaming ecosystem. Finally, we\u2019ve been witnessing more paid apps adopting free-to-download \u2018promotional\u2019 periods, which tends to increase visibility for a period of time before the apps return to a paid model.<\/p>\n

DFC: Are you tracking monetization based on category segmentation, or a game-specific focus? Put a different way, is your monetization analysis geared to specific sub-categories like Action\/FPS\/Sci Fi, or is that analysis title specific? One title in a particular sub-category may suck at monetization while another is outstanding. How do you advise clients what developer execution is trending hot within sub-categories? <\/span><\/em><\/p>\n

Patrick: Our tracking starts at the title level, and then we aggregate up. We compare the average strategies both within and across categories, observe outliers and high-performers, and generally can triangulate to best practices. By monitoring how these dynamics change over time, we can help publishers keep informed of the latest trends.<\/p>\n

DFC: Beyond your existing reports, what kind of data tools are you planning to offer to clients? <\/span><\/em><\/p>\n

Patrick: We see a clear opportunity to deliver additional value from our database through an interactive online format. We often work with industry analysts and researchers, and we\u2019ve observed that everyone has slightly different questions that way want to answer with data. In our position as a provider, our job is to make sure that these questions can be answered in the most efficient way possible, which we can achieve by giving our clients tools to interact with the information themselves. We have some ambitious plans in this respect, and will be rolling out the initial stages of this later this year.<\/p>\n

Priori Info<\/span><\/strong><\/p>\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
Company Name<\/b><\/td>\nPriori Data GmbH<\/td>\nUse this column for notes<\/td>\n<\/tr>\n
Headquarters<\/b><\/td>\nBerlin, Germany<\/td>\n<\/td>\n<\/tr>\n
Founded<\/b><\/td>\n2013<\/td>\n<\/td>\n<\/tr>\n
Employees<\/b><\/td>\n\u00a010<\/td>\n<\/td>\n<\/tr>\n
Website<\/b><\/td>\nhttp:\/\/prioridata.com\/<\/a><\/td>\n<\/td>\n<\/tr>\n
Executives<\/b><\/td>\nList 5-10 top execs<\/td>\n<\/td>\n<\/tr>\n
\u00a0<\/b><\/td>\nPatrick Kane<\/td>\nFounder & CEO<\/td>\n<\/tr>\n
\u00a0<\/b><\/td>\nAnders Lykke<\/td>\nHead of Sales<\/td>\n<\/tr>\n
\u00a0<\/b><\/td>\nMichelle Tran<\/td>\nData Scientist<\/td>\n<\/tr>\n
Business Model<\/b><\/td>\nFreemium & Subscription<\/td>\n<\/td>\n<\/tr>\n
Number of Products<\/b><\/td>\n5<\/td>\n3 Free Research Reports
\n1 Premium Research Report
\n1 Custom Data Service
\n(Bespoke analyses)<\/td>\n<\/tr>\n
Type of Products<\/b><\/td>\nFree Research Reports<\/td>\nFree-to-download<\/td>\n<\/tr>\n
<\/td>\nPremium Research Reports<\/td>\nPay-to-download<\/td>\n<\/tr>\n
<\/td>\nCustom Data Service
\n(Raw Data Delivery &
\nBespoke Analysis)<\/td>\n		One-off and subscription pricing<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

 <\/p>\n","protected":false},"excerpt":{"rendered":"

SEPT. 23, 2013 \u2022 Priori Data Interview: A big issue in the mobile and online space is trying to understand the size of the market and the potential for products in the marketplace.\u00a0 This information is very critical for developers and investors looking to make smart decisions.\u00a0 Unfortunately there has been a great amount\u00a0of hype […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[168,94,40],"tags":[389,12],"_links":{"self":[{"href":"https:\/\/www.dfcint.com\/dossier\/wp-json\/wp\/v2\/posts\/10555"}],"collection":[{"href":"https:\/\/www.dfcint.com\/dossier\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.dfcint.com\/dossier\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.dfcint.com\/dossier\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.dfcint.com\/dossier\/wp-json\/wp\/v2\/comments?post=10555"}],"version-history":[{"count":26,"href":"https:\/\/www.dfcint.com\/dossier\/wp-json\/wp\/v2\/posts\/10555\/revisions"}],"predecessor-version":[{"id":16255,"href":"https:\/\/www.dfcint.com\/dossier\/wp-json\/wp\/v2\/posts\/10555\/revisions\/16255"}],"wp:attachment":[{"href":"https:\/\/www.dfcint.com\/dossier\/wp-json\/wp\/v2\/media?parent=10555"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.dfcint.com\/dossier\/wp-json\/wp\/v2\/categories?post=10555"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.dfcint.com\/dossier\/wp-json\/wp\/v2\/tags?post=10555"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}