游戏邦在:
杂志专栏:
gamerboom.com订阅到鲜果订阅到抓虾google reader订阅到有道订阅到QQ邮箱订阅到帮看

开发者评App Store搜索算法问题及改进空间

发布时间:2012-05-03 17:24:34 Tags:,,,

作者:Brian Stabile

手机应用市场看似一个最容易发布产品的领域——你只要向苹果支付99美元年费,就可以让全球成百上千万用户立即从中购买并享用你的产品(或者批评产品)。两年前可以说是成立手机应用工作室的最佳时机,但现在遍地都是手机应用公司,诸多初创团队的涌现无疑加剧了开发者推广产品和实现下载量的难度。而iOS App Store本身的设计也在一定程度上成了我们走向成功的“拦路虎”。

你在iOS设备上打开App Store时就会在页面顶端看到一个搜索框,但不得不说它的设置确实有失水准——-就算你准确无误地输入应用名称,你的搜索对象也未必总是首个搜索结果!就算应用商店的搜索算法会考虑总体/每日下载量,评级和评价情况来排名,那也应该优先将准确的匹配对象显示为首个搜索结果。假如有多个应用名称相同,那么就应该根据它们的名称与搜索字段的相似度来排列搜索权重。但我们发布的《The Last Ace of Space》这款游戏过了三四个月后,在搜索框中输入这一名称时它才呈现为首个搜索结果。截止本文撰稿(游戏邦注:2012年5月2日),这款游戏的免费版本《The Last Ace of Spce Lite》仍然只能显示为第11个搜索结果(即最后一位),落后于另外9款在名称中根本不含相关字段的应用(它们甚至不含“The”和“of”)!

终于显示为首个搜索结果(from gamasutra)

发布三四个月后终于显示为首个搜索结果(from gamasutra)

输入“Last Ace of Space”字段,游戏的完整版显示为第5个搜索结果,而精简版仍是最后一位。

输入“Ace of Space”显示的搜索结果则是名称与这个字段一模一样的游戏《Ace of Space》(其精简版排在末位,而我们的完整版则是第6位),虽然这款游戏名称也含有我们的搜索字段,但却根本没有出现在的其他几次搜索结果中。

显示为第6位(from gamasutra)

显示为第6位(from gamasutra)

输入“The Last Ace”可以看到我们的游戏排在第6位,但如果不下拉屏幕根本看不到它的图标。

不下拉屏幕根本看不到这款游戏(from gamasutra)

不下拉屏幕根本看不到这款游戏(from gamasutra)

输入“Last Ace”却可以看到精简版和搜索结果排名上升了,这到底是什么状况?

App Store的这个搜索系统无疑让我们的游戏失去了许多下载机会——假如我告诉别人“去看看我的iOS游戏《The Last Ace of Space》吧”(我也确实经常这么做),那么对方根据完整的名称搜索这款应用的概率有多大?从我个人观察人们使用搜索引擎的经验来看,多数人会搜索“Last Ace”、“Ace of Space”、“The Last Ace”这些字段,假如搜索结果没有将这款应用显示在前5个排名(游戏邦注:iPhone用户如果不下拉屏幕,他们一次最多只能看到5个搜索结果),那么他们很可能就会放弃搜索,并且认为我是为了引起他们注意才编造了这个谎言。除此之外,我认为App Store搜索排名还应该考虑一些标准选项,例如“公司名称”、“应用名称”、“产品描述”、“评分”、“文件大小”(最好让处于非Wi-Fi区域的用户搜索到“不足50MB”的应用)。另外还应该设置一种应用下载“排队”渠道,以防止用户在某个时刻出于某种原因暂时无法下载应用。假设这种情况,有一款应用推出24小时限时免费促销活动,而此时我的iOS设备恰好内存不足,或者该应用超过了50MB,而我又正好处于非Wi-Fi区域,我就不能购买该应用,我就会错过这次优惠活动(此时用户定会备感受挫)。类似于iTunes应用商店中音乐及播客内容的操作方式,最好允许用户将应用加入“下载”菜单排队等候,直到其设备准备就绪才进行下载。

另一个问题是,用户差评对App Store应用排名似乎并没有多大影响。我当然理解,如果将用户评价纳入排名算法参照标准,有些开发者就可能利用这种排名系统,花钱雇人大肆诋毁竞争对手的产品,从而使其跌出排行榜前列。但我还是认为,苹果工作人员应该定期核查前200名应用,加强质量把关。目前形势来看,多数用户很少关注应用评价情况,他们通常会下载那些拥有出色产品描述,或者含有一些误导性/欺骗性截图的应用,或者一些擅长玩文字游戏,误使人认为它是某热门应用推出的官方衍生版本之类的山寨应用。这类应用通常都会收获很高的下载量,并由此晋升榜单前列,即便用户给予大量1颗星的差评。多数用户只查看应用商店中前20-25名榜单,他们心里会错误地假设“这些应用质量一定都很可靠”,然后无视其评价情况就下载应用,后来发现问题才给予差评。假如App Store搜索算法对评价情况给予更多权重,那么低劣或欺骗性应用就会迅速跌出榜单之外,就不会有那么多用户买到位居榜单前列的劣质产品。

我们重新审视App Store当前的状况,看到它不合逻辑的搜索算法,对用户差评以及恶意应用的放纵态度,就会清楚为何那些非热门应用开发公司无法获得下载量,即便用户搜索的是准确无误的应用名称!对我们这些没有多少广告预算的“小人物”来说,最好的方法是在进行推广时添加应用的直接下载链接(如果使用analog广告,就要含有QR码,如果是数字广告,就要含有应用在App Store页面的直接链接),因为你要是让用户自己去应用商店搜索,他们很难找到准确的结果,并可能由此失去兴趣,而你也将失去一次下载量。(本文为游戏邦/gamerboom.com编译,拒绝任何不保留版权的转载,如需转载请联系:游戏邦

Seeking a Better Search, or Why Apple’s App Design is Bad For Business

by Brian Stabile

If you’re like me, you’re a really cool guy with exquisite taste, and we should hang out. You also are trying to develop apps for the mobile market. Why wouldn’t you? The app market seems like the easiest avenue to launch a title- Millions of people worldwide can buy your product and start enjoying it (or harshly critiquing it) instantaneously, with nothing more than a $99 fee to Apple! What better time to start a mobile app studio than… two years ago! App companies are anywhere and everywhere at the moment, the (*metaphor warning*) ‘dubstep’ of the software industry – there’s a large demand for them, everyone in the industry has a side project developing them, you can’t tell the difference between most of them, and consumers and self-proclaimed ‘fans’ only really know of the same four or five of them. (Please rate my metaphor in the comments. Thanks.) The massive influx of new companies makes it massively harder to promote your releases and get downloads. The design of the iOS App Store has also somehow succeeded in beinganother enemy on the path to appcredible success.

Upon opening the App Store on your iDevice, there is a handy (Apple would say ‘elegant’, ‘mystical’, or ‘grandiloquent’) Search tab that opens a single textfield at the top.  However, it is plagued by a flaw of gigantor proportions – SEARCHING FOR AN APP BY INPUTTING ITS TITLE VERBATIM DOES NOT ALWAYS RETURN THAT APP AS THE FIRST RESULT! While you would expect the search algorithm to weigh the results by criteria such as number of downloads total/daily, number of ratings and reviews, the fact of the matter is, IF THE SEARCH STRING IS AN EXACT MATCH WITH AN APP TITLE, THAT APP SHOULD BE THE FIRST RESULT. If there are multiple apps of the same name, it should decide where they fit on the list by then factoring in the other criteria, then followed by adding a weighted score comparing how close the title’s name is to the search string to the rest of the results. We have witnessed this first-hand: after the launch of our game ‘The Last Ace of Space’, it took about 3 or 4 months before it became the first result when you searched ‘The Last Ace of Space’. As of writing this (May 2, 2012), the free version, ‘The Last Ace of Space Lite’ STILL shows up 11th place – the LAST result of the search, trumped by 9 other apps that literally do not have a single word from the search string in their titles – even ‘The’ and ‘Or’!

(Are these the results you would expect from this search?)

-Searching ‘Last Ace of Space’ puts the full version 5th on the list, and the Lite version is still last.

-Searching ‘Ace of Space’ returns a game exactly named that as the first result (with their Lite version dead last, and our full version 6th place), yet it didn’t even appear in the results for the other searches, disregarding the fact that the entire title was within the search string.

(Why was the app ‘Ace of Space’ not even listed on the other two searches?)

-Searching ‘The Last Ace’ puts our game at 6th place, which is not visible on-screen with the initial results returned by the search.

(Something is very wrong here.)

-Searching ‘Last Ace’ actually moves the Lite version higher up on the results! WHAT IS THIS MADNESS, AND WHO DO I DIRECT MY HATRED TOWARDS?

This system no doubt has destroyed many opportunities for us to get downloads of our game – if I tell someone I meet to “check out my iOS game ‘The Last Ace of Space’” (which I do constantly, believe me), what is the likelihood that person will search for the full title? More than likely (based off personal experience observing how people utilize most search engines), they will search ‘Last Ace’, ‘Ace of Space’, ‘The Last Ace’, and if the app is not found within the first 5 search results (that’s all that the user sees on the iPhone screen without having to scroll down), they will most likely give up and/or assume I made it all up to impress them. Additionally, there should be options to choose what criteria to search by, such as Company Name, App Name, App Description, Rating, File Size (Imagine if you could search for ‘Under 50MB’ apps when you are in a No Wi-Fi zone). There should also be a way to ‘queue’ apps that can’t be downloaded onto a device at the moment (for whatever the reason). For instance, if an app runs a promotion where it is free for 24 hours, and my device is either out of space or the app is over 50MB and I’m lost in a No Wi-Fi zone again, I can’t ‘purchase’ that app onto my account, so I miss out on the sale (NOOOOOOOO!). Apps should be queued onto a ‘Downloads’ menu, similar to how music and podcasts are in the iTunes app, and then downloaded later when the device is able to do so.

Another issue is that large amounts of negative reviews should count more negatively towards an app’s rank than they do at the moment. I understand that it COULD be possible to use one of those ranking networks where you pay money to receive reviews and rankings (which I’m pretty sure will get your account banned if Apple finds out) to tank a competitor’s app with an onslaught of one star reviews, but Apple should have people on staff to frequently test games up near the top 200 of the charts to check for quality control. The trends seem to indicate that most consumers disregard the app ratings anyway, which normally leads to this common scenario: people download an app that has a very appealing app description, or something misleading, such as deceptive screenshots, or a cleverly-worded title that makes the consumer think it is an official spin-off of a successful app. This results in a lot of downloads, and that app shoots up the charts, even though most people leave 1-star reviews. Most users only check the top 25-50 charts in the app store, and if the app makes it that far, there is a false assumption that it ‘must be good’, and it gets downloaded, ratings disregarded, and the new users end up leaving 1-star reviews of their own. If the review ranking held more weight, a low-quality or misleading app’s visibility would drop quickly off the charts, and less people would end up buying a poor product, despite being at the top of the charts.

Reviewing the present condition of the App Store, with its illogical search algorithm, and relaxed attitude towards ranking poorly reviewed and malicious apps, it becomes yet another obstacle for the companies whose apps aren’t blockbuster hits to get downloads, sometimes even when the customer searches for the exact title! For us ‘little guys’ without much budget to advertise, make sure that when spreading the word you include a direct link to your app (if using analog ads, include a QR code, for digital ads, include a direct link to your app in the App Store), because if you leave it up to the user to search for your app themselves, it might be too hard to find, and they’ll lose interest, and you’ll lose a download.(source:gamasutra


上一篇:

下一篇: