cool hit counter Three Predictions for the Future of App Built-in Ad Realization_Intefrankly

Three Predictions for the Future of App Built-in Ad Realization

Currently, advertising is the most common and widespread way for apps to make money. In fact, a recent study found that most apps derive more than three-quarters of their profits from advertising.

However, advertisers need to keep their eyes peeled as far as future avenues for making money are concerned. A survey of app publishers and ad partners, curated by InMobi and conducted earlier this year, revealed that more than half of respondents said that ad profits have shown a decline or stagnation over the past 12 months.

However, there are those who believe that mobile apps with built-in ads will remain hot for a long time to come, but as new techniques, campaign processes, and technology continue to emerge, the means for publishers to cash in may change.

This article will make three predictions about the future means of realizing advertising.

Guess one: waterfall ads

Models to be replaced by uniform bidding

Despite growing user interest in brands, mobile app built-in advertising profits are not seeing a rise, and one reason is because app publishers and their vendor partners still rely on the waterfall model. In fact, InMobi's survey found that 57% of respondents use this method of ad serving.

In the waterfall model, the sources of advertising demand (networks, transactions, demand-side partners, etc.) are provided in sequence, one after the other, with a full set of potential advertising inventory until the full inventory declaration is completed or the advertising service opportunity disappears.

There are a number of problems with this approach, perhaps the most problematic being that it prevents the potentially largest bidder from actually winning. For example, when a user opens a weather app that contains an ad, the first ad exchange can spend $5 (about 34 RMB) to get the opportunity to show the ad to the user, regardless of whether the ad at the back of the waterfall model is willing to pay a higher amount.

The unified ad auction solves this problem by allowing all potential bidders to meet face-to-face for each potential ad exposure. This is a far cry from the eBay bidding and the Sotheby's auctions. As competition increased and the frequency of ad exposure decisions increased, the unified auction resulted in a 48% increase in ad revenue for those publishers using the service and a 28% reduction in ad serving delays.

Not only that, but the 2017 online data supports this view. With the general trend of unified bidding, expect more and more app publishers to participate.

Speculation 2: App publishers will further

Use advanced, high-impact advertising models

The instinct for profit will lead to an ever-increasing amount of advertising. Once the core experience is disrupted by ads, user abandonment of the app rises.

This means that app publishers will be more concerned about the quality of the ads running in their apps than the quantity. More advanced, high-quality advertising models will also emerge in future application environments.

For example, properly placed video ads or native ads are more impressive and less obtrusive than traditional banner ads, and they can also generate higher cost-per-thousand CPMs for publishers. It goes without saying that a third of current mobile ad spend in the US comes from video ads.

Conjecture 3: AI and Machine Learning

Will have a huge impact on mobile app real estate

Artificial intelligence has ascended to the forefront of historical trends, with AI and machine learning permeating all areas including mobile app advertising real estate at scale.

It's worth noting that perhaps some of the AI applications will soon become popular.

. To ensure that advertisers are guaranteed a fair price for each ad exposure, while not having to worry about secondary price auctions or back-room bidding, ML algorithms will determine the ideal price for each exposure in a timely manner. Multi-arm bandit technology, on the other hand, will help highlight the needs of advertisers for particular users at different times, locations, applications, etc.

. Thanks to AI, the app user experience, including ad presentation and service formats, will be further optimized and personalized.

. AI can more efficiently identify fraudulent factors and remove them before they become a problem. ML algorithms can also learn the adaptation patterns of fraudsters and then instantly fill in the gaps before potential vulnerabilities are discovered and exploited.

The time is ripe for mobile ad real estate

Going forward, app publishers will introduce better ad presentation formats and utilize larger image layouts to deliver ad messages. Of course this is not all in isolation, as unified bidding, efficient advertising models and AI / machine learning will all come into play one after another. (This article was written by Kyon)

This article was edited and typeset by jqyjr

1、Interpretation and tensorflow implementation of the deep learning CTPN algorithm
2、Getting Started with Linux
3、Brain Book Notes Holistic Learning 1 Holistic Learning Strategies
4、What parameters should industrial robots focus on
5、PCA model plus a priori

    已推荐到看一看 和朋友分享想法
    最多200字,当前共 发送