Information flow advertising optimization driven by data monitoring!
|
In future advertising , the selection of information flow advertising platforms and the use of data monitoring tools will play an increasingly critical role. What is information flow advertising? "There will inevitably be a battle between Toutiao and Baidu." I believe many people in the Internet industry have heard similar assertions in the past few months. There is such a huge conflict of interest between Toutiao, which is positioned as an information distribution platform, and Baidu, which mainly engages in search business. The core point is actually information flow advertising. Information flow advertising refers to advertising that is intelligently pushed to people based on search data, social relationships and user attributes when users use Internet products or services, and is mixed with the original information content. Simply put, it is to hide advertisements in news and information flows, blur the boundaries between advertisements and information, and make advertisements less and less like advertisements. Target, a supermarket chain that was very popular in the United States in the 1980s, once tasted the sweetness of information flow advertising. At that time, Target attached great importance to the concept of data-driven operations . They set up a special data statistics department and analyzed the most valuable consumer group - pregnant women. Pregnant women often spend several times more than usual before and after pregnancy, and the items they buy most often are also easy to identify - milk powder, strollers, diapers, etc. Based on the above information, Target's data personnel screened out the pregnant women among the supermarket members and mailed them coupons for mothers and babies . Ideally, this approach should work very well because they accurately identify their target customers. But the results were not satisfactory. The supermarket received a large number of complaints. Many consumers believed that their information had been leaked and they felt uneasy or even panicked. After repeated consideration, the supermarket adopted a new strategy. They mixed the maternal and child coupons with other unrelated coupons and mailed them together to their target customers. Among the large number of coupons received by consumers, only a few are maternal and child coupons that supermarkets really want to send. As consumers' panic decreases, the promotional effect of coupons is getting better and better. Information flow advertising is such a model. Although it can locate users very accurately, it will still hide itself among a lot of irrelevant information, reducing users' panic about data leakage. It not only inherits the precise targeting of DSP ads , but also does not destroy the browsing experience brought to users by the platform on which it is delivered. Achieve brand promotion and marketing without causing consumer disgust. In the future, we will be exposed to a lot of information flow advertising, or advertising forms with information flow as the core. The model of mainstream information flow advertising and its current defects and shortcomings Blind design, blind placement, blind optimization - this may be one of the biggest problems with information flow advertising at this stage. Often, due to limitations in qualifications and experience, advertising and optimization personnel are unable to determine the most core performance indicators and develop a scientific and effective optimization process at the outset. Lacking experience and bearing enormous pressure from both parties, many relevant practitioners rush online without considering whether the advertising creativity is high-quality and whether the delivery platform is suitable, and then start rounds of aimless optimization. Let’s first look at the product mechanism of mainstream information flow advertising: eCPM (pre-allocated impressions) = bid * quality ( CTR ) When a user opens a piece of information on an information platform, a corresponding advertising request will be generated immediately. First, the system will perform an advertisement search and select suitable advertisements. These ads are then pre-allocated impressions. According to the above formula, the pre-allocated impression (eCPM) is equal to your bid multiplied by the quality score (CTR); and this CTR depends on the consistent performance of your account. After sorting the ads by pre-allocating exposures and filtering out some frequencies required by advertisers, real ad exposure can be achieved. At this time, when the user clicks on the ad, advertising charges are generated. Among the above product mechanisms, the part we are most likely to intervene in is CTR. In a business environment with fierce competition and limited budgets, our bids cannot be infinitely high. Through human intervention, we can let the system determine a higher CTR for us, so that we can use a lower bid and get the same or even higher exposure as our competitors. Ideal CTR optimization positive cycle: the advertising creative has a high click-through rate , is judged as a high CTR, gets a high pre-allocated exposure, and thus has a high advertising exposure, and good performance will be accumulated in the account's historical data. If you add new advertising creatives at this time, the CTR will be very high, thus getting high exposure again, and repeating this cycle to achieve the ideal CTR optimization positive cycle. Disadvantages and shortcomings: It is impossible to determine whether new ideas are high quality, and it is impossible to guarantee continuous high CTR. Actual cycle: There is no direction for optimization of advertising creativity, and high CTR cannot be achieved. Advertising is rushed online due to time pressure, and there is a lack of reflection and efficient optimization process, which eventually turns into a vicious cycle. The picture above is Converse’s advertising idea: the bold and innovative design may seem unsightly, but sometimes it can attract users’ curiosity and bring in a lot of traffic . For relatively novel creative designs like this, without the guidance of the launch results, it is difficult for us to judge the quality of the idea based solely on the design drawings. Information flow advertising optimization model with monitoring as the core Monitoring is the core and data is the basis. For most advertising practitioners, in order to avoid the limitations of qualifications and experience and establish an efficient and scientific optimization process, they must focus on data monitoring to guide keyword optimization, creative optimization and design optimization. Below we use a real case to introduce how to use data monitoring to optimize advertising creativity. In order to guide users to register, most local dating apps will place a large number of photos on the homepage to attract users to click. When it comes to the choice of photos, there are two creative designs that are at odds with each other. One is to use beautiful but not real pictures, and the other is to use real but not beautiful pictures. Party A considers from the perspective of software operation and hopes to use pictures that are very real but not beautiful enough in order to highlight the authenticity of the dating software. Party B considers user attributes more and hopes to filter out those who will not register by using beautiful but unrealistic pictures from the beginning, so that the remaining users will be more easily converted into paying users. Both sides have their own starting points. At this point, it is difficult for us to judge the pros and cons of these two creative designs based on experience. We can only adopt data monitoring methods to observe the effects of the two creative delivery methods and then guide the optimization of creative designs. It should be noted that the creative optimization process, in addition to considering images, also includes the matching of text and images, etc. The data-driven optimization system is much more sophisticated and complex than previous advertising optimization, and it is no longer possible to achieve a perfect result based on experience alone. Conventional data monitoring usually involves monitoring at the installation package/ landing page level. Different landing pages correspond to different installation packages, which are then assigned to different promotion sources. However, this can only determine the quality of a certain type of creativity on a certain page, and cannot achieve more detailed data tracking. To solve this problem, we used the data monitoring tool ptengine to generate advertising links. By refining the UTM parameters (detailed marking of the creative name, source, media, keywords, content, etc.), we were able to track data at the creative level. Through the above parameter settings, we can separate pictures and texts and achieve refined data monitoring at the creative level. Use different copywriting for the same picture, and then use different pictures for the same copywriting to achieve creative cross-contrast (as shown in the figure below). In this way, we can understand the most effective copy and the most effective picture respectively, and we can also know which combination of copy and picture has the best effect. When breaking down different creative ideas, we first classified all images based on their appearance, from 3 to 7 points. Each score was further divided into bare face, light makeup, and heavy makeup. If you use three photos of the same person, we define it as a set. Before launching, we set links for each creative, and each link corresponded to relevant traffic data (views, exit rate, average stay time, etc.). After clicking each link, you can also set the "Click to Download" button as an event for tracking. In this way, we can clearly understand the interaction after each creative brings traffic to the landing page and estimate the final download volume. When we cannot directly link creative ideas to conversion results, we can make a good estimate of conversions by setting the "click to download" button as an event for tracking. After a period of data collection , we combined the cost of each creative front-end with the number of interactive clicks on the back-end "click to download" button to calculate the interactive click cost. By comparing all creative ideas based on the principle of "more interactive clicks and relatively lower costs", we found the best image idea, the best copywriting idea, and the best image and text combination idea (marked in yellow in the figure below). The elements and designs used in these creative ideas are the general design direction of future new ideas. Next, we will introduce another case and see how to use heat map monitoring to optimize the design of the landing page. When doing marketing and information flow promotion, heat map analysis tools can help us analyze the user's behavior after visiting the landing page, check whether the design of the landing page is reasonable, and use data to drive the optimization of the landing page. When we have no way to directly monitor the final conversion results, the landing page, as an intermediate link, is also an important basis for estimating the conversion effect. For example: we can see where users clicked, whether some button settings may cause misunderstandings among users; in which areas users stayed and what content they were most interested in; the loss of users as the page scrolled, whether the current landing page is too long and caused users to lose patience, etc. By continuously improving the design of landing pages, we can enhance user experience and promote conversions. In the actual landing page optimization process, we often face the dilemma of "comparative testing first, then data" and "without data, it is difficult to promote changes to the landing page." At this time, optimizers need to introduce A/B testing, make small adjustments to the landing page quickly, use diversion testing to determine the optimization effect, and then return to guide the design of the landing page to achieve data-driven iterative optimization. It should be noted that data monitoring under normal conditions is not detailed enough, and we cannot determine which factor caused the change in the data. Therefore, when performing A/B testing, only one set of variables should be set at a time, and the testing and comparison should be carried out as carefully as possible to achieve the optimization goal scientifically and effectively. Information flow advertising optimization, with data and heat map monitoring as the core, truly realizes refined monitoring of the entire process from design to display, clicks, and landing page interaction, based on creativity and driven by data. Compared with other forms of advertising, information flow advertising is highly accurate, less intrusive, and highly integrated with the user's browsing environment. Combined with the optimization model driven by data monitoring, the work of advertising practitioners becomes more scientific, efficient, and process-oriented. In future advertising, the selection of information flow advertising platforms and the use of data monitoring tools will play an increasingly critical role. Choosing an excellent data and heat map monitoring tool can help you achieve twice the result with half the effort in designing advertising ideas, selecting delivery channels , improving work efficiency and establishing efficient optimization processes. Especially for less experienced advertising practitioners, data operation tools will become their best right-hand man.Mobile application product promotion service: APP promotion service Qinggua Media advertising The author of this article @Qu Haijia is compiled and published by (Qinggua Media). Please indicate the author information and source when reprinting! Site Map |
<<: Cooperative SEO training: Why can’t I rank high after doing so much?
>>: App Store logic: How to get Apple's recommendation position
Recommend
Anhui lifts lockdown: Which public service places can open? Attached notification details
As the number of confirmed cases in Anhui is redu...
Pear Diet: Why does a website's homepage ranking depend on what users want?
It is no use having a good website that just look...
Douyin VS Kuaishou - Competitive Analysis
Live streaming e-commerce has experienced rapid g...
How to awaken the product’s communication power? Share 6 methods!
Once the product is made, the big challenge is ho...
New user growth strategy: from traffic to retention
Traffic is the most valuable thing on the Interne...
7 Christmas marketing cases worth seeing, take them and thank you!
Christmas is here, and it’s time to have fun agai...
Zhejiang added 1 new local case and the country added 54 new imported cases!
From 0:00 to 24:00 on March 26, 31 provinces (aut...
When it comes to product promotion, are you a domineering president or a stingy boss?
The author of this article will share with you 4 ...
Ling'er Investment Research Diary "Golden Buying and Selling Points" Gold Buying Points & Gold Selling Points Private Edition
Ling'er Investment Research Diary "Golde...
Content operation, do you know the strategies of these four important links?
This is an era of information overload. Even very...
ClickHouse application practice and principle analysis
Course Description ClickHouse is an open source c...
The most comprehensive guide to promoting good products on Zhihu
There is a very good question on Zhihu: Why do or...
How much does it cost to be an agent of Wuxi Home Improvement Mini Program? What is the price of being an agent for Wuxi home improvement mini program?
Why should you be an agent for WeChat Mini Progra...
Which Olympic Games is the 2022 Beijing Winter Paralympics? Attached is the start and end times of the 2022 Beijing Paralympic Winter Games!
After Beijing's successful bid, it became the...
Guangzhou WeChat water delivery mini program function, how much does it cost to develop a bottled water door-to-door delivery mini program?
With the acceleration of my country's urbaniza...