One article to solve the search advertising problems on Toutiao and Baidu

Baidu, 360, Sogou, Shenma, and Juliang. Among these, the first four can be understood as traditional searches, and the huge amount can be understood as a new type of search. There are many search channels, including WeChat search, Quark, etc. WeChat search currently has no advertisements, but Quark already has advertisements. Since these two channels are relatively less advertising-oriented, they are not discussed in this article.

Baidu, 360, Sogou, and Shenma belong to traditional search . The characteristic of traditional search is search gene. Currently, these four search channels have two bidding modes, CPC bidding and oCPC bidding. In comparison, Baidu's oCPC is the most complete. The oCPC threshold of 360 Sogou Shenma is relatively high, and the technology is not very perfect. The method described in this article is mainly represented by Baidu bidding.

Massive is a new type of search . Its characteristic is that although it is mainly search-based, it also incorporates recommendation genes. Recommendations are generated by algorithms. The most powerful aspect of the massive platform is its algorithm. Therefore, to promote massive search ads, the main thing is to choose the right targeting and use smart bidding. We will describe the specific method in detail later.

Let’s take a closer look at how to optimize massive search ads. What are the new methods to solve the poor bidding results of Baidu in 2021? This article mainly explains 2 aspects:

Juliang is an advertising platform under ByteDance. The two main search camps of ByteDance are TikTok search and Toutiao search. The Douyin app now has more than 720 million monthly active users, and the Toutiao app has more than 290 million monthly active users. The monthly active users of the two major apps combined exceed 1 billion. And both have launched PC versions. Although some users are duplicates, the monthly active users of more than 1 billion are enough to illustrate ByteDance’s traffic strength.

Massive search ads have been available for a long time, but up to now, the advertising effect is not very good. Regarding the massive amount of searches, here I have sorted out three differences from Baidu:

The keywords don’t need to be too specific. It’s okay to have a plan with hundreds of keywords. However, the keywords must be relevant to the business so that the quality will be high. You can roughly divide the words into parts of speech, set up 4-5 plans, and put one type of part of speech in each plan to test the effect. The choice of words is mainly short words.

The difference between massive information flow and massive search advertising is that search advertising has smart selected traffic. When delivering, you usually need to turn on smart selection traffic. If the matching is messy, you can turn it off. After turning off the smart selection traffic, interest targeting and crowd packages need to be set as unlimited. They cannot be set as customized. Setting customized traffic will result in very little traffic.

When advertising on Douyin, Xigua, and Douyin Volcano version, you need to upload short video materials. If you only want to advertise on headlines, you can just upload pictures. If you invest in both types of media, choose programmatic creativity, set up both short videos and pictures, and wait for the system to optimize itself.

The materials should be more original and updated based on the effects. If the effect is still good, there is no need to replace it. When the effect is poor, you can add more new materials. In actual testing, some old plans have been doing well for a long time, while some new plans are difficult to gain momentum. Therefore, frequent new product launches do not necessarily lead to better results. It depends on the actual delivery results.

It is recommended to check out the ads of your peers more often, learn from their pages, and optimize your own pages frequently. A big reason for poor results is page problems, which must not be ignored. You can learn PS by yourself, which will be helpful for making web pages.

The above is the optimization method for massive search. In fact, it is similar to massive information flow. The core optimization point is still the material. Good material will lead to high advertising explosion rate. The second is keywords, and then targeting and bidding. It is not difficult to create more high-quality materials, grasp the keywords, and do massive searches.

When doing CPC bidding, the first and most important thing is the average price. A lower average price can bring more clicks, which means a higher conversion rate. Therefore, you can use the low price + wide matching method, which is suitable for 80% of industries.

How to keep the price low? Set prices based on the part of speech, set high prices for high-intent words and low prices for low-intent words. Track the average ranking of the account's keywords (which can be seen in the keyword report). Generally, with CPC bidding, the keyword ranking can be 2-4th. There is no need to pursue the first place in all words. Ranking first does not necessarily mean good conversions, and may also result in some waste.

How to wide match? We can use the display volume as the standard. The larger the display volume, the narrower the match, and the smaller the display volume, the wider the match. For small display volume, we can set phrase matching or smart matching. For large display volume, we can set phrase matching or exact matching.

The latest match of Baidu bidding was upgraded this year. The current exact match is different from the previous exact match concept. Exact matches now include both the exact word and synonymous variations, which is wider than previous exact matches. Therefore, even if you set an exact match, it does not mean that only words that are exactly the same will be displayed; synonyms can also be displayed.

There is no clear standard for the proportion of the three matching modes in an account. As long as the search terms are not messy, it is acceptable even if the matches are enlarged a little. It is not the case that the more smart matching is turned on, the worse the effect will be. Smart matching should be released according to the actual situation. If the match is good, set more; if the match is poor, set less.

By continuously improving quality, we lower prices. By combining matching and frequent negative words, the average price can be gradually reduced. After the average price drops, the next step is to optimize the page. Over time, the results will definitely improve. Because the average price is lower and the number of clicks is higher, the conversion rate will inevitably increase.

Compared with the first type, high price + precise matching is suitable for 20% of industries. In some industries, if you try low price + wide matching, the effect will not be improved. At this time, you can try the high price + exact matching method. The advantage of this model is that after setting a high price, the ranking will be higher, and after setting a precise price, the matching will be better; the disadvantage is that it is not suitable for small-budget accounts, because after setting a high price and precise matching, the average price is high, the number of clicks is small, and the conversion is unstable. When traffic is good, conversion rate is high. When traffic is poor, the conversion rate is very low and there is a possibility of wasted time (consumption but no conversion).

Therefore, this method can be used as a second attempt when the effect is not good. Set the price a little higher, set the match to be exact, and make sure the keywords and pages match well. For example, manufacturer words correspond to the manufacturer page, and price words correspond to the price page.

According to my own actual tests, the effect of using high prices + precise strategies in a few industries is quite good. Therefore, you can test it when low-price wide matching does not work.

The advantage of a 0 threshold is that it skips directly and saves data accumulation, while the disadvantage is that it is not very stable. When choosing oCPC bidding method, we usually consider two factors: one is the budget and the other is the daily volume.

If the budget is large, follow the conventional method; if the budget is small, follow the CPC or oCPC with low threshold/eCPC; if the scale is large, follow the conventional method; if the scale is small, follow the oCPC with low threshold/eCPC, or 0 threshold.

Suppose you only have a budget of 300-500 per day, which is not a large amount. In this case, you may not be able to advance to the second level if you run regular, low threshold or eCPC, but you can advance directly to the second level if you run 0 threshold. In the newly created oCPC delivery strategy, choose Skip and the threshold will be 0. Before reaching the 0 threshold, it is better to have a lower average price in the account.

After establishing the 0 threshold, remember to turn on the automatic expansion switch at the planning level (recently upgraded by Baidu Marketing).

The above are the three mainstream methods of Baidu bidding. Under CPC bidding, try the low-price wide-match method first, then try the high-price precise method. Under oCPC bidding, there are three strategies: regular, low threshold and zero threshold. If you have a small amount, the 0 threshold is still worth a try.

The current scale of massive search ads is not very large. The placements can be selected from Toutiao + Douyin + Xigua + Douyin Volcano. When multiple ad placements are combined, the traffic will be much larger. Xiaosou, 360, Sogou, and Shenma need to be run according to your actual situation. 360 is more suitable for PC, but the match is poor; Sogou and Shenma are more suitable for mobile (Sogou's PC traffic is second only to 360). Currently, Baidu is still the largest in scale. In addition to Baidu, we can also consider massive search.

The above is my sharing on bidding advertising optimization methods. These methods and strategies are suitable for all industries. Whether you are CPC or oCPC, you can basically optimize your account from these dimensions. If you haven’t been seeing good results recently, you can try this, maybe there will be a turnaround.