Optimization example of Baidu bidding in the machinery industry

In October, I took over the Baidu bidding account of a small client. The daily budget is 150. Many larger teams may not do it with such a small budget. Because the people I was working with were very sincere, I took the job. It is not easy to run a large search of 150 people a day, but there is a way. The operation was carried out on October 10th, and this is the conversion situation by 31st. It’s not particularly good, so I’ll share some of my personal experiences and see if they can help you.

1: Positioning strategy

First, set a strategy.

Bidding strategy. For small accounts, they can only run CPC, especially 150 a day, and they cannot run oCPC at all. After the second level of oCPC, the click price of a word may be seventy or eighty. Therefore, for such small accounts, the most appropriate method is low price + wider matching to lower the average price. The average price decreases, the number of clicks increases, and with more clicks, the probability of natural conversion will be higher. Look at my data. Before October 15, the average price was greater than 1.9 for 4 days, but after October 26, the average price has dropped to around 1.2. So from the perspective of average price, it is very effective.

Device policy. In the machinery industry , everyone knows that although the traffic of PC is relatively small, its quality is relatively higher. Although there is a lot of traffic on mobile devices, the quality is actually average. My original intention was to invest in PC instead of mobile, but the account was hosted by Jimuyu. Since the Jimuyu mobile page performed better than the PC page and the budget was small, I had no choice but to change my strategy to invest mainly in mobile, and set the computer bid ratio to 0.3 to ensure that the PC could have some exposure and consumption.

Time period strategy. There is no doubt that when running searches, the quality of traffic is relatively high during the day, and relatively low at night, so the quality of traffic should be high during the day, low at night, and even lower in the early morning. The reason why I run in the early morning is mainly to improve the quality. It’s not like there will be clues in the early morning. The following is my time period ratio. 8-9 am, 12 noon-2 pm, and 7-12 pm are all the same ratio. 0 am to 8 am is one ratio, and the working time period is another ratio. This ensures that you arrive at work at least within the budgeted time.

The above are adjustments made based on the budget, bidding strategy, time period and device.

2: Detailed strategy

Detail strategy. Actually, there are three points: bidding, matching and negative words. These are the three tricks of Dasou.

Bidding, based on a budget of only 150, there is no need to bid high. You don’t have to worry about the prices offered by other competing products; all you have to worry about is your ranking. With a budget of 150, generally speaking, ranking 2-4 is enough. For ranking, you can refer to the average ranking in the keyword report in the data report, which is relatively accurate. The rankings from front-end searches and real-time query rankings are not very accurate, so don’t use them as a reference. Another point to add is that there is no need to consider the ranking when making oCPC bids, as it is meaningless. When bidding on cpc, you can focus on the ranking.

Generally speaking, to look at ranking data, you need to download reports for three days or more. Don’t just look at one day’s data. The data for one or two days is not very accurate. You need to look at data for at least three days. Then make fine adjustments based on your ranking, keeping good words at second or third place and average words at fourth or fifth place. If the keywords with high intentions are ranked low, the price will be increased by 5%-10%. If the words with poor intentions are ranked high, the price will be reduced by 5%-10%. With each adjustment, the average price can be slowly reduced.

Matching strategy. It can be said that matching is the most core part of CPC bidding. Matching is not only related to traffic quality, but also to the average price. As mentioned before, for small accounts, the most suitable method is the low price + wide matching mode. The broad matching here mainly includes phrases and intelligence. For small budget accounts, you can also set precision, but not too much. Look at my account. I never set it to be precise. Even if I do, it won’t exceed 10. Not all accounts can be set to smart mode, otherwise the matching will be very poor.

How to set phrases and intelligence? Everything is based on display data. If a word has not been displayed after a week or half a month after phrase matching, I will set the match to smart matching and reduce the bid by 10%. For smart matching words with a large display volume, shorten them to phrases and increase the bid by 5%. What kind of idea is this? It’s very simple. Words that are not displayed cannot get clicks, so we need to loosen up the matching to increase the display and click volume. Under smart matching, words with too large a display volume are prone to random matching, so they must be reduced. This keeps the match from being too bad.

(As shown in the picture, select the phrase, change the undisplayed ones to smart ones, and change one displayed one to the phrase, so the words that match the phrase in the picture should have at least one displayed one)

(As shown in the figure, select smart matching, and the words with large display volume will be reduced to phrase matching, so the smart matching words in the figure will not be too large no matter how large they are displayed, and the smart words with high display volume will be changed to phrase matching)

Generally, you only need to fine-tune the matching once a week. You don’t need to adjust the word count that much each time, and it mainly depends on the display data. Of course, you can also adjust the match by referring to other dimensions, such as matching degree. From my experience in advertising over the past few years, I know that data is the primary consideration, followed by matching, and then other dimensions.

The third key point is the word "yes". Among the targeted suggestions of Baidu Search Optimization Center, there are two very good suggestions, one is to remove conflicting negative words, and the other is to remove unpopular negative words. I use these two functions a lot. If my account has conflicting negative words or unpopular negative words, I usually remove them. Clean up the useless negative words, so as to increase the capacity of the negative word library. Another point is that conflicting negative words will cause the account's keywords to not be displayed, so they must be cleaned up in a timely manner.

The strategy of denying words is generally to deny words once every 3 or 5 days, and to deny some words with irrelevant traffic in time. There isn't much to say about this ring. One more point to add is that there are three modes for expansion: not enabled/matching with smart selection/automatic expansion. If you run CPC, either not enable it or match with smart selection, and do not choose automatic expansion, otherwise the search terms will be messy. If you run oCPC, then turn on automatic expansion. For small accounts, I suggest not to enable it. The budget is limited and you don’t want to add any more low-quality traffic.

3: Conversion strategy

Through account positioning + detailed strategies, the account can basically ensure that three things are done well. First, the average price will gradually decrease, second, the quality will gradually improve (all-day investment), and third, the traffic will gradually become accurate (matching optimization + negative word optimization).

The next step is to increase conversions. Improving conversions mainly starts from the page. Since the page is invested in Jimuyu, there are two optimization elements involved, one is the page content layout, the other is the conversion component setting.

For the layout of page content, we use the mobile phone screen as the standard. The large search page is usually 4-6 screens long. The first screen is mainly the banner + form component, the second screen is the content that meets the user's pain points, the third to fourth screens are information such as self-advantages and price discounts, and the fifth to sixth screens are secondary information and secondary conversions (sub-forms or bottom-up conversions). The focus of the landing page is on the first two screens. In addition to the page content layout, the page content is also very important, and all content should be produced around user pain points .

The second is the conversion component. Generally speaking, when doing a big search, the mobile page needs to have a form + consultation + phone number + WeChat, and a callback can also be added on the PC. Jimuyu’s components also include coupons, lotteries and Baijiahao. But these are only used for activity pages and promoting Baijiahao. Since the account I run does not have a consultant, I did not add consultation, but I did add forms, phone calls, and WeChat, which can increase the conversion rate.

For small budget accounts, you can only improve your conversion rate by fully utilizing each conversion component. This is something that many bidders overlook. If you are making an event page, you can also use coupons and lucky draws.

After 21 days of optimization, the account clearly saw some results after three days. Then starting from the 19th, except for a few days without clues, there were clues on other days, and the amount of clues also increased significantly compared to the initial stage of the launch. Therefore, examples can prove that this approach is effective. Under the current circumstances, is it still possible to increase the number of leads? I think there is still room for improvement. First, we can lower the average price. Second, we can optimize the pages and make new versions with more marketing-oriented pages. This way, there should be room for optimization in conversions. If the budget is increased to more than three to five hundred, the effect should be better.

The above is an example of a small account operated by our team. This can give you a reference, and I hope that friends who have small budget accounts can also get inspiration from it.

I have managed medical accounts with daily spending of 10,000 or 20,000 yuan, and I have also managed accounts with daily spending of 150,000 yuan. Regardless of whether it is a large account or a small account, the policy must be tailored to the account. In some industries, a small budget really won't produce any results, but in some less popular and relatively traditional industries, the results actually have little to do with the budget.