Management optimization method of super-large SEM account

Intermediary transaction http://www.aliyun.com/zixun/aggregation/6858.html ">seo diagnose Taobao guest cloud host technology Hall

Yesterday and a friend on the Super large account SEM management of the problem made a little discussion, think may be useful to everyone, sorted out added, posted.

The account structure involved in the question and answer is in fact the most important part of the medium and large account management, this time not unfolding, if the reader is interested, can be a separate discussion.

Q: More than 100,000 keywords of the Super large accounts, how to group easier to manage it? Now this account is mainly divided by type, such as industry/general/doubt/competition, but feeling management is very cumbersome.

A: For large accounts, the account structure is the most important part of SEM management. In essence, the size of the account, the amount of keywords in SEM management and no essential differences. The main problem with large accounts is that there is too much content, easy to get lost in the details, resulting in the inability to effectively identify priority issues and solutions. A clear and reasonable account structure is designed to improve the readability and operability of accounts and prioritize, and management is not cumbersome. In principle, I think that even millions of of thousands of keywords do not need dozens of people to operate, 35 is sufficient.

Q: Yes. Now there is a problem. My client has a couple of little girls on their own, and they often take the low quality out and rebuild their plans. As a result, I went into a look of the crazy, messy, big move, the estimated data volatility will be more severe.

Answer: Take over new account to avoid big move, move too much easily cause violent fluctuation. Even if the new account structure should be gradual, for customers, on the basis of a stable optimization more meaningful. Volatility itself can also cause more obstacles to optimization.

Q: So just the top 100 adjustment is OK? I'm going to sift out top 100 consumption keywords, and related search terms

And then group again.

A: You can do this, for example, choose one of the largest project industry to do a sample, choose the most meaningful keyword reorganization a Test series. Then optimize this series. After the initial optimization, replace the other project/industry with the same test group concept, change it a little, don't get rid of it all at once.

Q: I have seen you previously write Baidu Intermediate training ppt, to test a group of keyword combinations, this does not have a system, it is difficult to achieve it?

A: Not necessarily difficult, Excel is enough. The main structure is to build their own first, do the test when the layout will come out, and then gradually adjust.

Q: Yes. Do you have a sample of Excel data analysis on your side? I'll see how you do it. Demo data on the line.

Answer: No ... This is not complicated, basically is a simple sort and effect comparison. But you need experience? structure, in addition to their own organization, but also with their webmaster or business owners to chat, with the site structure or business processes, to report readability and operational enforceability as the goal.

Q: I want to know what you are doing in these optimization efforts, is to focus on which aspects of data fluctuations. It's intuitive to look at a table.

A: Of course the effects of the old and new series are compared. The small test group is this purpose, to ensure that the data before and after testing can be compared to the consumption and conversion as the main indicators, display click and CPC as a means.

Q: Is the ad group guaranteed to be relevant? The relevance of creativity and keywords, different purposes, the structure will also have a difference.

A: Your Test series is designed with the best goal, otherwise it's meaningless. After the test is successfully replaced, the old words are phased out to make sure the new structure is clear and controllable.

Q: Yes. I see. Sometimes it's good to start planning, but later optimization is very likely to mess up the account.

A: Well, it's a matter of experience. Or the idea is not clear, a lot of things will lose the order of precedence. Do SEM management to get out, always understand the core of the problem, at any time can grasp the main line, Gangjumuzhang, will not be inundated by a lot of details later.

Q: In this case, I keep adding quality search terms when I test, so it's difficult to compare. If not, only in the CPC, creativity, landing these several links to optimize.

A: The test only for the old, do not complicate, expanding the word is to replace the old keyword old series of things. Optimization must be implemented step-by-step, each time for the limited target, otherwise you do not know how to evaluate the effect. You do account optimization a little anxious, always want a step in place.

Q: Yes. What kind of steps do you usually have here? Yes, I'm so mature and sophisticated. Mainly customers are also anxious, there is no way.

A: Some jobs are Sumer can not be anxious, the key is to confirm the optimization and optimization means to replicate, which requires data accumulation to verify. Light anxious useless, buy lottery tickets can not be the second time, replication is the key to system optimization.

Q: You probably guide the implementation steps. I keep a good record and follow this to see the effect.

A: First of all, we need to address priority objectives. From the consumption transformation of two angles, select the most influential series or keywords, set up an ideal test group or Test series. Now Google accounts in Europe and the United States to support testing in the series settings, can be in the running state of the test. Chinese users use Baidu or Google China, the use of traditional means: is to stop old keywords old series, start Test series, compare the test before and after the same time cycle effect.

Q: Google's test is the old and new series/group partitioning, setting different two CPC bids to see the final comparison results?

A: not just the CPC. In fact, when the first step of optimization, CPC is not important, the important thing is the structure. The ideal structure must be implemented in the Test series, then CPC, advertising, LP, dependencies, and so on. After the initial test is complete, copy the pattern to the other target series. In the new series that has already completed the new and old replacements, it is possible to further expand the word/budget/improve advertising and other optimization. In fact, it is not complicated, the main idea is clear, each step is aware of what they do, whether the goal is achieved, whether and the means of implementation to match.

Q: Well, I've recorded it. Another problem is that some industries are more volatile, such as a 30% of the normal volatility of the week.

Answer: One is to lengthen the observation period. The second is to lock the series, grouping or keywords, eliminating fluctuations to see the overall effect. Fluctuations are normal, but all grouped words fluctuate together, if there is no external influence is not normal.

Q: Overall volatility is normal in some industries.

A: You don't understand what I mean. If you think a certain fluctuation is normal, put the fluctuation range into effect observation, which is the same as the meaning of lengthening the observation period. That is, if 30% of the fluctuations are normal, then our optimization can target 50%. On the other hand, you can reduce the impact of short-term volatility by weighting all the data in the 7th average, the 15th average. After all, it's a matter of thinking: you set up a structure and optimization plan, how to implement and effect evaluation.

Q: Thank you very much, listen to this one say, feel the thought opened. Always wondering how to weighted from data analysis to eliminate the effects of normal fluctuations.

A: You are welcome, you have to be patient, the main is to identify problems, determine the solution to the problem, confirm the problem is resolved. Small start, the method is correct, the little improvement can also be copied amplification. But if a lot of operation, do not know what useful those useless, it is completely unclear, the customer is anxious their own urgent useless.

Q: Testing the series, is not to put too many keywords and groupings?

A: You can effectively manage as a standard, if you are sure you can manage a large number of keywords at the same time, not afraid of more. But generally speaking, the more keywords, the higher the requirement of SEM management experience.

Http://semwatch.org/2011/03/large_sem_acct_management