The shortcut from a novice to an expert in SEM bidding, you will regret it if you don’t read it!

The growth path of SEMer is a process of fighting monsters and upgrading, constantly accepting challenges, and constantly entering and exiting the pit.

1. Say goodbye to SEM, practice only for a gorgeous turn
The growth path of SEM is a process of fighting monsters and upgrading. It requires constant practice, constant acceptance of challenges, and constant entry and exit of pits. In fact, it is only to bid farewell to SEM and achieve a gorgeous turnaround.

As a novice, you are just starting to get involved in paid channels and are gradually becoming familiar with various account operations. Your understanding of things with the same structure and similar meanings is constantly deepening. You have some experience in various targeting methods such as keywords, interests, ID number packages, remarketing, media, etc. You pay attention to some basic data, such as impressions, clicks, consumption, consultations, visits, transactions, click-through rates, average click prices, etc.

After a first glimpse at the basics, I had a rough idea of ​​what SEM is. It is about using money to buy traffic and achieve conversions. I started to have some clues about the work, but at the same time, I inevitably fell into the first pitfall of SEM: quality. Every day, I think about how to optimize the quality, how to reduce the average click price, and how to use the same amount of money to bring more traffic to the company. I believe that this will greatly improve the results, and I have no doubt about it.

As a rookie, I finally jumped out of the first pit. It’s up to me what to do. I don’t care about the quality score. They are all kidding me anyway. The more important the words, the lower the score, and the more useless the words, the higher the score. Haha, but I immediately fell into the second pit, a very deep pit: data analysis. I don't know where I heard about data tracking, URL tags, web page/event conversion goals, SDK embedding, etc., and I felt like a great guy. I finally figured out the ins and outs. I talked about data all the time, was full of energy in company meetings, and fought back against all those who challenged me. Everything was spoken with numbers, and I placed numbers on an unquestionable altar. At this time, you are so silly and cute.

The old bird had an epiphany and woke up from his obsession with data. Many times, epiphany only takes a moment. To truly become a qualified SEMer, I understood a very simple thing: I am engaged in online marketing, not just data analysis. Data is only one of the auxiliary tools in the marketing process and I cannot always stay in data analysis. At the same time, I also understood that the relationships between things are intricate, both direct and indirect. I wonder if the data analysis I have done is too absolute, and whether the attribution model (first/last/time/position) has been considered. In any case, whether it was a natural outcome or a fluke, I finally climbed out of this very deep pit.

As experts become more skilled, they transform from online promotion to online marketing. They start to think more about what can be done to help the final result. They don’t care too much about the gains and losses of a city or a pond, but focus more on the battleground: advertising creativity and pages. Advertising copy, material matching, special topic planning, copywriting and prototyping, design and front-end communication, these things start to fill you up, very full and very slowly. When you successfully create one topic page after another and begin to unify the style of the pages, forming a topic website composed of topic pages, you can finally take a breath. The work is finally done. All you need to do next is to conduct continuous A/B testing on the pages or local heat map analysis. At this point, you are already a top SEM expert.

You, who have been promoted to management and have never persuaded or dismissed your subordinates, tossed and turned, procrastinated again and again, and finally made a painful decision - I'm sorry, brother, you have to go. You are wasting your time and youth here. You might as well find a more suitable platform for development. It is not easy to be a manager. After this incident, my state of mind has been improved. I understand that managers are responsible for both the company and the employees. When conflicts arise, they still have to stand on the side of the company and cut the Gordian knot. Unsuitable employees should be persuaded to leave in a timely manner. It is good for them, you and the company.
It’s a gorgeous turn, basically saying goodbye to SEM, a very important stepping stone in your career development, and it has completed its mission. Your work content has also undergone a complete change, including department structure, job allocation, recruitment, training, assessment, retention, and daily trivialities. It is lonely at the top. From the heart-wrenching work, you gradually understand that finding the right person is more important than anything else!
It is not easy to make a gorgeous turn, and too many people cannot complete these steps. At the end of the article, I attach a famous line from Kong Shangren's "Peach Blossom Fan·Afterglow" to express some sadness at this moment: I watched him build a red building, I watched him entertain guests, and I watched his building collapse! I once slept elegantly on this pile of green moss and green tiles, and witnessed the rise and fall of the past fifty years.

2. Changes in Enterprises’ Understanding of SEM

Traffic is king. At this stage, most people believe that the purpose of SEM is to bring more effective traffic to the company, and the focus of work is mainly on digging for keywords. After Baidu promotion introduced quality, the focus was placed on grasping the quality. By optimizing the quality, CPC was reduced, and more effective traffic was brought to the company under the premise of the same budget.

Conversion is king. As CPC climbs, we begin to care less about traffic and focus on results. We understand that traffic diversion is not the goal, but only conversion. Our work focuses on organizing conversion words, and we monitor the words that can bring conversion to ensure their ranking.
Cost control: finding conversions is not necessarily a good thing, because the advertising investment may be too much and the conversion cost may not be reasonable in the end. At this stage, you will be obsessed with data analysis and start to study data tracking, attribution models, code deployment, etc. After realizing the power of data analysis, you may also go to extremes and regard data work as everything.

Process control has broken away from the shackles of promotion and reached the height of marketing. We have begun to understand the importance of teamwork. The conventional industry marketing process can generally be divided into: promotion and traffic generation - page browsing - online consultation - telephone invitation - door-to-door visit - transaction payment - word-of-mouth communication. Every link and every position in the entire marketing process is very important. At this time, SEM practitioners, in addition to data analysis, began to be interested in advertising copy, page design, consulting language, etc.

There are more and more marketing systems and channel management. Sometimes you will find that the marketing process is not stable enough, and there will be emergencies from time to time, which cannot be controlled well. I always feel that there is still something missing. When thinking about such problems, I gradually realized the importance of the marketing system. Through continuous practice, I finally realized that only a strong marketing system is the key to a company's SEM.

3. Is it a qualified SEM?

Account operation is the most basic part of SEM. The account mainly involves two parts: search advertising and information flow advertising. Whether you can skillfully design the account, build a high-quality structure, and produce advertising creativity, from the creative entry point, copywriting level, image processing, and even to the editing of advertising videos, after the account is built, whether you can implement a reasonable delivery strategy and make corresponding adjustments to different issues.

For novices, data analysis may seem difficult when faced with a bunch of complex data. In fact, it is not complicated after a little sorting. Data analysis is nothing more than segmentation, and then evaluating the quality of the segmentation, investing more in good areas and less in bad areas. In terms of thinking, we need to understand three main issues: dimensions, indicators, and methods, and the data part is basically solved. When you are able to use evaluation indicators such as cost and input-output ratio based on dimensions such as channels, devices, regions, time periods, plans, keywords, etc., and then master methods such as pivot tables, Countif, Vlookup, etc. to make specific Excel tables, you basically understand SEM data analysis.

Special topic planning, SEM consists of three parts: account operation, data analysis, and effect optimization. Account operation is the most basic thing, there is nothing much to say; data analysis, although it sounds advanced to laymen, is actually the simplest thing in SEM, because a novice with 3-5 days of experience can do exactly the same as an experienced person, just at a slower speed; effect optimization is nothing more than doing three things:

1. Basic adjustments to the account are still account operations, such as structure, creativity, detail settings, etc.;

2. Adjustments based on data analysis are to apply the conclusions of the analysis and spend money more efficiently;

3. It is an attempt at marketing planning, with the focus on special page planning.

If we sort out the above-mentioned points, we will find that what really determines the SEM differences between companies is not account operations or data analysis, but the special planning in effect optimization. The sad thing is that topic planning is a place that many SEMers have never been to, and many people are still working hard on accounts and data. Therefore, if you want to become a SEM expert, please focus more on the planning of special pages.

4. What to do if SEM effect is not good

This is a very boring question, but at the same time it is a problem that every SEM practitioner must solve. Next, from the perspective of the most common marketing funnel, I will try to fully summarize the factors that affect the effect, to help everyone sort out possible problems in your work (finding problems is generally more difficult than solving them). If there are omissions, you are welcome to add them. For example, taking the traditional industry as an example, the funnel can be divided into: display-click-arrival-browse-access-conversation-lead-invitation-visit-transaction-word of mouth.

Factors affecting display: industry popularity, advertising budget, number of channels, delivery area, promotion period, device, bid, and selection of targeting method (mainly word volume and matching in bidding, and reasonable coverage of targeting method in information flow)
Factors affecting clicks: creative quality (directly affects click-through rate), channel ad space (high-quality ad space has a more reasonable click-through rate, search is generally higher than information flow, information flow is generally higher than DSP/network alliance), competitiveness (in bidding, it mainly looks at the comprehensive ranking index = quality x bid; in information flow, it mainly looks at Ecpm = click-through rate x bid, which are essentially the same)

Factors affecting arrival: traffic quality (generally invalid/malicious traffic is not much, mainly because the delivery is not well controlled, the crowd is not accurate or the creativity is not clear), loading speed (server and web page itself), code deployment (the arrival volume counted at the bottom is definitely less)
Factors that affect browsing: page design (don’t make it ugly, try to make it comfortable), page content (what users want to see and what you want to show them, control the ratio), page link relationship (according to needs, even special pages can have mutual recommendations to encourage viewers to browse in depth)

Factors affecting access: Deployment of marketing probes (high-frequency deployment is fine, but user experience can be harmed. Don’t listen to the designer’s nonsense. Don’t forget that we are in marketing and our goal is to sell), customer service software settings (suspension, invitation, forced dialogue, different settings will result in very different access situations)
Factors that affect conversations: customer service software settings (interface decoration), automatic language and customer service's first three sentences (inducing visitors to start a conversation), customer service work attitude (even if the customer doesn't speak, you should say a few words to induce them)

Factors that influence leads: Customer service consulting techniques (ask questions while answering, quickly take the initiative, and ask for contact information), event design (events that are truly attractive)

Factors affecting invitations: attitude towards telephone invitations (frequency of telephone follow-up visits), telephone invitation language (attractive event design and invitation language level)

Factors that affect visits: weather (it is really difficult to visit during a typhoon), location (clearly inform the visit route), appropriate return visits (ask the customer to promise to visit multiple times)

Factors affecting transactions: comparison with peers (when peer brands are strong or activities are too crazy, the competitive pressure is indeed great), environment (don’t pursue luxury, but it has to be acceptable, at least clean), front desk reception (no need to be too enthusiastic, just moderate, a glass of water/tea/coffee is fine, no need to open a 1982 Lafite), consulting level (generally popular consulting sales, digging out pain points, magnifying needs, closed guidance, eliminating resistance points, the final key is to dare to help customers make decisions, don’t be embarrassed to ask for money), teamwork (when you can’t do it, replace someone, if it really doesn’t work, let him go, and clean up the mood to receive the next one)

Factors that affect word-of-mouth: product quality, service level, word-of-mouth marketing design (active requirements, incentives), online and offline brand packaging (most of the time, the product difference is not big, so positive guidance is needed)

If none of these meet the criteria, then you can only blame the market (peak and off-seasons, policy influences, emergencies, etc.). In fact, sometimes nothing has been done, it is just normal fluctuations, and it may recover the next day. Spend more time identifying the problem and then make adjustments.

Funnel analysis can help us find the problem, and the next step is to solve the problem. Generally, we can consider optimization from three levels.

Basic optimization mainly involves implementing the basic settings in the account to reduce unnecessary losses, such as structural optimization, click-through rate improvement, filter settings, etc.

Data analysis is a key to doing a good job in SEM. It sounds difficult, but in fact, the easiest part is to identify the good and bad advertisements, find the good ones and invest more in them, pick out the bad ones and invest less or even no money in them.

Marketing planning is the core and the most difficult part of SEM. It is mainly reflected in high-quality advertising creativity, high-quality special pages, successful marketing activities, strong consulting conversion, etc.

5. The way to choose SEM

First, understand the common SEM channels, and then consider the trade-offs.

There should be no need to explain the ads in the search engine results page (SERP). CCTV has exposed them many times, and the explanations are very thorough and the words used are very violent;

Information flow ads: ads inserted in the information flow. Many so-called information flow ads are not real information flows.

Streaming ads are very popular, so I followed the name to take advantage of the popularity;

DSP advertising: an upgraded version of affiliate advertising. It is generally said that affiliate advertising sells advertising space, while DSP sells targeted audiences. In fact, affiliate advertising is not easy to sell, so it is sold under a fancy name.

Advertisements based on website alliances or APP alliances have a low entry threshold to join the alliance and are difficult to manage, so there are many invalid clicks.

Taking Baidu's commercial products as an example, they mainly include brand zones, brand starting lines, brand Huabiao, commercial Aladdin, customized non-standard advertisements, etc.
Brand Zone and Commercial Aladdin basically start from one million, and often cost tens of millions or even hundreds of millions. They are played by wealthy companies. Brand Zone is generally a standard feature of large companies, while Commercial Aladdin is used by relatively few companies, and is mainly invested by large platform-based companies.

We can also do customized non-standard products worth hundreds of thousands, mainly to increase exposure during activities;

Brand Starting Line and Brand Huabiao can be made with a few thousand or tens of thousands of dollars, and are generally suitable for consideration by small and medium-sized enterprises.

Now that we know the common channels of SEM, how can we allocate the budget reasonably? Do companies usually focus on results first or brand first?

Taking direct conversion as the starting point, a positive cycle of corporate cash flow can be achieved. To exaggerate a little, it is a compound interest model. The money earned is used to continuously increase investment, and the snowball becomes bigger and bigger. Of course, such a compound interest model is mainly suitable for projects with relatively short conversion cycles and rapid turnover cycles.

The main purpose is to do brand communication, conduct continuous high-frequency bombardment on the minds of potential audiences, and strive to realize the corporate brand, and rank at the top or even first in the mental ranking of users. For example, when you mention instant noodles, you will immediately think of Master Kong; when you mention mineral water, you will immediately think of Nongfu Spring, which is a little sweet; when you mention safe cars, you will immediately think of Volvo, the Thor's Hammer.

You may think this is a very cool idea, as it can both build a brand and create results. In fact, the meaning of product and effect integration is that there is no effect and no brand. To put it more nicely, it means there is some effect and some brand. Confident channels call themselves performance channels, while less confident ones call themselves brand-effect integrated channels. Brand channel is a term used exclusively when encountering customers with brand demands. Generally speaking, no platform will specifically say that its channels have no effect and can only boost the brand.

SEM channels are mainly oriented towards performance marketing, so the choice of channels is simple. Generally speaking, priority is given to bidding for performance marketing, followed by information flow that combines product and effect, and finally DSP and affiliate advertising for brand marketing (information flow in a few industries is better than bidding, but generally speaking, they are placed in the above order).

The specific budget allocation is made according to the effectiveness of the channel. Assuming we take the order volume as an example, we will use the order cost to evaluate the effect. Channels with low costs will increase investment, while channels with high costs will reduce investment. As the budgets of channels with good results increase, the accuracy starts to decrease, and the results will gradually decrease; similarly, as the budgets of channels with poor results decrease, the accuracy starts to increase, and the results will gradually improve.

The optimal configuration is to achieve basically the same cost indicators for each channel by continuously adjusting the budget of each channel (the attribution issue will not be discussed in this article). At this time, the budget allocation is the most efficient. When it is fair, it is inefficient; when it is efficient, it is unfair.

Every channel has its own value. We can only take one from the vast ocean. We just need to identify the part that suits us and allocate the budget reasonably. So in a sense, there is no good or bad channel, only big or small.