SaaS Advertising Tips

Marketing genius John Wanamaker has a famous saying in the industry: I know that at least half of my advertising expenses are wasted, but I don’t know where exactly they are wasted? As the epidemic continues to recur, most companies have budget cuts, and how to acquire the most potential customers with the least money has become a top priority for marketing departments.

Advertising on the B-end is different from that on the C-end. C-end products cover a large population and user base, and can achieve good output by taking advantage of hot topics. However, B-end products cover a small population and have a long transaction cycle. It is necessary to find key issues through data construction, refined data operations and market dataization in order to optimize marketing strategies.

This article is compiled from Cui Niu Time, and is about "Sharing Experience in Paid Advertising" brought by Xu Wenlong. It mainly breaks down how to do a good job in paid advertising. The specific content is how to do a good job in positioning, account building in four aspects: algorithms, operating skills, user size, and resources, as well as thinking on the data model of advertising.

The core content of the full text:

1. How to do a good job of targeting

2. Account building experience sharing

3. Launch a data-based model

01 How to do a good job of targeting

Before getting into the main topic, let me share with you an interesting phenomenon. These are the promotional pages of several refrigerators. Their functions are similar, but they are all labeled with "mother and baby" scenes instead of "beer" scenes.

Please think about two questions first:

1. There are so many scenes to choose from, why do you have to choose a mother-and-child scene instead of a man watching a football game and drinking beer? Out of curiosity, I actually searched it and found that no one really went to get beer.

2. Are the customer portraits for the mother-and-child scenario and the customer portraits for drinking beer and watching football games consistent in terms of gender, age, interests and hobbies?

The results are obvious. The conversion rate of targeting mother-and-child scenarios is higher, and the customer profile of this scenario is different from the customer profile of watching football games and drinking beer.

Back to the topic. How to do market positioning for advertising?

First I need to introduce a theory: Tasks to be Done, proposed by Christensen. I have extracted 5 marketing related principles from the main principles.

1. People buy products and services to get “jobs” done.

2. The tasks are functional and have emotional and social components. The essential difference between To B delivery and To C delivery is that users and customers are separate and they are usually not the same person.

When we do To C advertising, we usually just need to solve a user's moment, create a conflict for the user, and impress him with the emotional part, which may lead to a purchase. However, when doing To B advertising, there must be both a rational part to satisfy users' business needs and an emotional part to satisfy their emotional needs.

3. The backlog of tasks is stable over time.

4. To-do tasks are not related to the solution. For example, to get from Beijing to Shanghai, people would have chosen horse-drawn carriages or walked 2,000 years ago, but today we can choose to travel by train, plane, or drive ourselves.

5. Success is analyzing tasks, not products or customers as the unit of analysis. It can be said that this is the most useful principle in marketing.

So, how do you identify your customers’ to-do tasks during user research? After multiple user surveys, we summarized four major categories of problems and divided them into 28 small problems (as shown in the figure). Within the scope of the customer's wishes, ask these questions as detailed as possible by designing the script.

After these questions are asked, they need to be analyzed from the following dimensions to clarify the customer's tasks to be done.

1. Difficult moments. Interviewees bring up events or contexts that led them to struggle and seek new solutions. For example, users are working hard to increase the amount of leads they acquire each month, but they won’t advertise on TikTok.

2. Motivation. The interviewee wanted to use a new method to accomplish this task. For example, users want to build a sales funnel to obtain leads and automate the process of lead generation.

3. Driving force. It is the emotional force that causes users to seek new solutions and no longer work the same way. For example, a user is now frustrated because they have not met their lead acquisition goals in the past few months.

4. Perceived value. How the interviewee envisions the current situation becoming better once a new solution is available.

5. The value of experience. How the interviewees experience the product.

6. Solutions considered.

7. Functional tasks. The core tasks that the interviewee hopes to accomplish.

8. Emotional tasks. What the interviewee wants to feel or avoid while performing the core task.

9. Social tasks. How the interviewee would like others to view them.

After we analyze the problems in these 9 dimensions, we can help you output results in the delivery process, whether it is from the dimensions of video content, landing page, creative title, keywords, etc.

02 Account Setup

Regarding account building, I divided it into 4 quadrants:

The first quadrant is algorithms. On the one hand, there are the platform’s requirements for content, and on the other hand, there are the data that should be paid attention to from the algorithm level.

The second quadrant is operational skills. On one hand, it is about writing creative titles, and on the other hand, it is about getting accurate sales leads by writing tags.

The third quadrant is customer volume. Including user coverage and channel selection.

The fourth quadrant is resources. Including channel rebates.

algorithm

Currently, all information flow platforms require users to provide high-quality content. So what are the criteria for judging high-quality content?

1. Content

The video content of an account should revolve around one theme, not multiple themes or multiple product lines. The content can be presented by changing people or scenes.

(1) The video content should be based on the hot topics of the platform, such as videos of beautiful women and luxury cars that are easy to attract attention. I have done a test before. For the same oral broadcast, the click-through rate can differ by about 1 percentage point when read by a beautiful woman and a middle-aged man.

(2) The video should use real scenes as much as possible, highlight the theme in the first 3 seconds, have clear voice, and the picture should not be blurry. Usually, a video material with blurry images or closed spaces will not be judged as high-quality material by the system.

2. Data

During the delivery process, although the official platforms such as Douyin have never stated that there is an account scoring mechanism, this is not the case. Both Douyin engineers and the "dou+ assistant" live broadcast said that the system will give a score to each account. To keep your account at a high score, you need to pay attention to the following data:

(1) The playback rate of the first 3 seconds. The playback rate of many advertisements in the first 3 seconds is as low as less than 3%, and the advertising costs remain high.

(2) Completion rate.

(3) Click-through rate.

(4) Number of shares and likes.

(5) Conversion rate.

The data for different industries are different and there is no clear standard. The data is greatly affected by the average order value. If the average order value is high, the data will be low, and if the average order value is free, the data will be very high. This cannot be compared. What is important is to pay more attention, try to optimize the data and ensure the stability of the account.

3. Account Budget

Account budget is an important factor affecting account quality score. If you normally only run 20 leads, but suddenly run 50 today, you will find that the quality of leads obtained by your account has dropped significantly in the next few days. In order not to affect the quality and rating of the account, do not adjust the budget too much and keep it as stable as possible.

Operation Tips

Operational skills are important factors in increasing exposure, acquiring accurate customers, and finding high-quality customers. It not only involves targeting based on product features and target groups, but also things that are overlooked, such as the writing of creative titles and tags.

1. Creative title writing

There is a big difference between C-end and B-end delivery. On the C-end, people tend to take advantage of hot topics and use XX activities to attract customers. For example, "XX wedding photography platform offers free limited-time gift" and "Shocking price XX". This type of title can give customers a sense of satisfaction from making a quick purchase. But this type of operation is completely unsuitable for the B-side and will only bring in some people who want to take advantage of the situation. When many "freeloaders" leave money in your advertising account, the system will think that this group of people meets your standards and needs, and will constantly change the positioning of the system's population model.

I have summarized four creative title writing methods that can increase the display volume and help the system locate the precise audience. They are suitable for all channels such as Baidu, Toutiao, and Tencent. The formula is as follows:

(1) XX customer + product name + achievement

Example: The marketing system used by Baidu Auto obtains 300 leads a day.

(2) Product name + achievement

Example: The Blue Sky inspection system is easy to use without training.

(3) User group + product name + results

Example: A restaurant store needs this store inspection system, and one supervisor can easily manage 10 stores

What needs to be noted about the first three formulas is that when describing the results, it is best to describe a phenomenon rather than a conclusion. For example, some customers have their own traffic and are your typical customers. So at the moment of communication, it is a good empowerment to let them introduce their brand into your advertising.

(4) Location + User Group + Product

Example: Restaurant owners in Chengdu are all using this financial system.

You know, on the one hand, locality can make advertising more targeted, and on the other hand, it can enrich the title model.

2. Label formula

Tags are a targeting method unique to the TikTok channel. Creative classification and creative label are the advertisement’s self-introduction to the system, helping the system understand the product’s functions and thus helping you find the right person; the idea behind selecting creative classification is “the more accurate the better”, choose the category that the advertisement belongs to; the idea behind creative labeling is “the more accurate the better, the more detailed the better, and the more the better”.

The tag formula for Douyin placement in the To B industry is as follows:

(1) Industry + target group + functional description

(2) Competitive product + functional description + achievement description

(3) Target group + well-known brands in the customer industry + functional description

These three combinations are my experience when working on the B-side. On this issue, I would like to share with you the differences between writing labels on the C-end and the B-end.

C-end products cover a large population and user base. When launching products, the tag strategy adopted is to add hot events or people related to the product who have their own traffic. The search frequency and event reach frequency of this type of character keywords are high, and the user base itself has a high match with the viewers, so it can produce good output most of the time.

However, B-end products have clear industries and functions, and a small user base, and are not suitable for C-end label targeting methods.

User volume

1. Number of users covered

Oftentimes, there is a problem when we place advertisements: they are not displayed within a reasonable market bidding range. 80% of the reasons for this situation are small user population coverage. When the user base is larger than 10 million, there will be good performance; when the user base is smaller than 10 million, even if there are a lot of outputs, it will be difficult to show.

For example, a friend of mine has a product with a unit price of around 500,000 yuan. In order to save sales manpower, he is very strict in targeting. When placing advertisements, the area is targeted to a specific geographical location such as Xidan Financial Street in Xicheng District, Beijing, and the price of mobile phones is positioned at more than 10,000 yuan. But there was no consumption in the account. Is it a channel problem? We investigated the data of peer competitors on Douyin and found that the average daily number of accounts was over 50,000.

My suggestion is: appropriately relax the targeting to expand the user base, and try to cover more than 1000W people with the planned advertising. The targeting should be supplemented by creative titles and label writing.

2. Channel selection

For example, the user groups of maternal and infant products and pet products are mainly concentrated on vertical platforms such as Meiyu Maternal and Infant and Mayu Pet. Previously, Lu Yang wrote in "Short Video Marketing Campaign: How Can To B Enterprises Break Through? 》I have shared the platform features in the previous article, so I will not repeat them here.

resource

The resource section mainly talks about channel rebates.

If you have long-term plans for investment, I hope you can maintain good relationships with channels, which will bring invisible benefits, such as market price data, competitor data, and even free resources. For example, Tencent gave me several free MBA advertisements.

What is more important is the rebate. The actual recharged money is different from the money reflected in the account. Each channel will provide channel rebates to users. In 2018 and 2019, the Douyin channel could still offer a rebate of 37% to 40%. Starting from 2020, the policy of the Douyin channel has gradually decreased.

At present, there are still great differences in the policies given by agents in some small channels, such as 360, Sogou, UC, etc. On the one hand, it depends on the level of agents they are, and on the other hand, it depends on the relationship with your customers. It is recommended that you make more comparisons and find a channel that can provide suitable agency policies.

03 Launching a data-based model

The most difficult thing to adapt to when switching from C-end to B-end is the long transaction cycle on the B-end. At the beginning, there was no transaction volume and lack of data support, making it difficult to confirm whether it was a problem with the delivery or other problems, and there was no way to find the direction for optimization. Later, through statistical data models, I found that my biggest problem was in the stage of awaiting customer transactions. Therefore, data construction, refined data operation and market datafication are crucial.

At the beginning, the data we focused on included data volume, data cost, effective quantity, and transaction rate. In terms of customer profiling, the marketing department’s responsibility is to obtain enough data to meet the MQL standards, focusing on efficiency rather than transaction rate. Later, we will use the data-based model to count the sales milestones, and when each stage is presented, we will find problems and make optimizations. The dimensions include: matching of portraits, results, solutions, as well as customers to be closed, closed customers and additional purchasing customers.

Customer profile: Customers who meet MQL’s standards, such as whether they are from the finance department, human resources department or the boss, whether the company has 100 employees, whether there are technical personnel, whether there are more than 3 stores, whether the revenue reaches 10 million, etc.

In terms of operations, the data conversion rate at this stage is low. You can check the content in the advertising promotion to see if there are any positioning errors, or start from the private domain traffic operation and perform refined operations to improve the portrait.

Outcome match: whether the customer’s problem matches the description you made in the promotion. For example, there may be deviations when communicating tasks and they may not be executed properly. When the outcome matching conversion rate is low, the outcome description of the implementation task needs to be adjusted.

Solution matching: The solution recognized by the customer can solve the customer’s current problem.

When serving B-side clients, we all have scenarios we want to target, such as matching refrigerators with maternal and child scenarios. We hope that all customers come in based on maternal and child needs. Although some customers may buy because they drink beer, these two aspects do not conflict. When we make a choice, we should judge whether our choice point is suitable for large scale and large scene.

In the current increasingly fierce market environment, the cost and difficulty of obtaining leads are increasing. In addition to expanding more channels for obtaining leads, we must also optimize and improve the conversion rate at each stage internally to ensure production. Currently, most companies are just trying to find ways to obtain clues, but they have a weak awareness of market dataization, which is also one of the key factors affecting investment.