The road to "customized marketing" for online education in 2020

The road to "customized marketing" for online education in 2020

The battle for customer acquisition in online education is becoming increasingly fierce. With the support of capital, the battle for customer acquisition in the education industry is becoming more intense. However, we should not forget that the essence of capital blessing is to exchange capital for time. In short-term battles, advantages come quickly but also go away quickly. Whether you can win in the long run depends on whether you have a mechanism to control customer acquisition costs and advanced marketing capabilities.

Taking the recently popular GSX as an example, its weighted average customer acquisition cost in 2019 was around RMB 470, which is significantly lower than other competitors in the same period, but still increased by 113% compared with 2018. The continued rise in customer acquisition costs has led to some important innovations in customer acquisition methods. (Data from Caixin.com)

From the perspective of customer acquisition, marketing has experienced three major turning points: the creation of product promotion catalogs, the popularization of online searches, and the rise of social network marketing. Every turning point is a step towards more accurately conveying information/supply to those in need.

The fourth turning point that is happening recently is no exception - customized marketing.

We can all feel that social network marketing has been widely used, but its effect has gradually weakened. The way to produce and push information on social networks is becoming more and more convenient. A large number of public accounts, video self-media, and online and offline circles of contacts have led to customers facing a large amount of marketing information at all times, thus returning to the starting point of being unable to obtain information efficiently. The marketing effect will naturally gradually deteriorate.

The emergence of customized marketing can solve this dilemma well. That is, on the basis of mass production, enterprises regard each customer as a separate market segment and make a marketing mix based on the individual's unique needs to meet the needs of each customer. At this point, "how to provide different marketing services based on the personal preferences of different potential customers" has become the core issue of customized marketing.

The main value of data to customized marketing is to efficiently find accurate target customers. Here we need to lay the foundation for a concept - the data patterns in traditional CRM marketing and digital marketing are completely different, and customized marketing is a marketing model that flexibly combines the two for my own use.

As shown in the figure above, in traditional CRM marketing, the core of data application is customer data. In the entire closed loop, it revolves around the main line of "collecting customer data → finding target customer data from a large amount of data → targeted marketing → re-collecting new customer data generated in marketing."

In digital marketing, the core of data application is the algorithm for identifying customers. The main line is "setting up marketing based on the possible behavior of target customers → collecting customer behavior data → triggering corresponding marketing activities for target customers through matching → improving the trigger condition algorithm for target customers based on marketing results."

There is a vivid metaphor:

The traditional CRM marketing strategy is like an artillery piece. After discovering the target customers, it will fire a bunch of marketing content at them, hoping to hit them. Digital marketing is more like a "mine soldier". It finds places where target customers often appear, buries triggering marketing content, and just waits for target customers to appear to trigger marketing activities.

Customized marketing is like a smart command center that can flexibly command artillery and mine soldiers, formulate combat plans, and maximize the value of existing data.

Take the new customer conversion path of an online education brand as an example:

From advertising - retaining information - trial classes - telemarketing, to final purchase, the data shows that the brand's conversion rate failed at the telemarketing step.

After analysis, it was found that a very important reason was that marketers could not make estimates and judgments in advance about the customers who left their information, could not distinguish the quality of leads during telephone sales, and did not have enough information to make personalized conversions.

After discovering the problem, the brand combined digital marketing with traditional CRM to develop a customized marketing strategy:

First, the brand conducts correlation analysis and learning based on the data of historical successful/failed customer conversions, and uses AI to establish a relationship model between "successfully converted customers" and "customer characteristic parameters." This model can stratify customers according to the behavior and attributes of new visitors (for example, stratify them according to the probability of possible conversion), so that a more personalized landing page display can be made at landing, improving the customer experience. In the subsequent telemarketing stage, a more reasonable follow-up strategy can be determined through data screening by the CRM system and newly added customer stratification.

In addition, there is another very important value in utilizing existing data assets, which is to balance the final customer acquisition cost. We all understand that extremely low acquisition costs will introduce low-quality traffic, consume a lot of customer service manpower costs for communication and screening, waste a lot of sales time for conversion, and waste the resources of trial class teachers. Therefore, if brands can utilize historical + real-time data and intervene at the front end of marketing, they can better control conversion costs.

Case: A K12 institution’s “traffic stratification + cost stratification” marketing

The landing page of a well-known K12 brand will direct new customers to the trial class page. When customers come in, they need to fill in their mobile phone number, grade, subject and name. This certainly cannot achieve traffic stratification. So how does this brand stratify?

It is through a clever operation setting - after filling out the trial class form, a pop-up window will pop up, prompting you to follow the official account QR code, so that you can give priority to the top-rated teachers for trial lectures (and you need to make an appointment with the teacher through the official account). Through such step settings, users are divided into four categories according to their behaviors:

  • Category D: Only click the first step to submit but do not complete the verification;
  • Category C: Verification completed but no public account binding;
  • Category B: following the public account but not making appointments on their own;
  • Category A: Completed the self-scheduled classes.

The brand's marketers stratified costs for the above four types of users and adopted different marketing methods:

  • Category D: The lowest intention, using the lowest-cost AI customer service robot to conduct a carpet scan, which can be done through automatic text messaging and phone calls.
  • Category C: They have intentions, but their initiative is average. They will be screened manually by TMK before being sold to brands.
  • Category B: The intention is relatively high, and it is followed up with general brand sales.
  • Category A: Assigned to top sales, focused on follow-up.

Such a simple form process divides customers into four levels and then matches resources based on conversion possibility. After a series of operations, the comprehensive cost of traffic (including backend costs) dropped by nearly 15%.

Marketing is achieved through a series of links. If the entire enterprise has such a marketing idea, then every link now will definitely have room for further optimization. If you can use your imagination to acquire customers, data assets can do a lot in customized marketing.

Today, community operations in online education have become very common, mainly through live broadcasts, open classes, paid courses and other sources, to guide customers to the brand's private domain operations. The personalized community involved in this article goes one step further than mainstream conventional communities. Its core is to establish trust between customers and brands, and between customers themselves, so that the community can extend the customer life cycle and trigger a word-of-mouth effect.

Let’s first look at the success factors of the current mainstream communities:

  • Functionality: There are more than one operator and one KOL (such as a course instructor) in the group, who will organize group tasks, share information and answer questions.
  • Activity: When the community has a certain number of members and about 15% of the members are willing to actively share information and resources, this is a basic condition for the community to have commercial conversion potential.
  • Orderliness: There are clear rules within the community, such as how to introduce yourself after joining, how to prevent advertisements unrelated to the group, entry barriers and exit mechanisms, etc.

Personalized communities also need two more elements:

  • Strong connectivity: Members join the community because of common learning or a certain pain point. Operations should take advantage of this commonality to guide more social behaviors among members (whether it is online and offline communication, resource integration and exchange, knowledge sharing, etc.). Only in this way can an atmosphere of safety and trust be established, making brand conversion more natural. There is another factor that is easily overlooked here - trust and interaction among members can amplify loyalty and purchasing power.
  • Personalized operation: a marketing plan tailored to the characteristics of each community, rather than a one-size-fits-all mainstream plan. Personalized communities do not value a one-time conversion, but continuous repurchase; personalized communities do not value a certain product, but the purchase of more different types of products based on brand trust. Therefore, personalized communities are more flexible and growable rather than being a simple "class group".

It should be noted here that no matter what kind of gameplay today's community adopts, the final step is to realize its commercial value (commercial value can be either product sales or word-of-mouth communication). The principle of personalized communities is to enhance this commercial value through "strong relationships + trust chains".

Case Study: A children's programming education brand achieved high stickiness and conversion through "small communities"

A certain children's programming brand is well aware that only a high renewal rate can guarantee stable cash flow and profits. Therefore, the operation staff will make a relatively detailed user portrait of each family (children + parents), and then group them according to the child's age, programming foundation, and actual geographical location. Some people may not understand why online teaching should prioritize geographical location grouping? In fact, this is to strengthen the social attributes of small class communities.

The brand’s class division system in a certain first-tier city explicitly includes residential location as a factor. Small classes have fixed time periods and fixed classmates. The teacher will set up groups for children and parents respectively. Each group will have at least three staff members involved: lecturers, tutors, and class teachers with service and sales natures.

First of all, there is a sufficient warm-up period before establishing the group. The class teacher will spend 2-3 times organizing several families to introduce themselves and get to know each other.

Next, because of the geographical advantage, after several families become connected through the course, they will most likely become families that have offline interactions with each other, that is, children become friends and parents become objects of mutual sharing and comparison. This stickiness will cause the preferences of this small group to be bound together - in addition to renewing their classes, they will also jointly purchase other services of the brand, such as study tours, summer camps, etc. The essence of this small community becomes a growth group for several children, and the attrition rate will be greatly reduced.

In addition, the community’s activity design also stimulates word-of-mouth communication among parents. Through various small competitions, the winning parents are inspired to feel proud and forward the message. And because parents’ circle of friends has its own precise traffic, such dissemination often leads to a good recommendation effect.

Of course, this model also has an obvious weakness: once customers churn, it may be a complete churn of a small group. This requires the brand's customer operation system to be more agile and able to conduct timely public opinion analysis.

The brand in the case has encountered this situation before. Through public opinion tools, they discovered that the attrition rate in a community with a common feature has recently begun to increase: small course groups for 8-9 year old children.

The operations staff observed and conducted in-depth interviews with community members and found that, in the minds of parents, "third grade" is an important psychological turning point - even many Buddhist parents begin to become more utilitarian when their children are in the third grade, and pay more attention to whether online courses can improve their academic performance. But the problem is that the programming series courses are more inclined towards thinking training and quality improvement. This is the main demand for parents of students below grade three. Once they are past grade three, many parents will communicate with each other privately to reduce the proportion of extracurricular classes. Then even loyal customers with a good experience may be lost.

After discovering this, the brand mobilized resources from multiple departments including lecturers, operations, and sales to jointly develop a response strategy for this part of high-churn users - starting with a small-scale trial, through parent KOLs in the community, emphasizing how programming can help children's science grades, and reducing the burden on children by adjusting the course settings and reducing after-school exercises. Finally, we will provide greater discounts to several core families in the small community to encourage them to renew their registration and influence other parents in the group. After this marketing method achieved good results, it was applied on a large scale.

This practical case also reminds us that the idea that "social marketing is a marketing method that can save manpower" is "very stupid and naive." Community is a typical labor-intensive and high-intensity job. Behind a good community, there must be a lot of people working and resources invested. In addition to daily operations and promotions, we must also combine product features, continuously maintain insight into customers, and guide them in a direction that is beneficial to the brand in the long run.

Of course, this process can now be combined with many types of operations and analysis tools to make community building, maintenance, and tracking more efficient . But it also poses higher challenges to operators, requiring more marketing imagination to meet customers' demand for personalization.

Customized marketing is to provide different product promotions and services based on the personalized needs of customers, but customer needs vary greatly. Some brands believe that adopting customized marketing based on different customer needs will inevitably significantly increase operating costs and even affect overall profits.

Here are three ways to dispel this misunderstanding:

First of all, customized marketing is to better respond to the growing customer demand for personalization. These include theory, technology and execution strategy, which are the innovative upgrades that companies make to product marketing after changes in the market environment. Cost investment is only one step, and whether they can benefit from it depends on their actual application capabilities, and this is also a challenge that companies must face.

Secondly, in terms of cost, from the perspective of the entire product supply chain, customized marketing does require more manpower and resources in the early stages; but once the mechanism is on the right track, the cost of service quality and resource allocation will be reduced a lot, and the brand can match customer needs more accurately. And when the market environment changes, marketing can respond flexibly through micro-adjustments and micro-innovations, avoiding drastic and high-cost adjustments.

Finally, customized marketing does not have the wide coverage of large-scale marketing, but the marketing experience is significantly better than in the past and will improve the discovery and conversion of effective customers (such as higher premium ability).

Indeed, given current technology, it is unrealistic to achieve customized marketing that satisfies all customers in a one-size-fits-all manner. However , the 80/20 rule of corporate statistical analysis also proves that 80% of the profits of most companies are often generated by 20% of high-quality customers.

Therefore, one value of customized marketing is to discover and screen high-quality customers as quickly as possible so as to tilt resources and services toward high-value customers. At the same time, operations with high-quality customers can also increase customer stickiness, form word-of-mouth marketing among high-quality customers, bring in equally high-quality customers, and improve the conversion efficiency of high-quality customers.

Therefore, customized marketing is not a process of perfectly solving the diverse needs of thousands of customers, but a process of more accurately positioning different categories of customers, discovering high-quality customers based on data insights, and further improving the conversion rate of these high-quality customers through personalized services.

There is no doubt that the core competitiveness of an educational brand is its teaching ability, but the conventional understanding of teaching software and hard teaching staff is not all of today's educational products.

Educational services have the particularity of teaching students in accordance with their aptitude. For every customer, the more advanced the course, the better. Instead, the best product is the one that suits him or her. Therefore, the process of creating core products should maintain sufficient flexibility and customization, which requires the brand to continuously build its capabilities in marketing operations. That is, the marketing process is also the process of cooperating to complete product creation.

From the perspective of service quality, marketing operations are definitely part of the product, and its value in educational products will become higher as students' demand for personalization increases.

Finally , based on our Lingxi’s insights into customers in the education industry, the service chain of online education is very long. If we can improve 1-2% in each key link of marketing, the cumulative improvement will create a very high competitive threshold . I would like to dedicate this article to those in the education industry who are struggling in this capital carnival. I hope they will have the courage to face challenges while keeping a long-term vision.

Author: Agnes

Source: MartechtoolLingXi

<<:  “Soul” product operation analysis!

>>:  What are the functions of Lanzhou Pet Mini Program? How to develop a pet store mini program?

Recommend

Tool Inventory | One-click query of information flow advertising materials

Are you still using Baidu to search for informati...

How do algorithms influence user decisions?

In the past two days, articles about Internet tec...

The secret of a hit product: creating influence

Creating product influence is not a one-day job. ...

3 elements of APP operation and promotion strategy!

All resources have a focusing effect. For product...

Amazon Advertising Bidding Mechanism & Ranking Principles In-depth Analysis

When placing advertisements, Amazon sellers first...

How to build user portraits for online education!

The implementation of the data-driven operation c...

How to correctly place ads on TikTok?

1 Behind the TikTok phenomenon As of mid-June 201...

How to like Douyin? Let us teach you how to get likes on Douyin!

Anyone who has used Tik Tok knows that if you wan...

The 10 most common questions for beginners learning SEO

Before reading this article, novice SEOs can revi...

Zhihu operation full set of detailed operation guide

The full text is about 5,000 words, and it takes ...

Xiaohongshu promotion method: How to promote effectively?

When you have a fan "offline", and he s...

How to increase the number of APP downloads from 0 to 3 million?

In April of this year, I was looking for product ...

Can the female anchors who earn millions a year support the future of Momo?

At the end of October, this girl was not too cold...