推薦連結 : 實用英語課程
New Big Data Marketing 4P, real-time status and predict consumer dynamic, zero-day, personalized marketing zero error, a person is a niche market, marketing hit rate of 100%.
When combined with big data marketing ── big data marketing, will become the most revolutionary marketing trend, big data marketing may even subvert pursue marketing theory 4P nearly half a century: Product (product), price (price), promotion (promotion ), passage (place). Great marketing will generate data under a new 4P: people (people), performance (performance), the step (process) and prediction (prediction).
First proposed a new theory of 4P Kimberly Collins is the world's most authoritative IT research and consulting firm Gartner vice president ─ Company (Gartner Research), and we will be the last P (profit, profit) is corrected to predict (prediction) .
From the old to the new 4P 4P, big data marketing exactly how to subvert the traditional marketing?
First, companies should be from the past, "the commodities" thinking, turning human-centered "business customer", and big data era, is to provide the best opportunity to view transformation. Big data allows "one to one marketing", "personal marketing" is no longer a fantasy, but basic services.
Every time a consumer purchases, buying cycle, buying properties are not the same, but the traditional marketing, you can not do very detailed personal marketing, mostly mass marketing or community-based marketing. Such as traditional marketing based on consumer needs and differences, will the market segments for each of the different ethnic groups. For example a sports shoe brands have seven market segments, then with seven marketing activities, they often have to spend a lot of manpower and resources.
But the era of big data marketing, market segmentation may be 70 000, the corresponding marketing activities have 70,000, while the implementation of the system or platform, if the same as traditional marketing can only rely on manpower to carry out these activities in 70000, - an almost impossible task.
Everyone's buying behavior and consumption habits are not the same, but because the traditional marketing and human resource constraints, therefore, is often the only person to go along with marketing, rather than marketing to meet people; when companies can only perform 7 marketing campaign , naturally only people looking for activities, rather than the difference of the activity to find someone (people to campaign). But when the marketing campaign has the ability to become as many as 70,000, every consumer can be paired event from 70 000 to his activities in the most appropriate, thus flipping the old marketing point of view, turn out to be people looking for activities, rather than activities Find someone.
Thus, from the crowd to individual marketing activities, market segmentation is getting smaller, or that each customer has become a niche market, personalized marketing emerged.
1st P: people
NES model, the more points only five kinds of customer problems
In the era of big data, human-centered, consumer two biggest feature is the presence of heterogeneity and variability, variability among which the most difficult to master.
The figure below illustrates
For customers, we have constructed a model NES:
N = new Customer (New Customer)
E = Existing customer (Existing Customer)
1.E0 main customer: Personal time buying cycle twice repo man
2.S1 sleepy audience: more than 2 times the buying cycle is not personal repo man
3.S2 half asleep Customer: more than 2.5 times the buying cycle is not personal repo man
S3 = Customer sleeping (Sleeping Customer), purchase frequency than personal buying cycle three times did not repurchase, the repo rate is less than 10%.
NES model that is to immediately grasp the changes of customers and design, this 3 labels and 5 groups, based solely on actual consumer transactions calculus, and the ability to dynamically update with information corrected.
The first big marketing data is a P "consumers." NES consumer model can be divided into N (new customers) via real-time calculations, E0 (main customer), S1 (sleepy customer), S2 (half asleep customer) and S3 (sleep customers) five kinds of labels, as customers of sleeping deeper and deeper, the brand the opportunity to effectively lower the wake, wake relative costs will increase dramatically.
Marketing people past the marketing budget and work time, according to the grouping customer natural properties are allocated, habitual from customers past the cumulative contribution of consumption and previous transaction history, combined with sex and spending power index of customers to determine their label set and mode of operation; but it is clear in these data behind, we underestimated the time to influence consumer motivation interference, through the concept may be just an average figure of "about" defines the repurchase of no more than 180 days customer is the so-called sleeping customers.
But in fact, when we model the data transmitted through the NES calculus we found that many customers as early as 120 days or so into the S3 (sleep customers) stage, time is defined not only allow enterprises to brand non-discriminatory missed a key opportunity to wake up, and then in the low S3 stage did not get to wake up the rate of remediation, these seemingly trivial details, are a waste of valuable corporate resources and costs.
Immediately grasp the actual status of each consumer is the big data marketing the most important first step, if we have the ability to tailor each customer a dedicated communication point in time, it's time to put out fishing net marketing to close up.
2nd P: Performance
Each shop can do their own marketing
Big Data Marketing second P is "Performance", "profit" is the common goal of business, profits affect many factors, the truth is the revenue impact of three objectives: to increase the number of customers, customer price increase, active lift.
If the front of people talking about the state of the customer, then the customer is talking about dynamic performance.
The flow of customers is reflected in the store-oriented point of view, it is often observed that when companies view profit revenue, we found that the number of visitors fell directly determines the number of new customers slack, resulting in revenue decline is the culprit, so immediately decided to drop under million budget, arranged to store gift activities to enhance the recruitment of new customers, in one fell swoop can expect increasing sales momentum, boosting performance.
As a result, the number of new customers does significantly increased, but revenue was still no improvement.
The original data hiding behind the real reasons for the decline in revenue caused by the store, in fact, high customer loyalty and contribution of a large number of rapid loss of new customers to enhance the activity of small revenue assistance, priority should be to first find out the reasons for the loss of customer loyalty, develop programs to restore customer, seek first break before going to complement and reinforce strokes off.
Wrong data will be the wrong meaning, not only will the store consumes unnecessary marketing budget and time in vain to solve a problem, and further to the competitors can take advantage of the machine, between the one to come, victory or defeat sentence, to accidentally Down?
In other words, KPI every store operations should be personalized setting, for example, in Taipei to do when the new customer, Kaohsiung should probably do the loss of customers, Taipei No. 1 store in new customers do when the store might Taipei 2 should do the loss of customers, the end to see data for each store to decide what it should give priority to the most improved Yes.
Chapter 3 P: Step
Identify priority, priority issues at stake
Big Data Marketing P is the third "step" through a hierarchical implementation of the Heart, to improve revenue equation. When revenue three variable problems, should take what kind of strategies to solve the problem.
When the store found that revenue decline, view items sales, the number of visitors and customer price and other data, assuming that the number of customers is not enough, should think of ways to add new customers, or to find ways to retain old customers. If the customer is not enough activity, the question of loyalty, it can be purchased for an early customer, or to increase regularly care for existing customers, and warn the time of purchase, and other various actions to enhance the activity of customers.
If the problem is in the customer's customer price is insufficient, further to review, what is not enough new customers or old customers is not enough. If the new customer's customer price is not enough, usually because of a variety of promotions for new customers, while new customers come in with a very favorable price, usually come close are unhealthy members, then he probably is no way to continue to contribute to the value of .
The new problem is presented by the customer's customer price is too low, while the rate of new customer repurchase is too low, a new customer conversion rate is too low, the main is to do a promotional improper way.
4th P: Prediction
Accurately forecast customer repurchase next time
If the process to tell people from marketing revenue flow equations and members can be controlled, and that the last part of big data marketing "project", declared that such control can be intelligent monitoring and execution.
Customers like water, water to flow from the first purchase of new friends, the lopsided loss of customer, the process is the norm. But then with a big data marketing but we can make an early response programs. Analyze future data. From Analysis to predict, calculate the members' later time "to let the stores at the right time for the most likely home customers speak.
Presumably, the customer as the status of the sink device stage 5, the new customer (N) flows all the way down sleeping customer (S3). "Intelligent control" can be detected in every aspect, when the water level in the tank or pipeline traffic anomaly occurs, the system automatically make it back, repair, or warning, which is part of the problem, and based KPI automatically make the optimal adjust the settings.
For example, when a customer flow from the main tank S3 customer sink slumber customer, which means up to 90% chance that the customer will be completely lost, namely the so-called break up stage. If male and female friends, before going to recover until the fast break, only half done, the chances of a successful restore is low.
However, when customers began to flow to the main sleeping customer, it must go through S1 (sleepy state), S2 (half asleep state) two sinks. Through intelligent control, can be detected in advance sleepy stage S1, when the customer has found a little alienated, intelligent control will give care or reminders to control the amount of water to reduce the loss of customers.
"Smart Control" can do real-time detection, zero-day communication and personalized messages, instant and timely adjust completely zero-day, zero error, which is the essence of the whole big data marketing.
The figure below illustrates
NES customer level view of the customer more points only five states
Going with the flow, make more low-level water is pumped up, the need for greater momentum. Similarly, NES customer models like three layers of down the sink, to prevent customers from slipping E0 (main customer) S3 (sleep state), to establish a regulatory mechanism, when a customer enters S1 (sleepy state), do the contingency measures.
留言列表
