Current location - Recipe Complete Network - Catering industry - How does o2o in business circle play big data?
How does o2o in business circle play big data?
Online data mainly include: visit volume (IP UV PV), average browsing time (page views), new UV ratio, bounce rate, conversion rate (registration, order, payment), traffic source (search, direct access, connection, area, promotion), webpage opening time, website hotspots, search analysis, etc.

ERP data mainly include: order quantity, customer unit price, gross profit margin, secondary purchase rate, loyal customer conversion rate, customer churn rate, sales rate, out-of-stock rate, commodity price change, SKU quantity change, turnover rate, return rate, category sales ratio, member registration amount, registered member conversion rate, etc.

The complaint data of customer service return visit questionnaire mainly includes: complaint classification, UI impression, category impression, price impression, website function impression, logistics experience impression, after-sales impression and so on.

The above data are interrelated. For example, when analyzing the effect of promotional activities, it is necessary to analyze the changes in the number of visits, the change in the conversion rate of registered orders, and the changes in the sales volume of promoted goods and normal goods.

How to analyze the data?

Some companies set up special data analysis departments to not only provide data, but also complete data analysis. This way of working, although the basic data is accurate, the analysis results may be quite biased. Because data analysts are unfamiliar with business, their understanding of all kinds of information is not as good as that of marketing department, operation department and other business departments.

For example, the sudden decline in the sales proportion of a certain category may be due to the change of marketing methods of the marketing department or seasonal factors. If the data analyst doesn't know this information, he may simply judge that the customer doesn't welcome this kind of goods and make a decision to suggest that the commodity department reduce the proportion of this kind of goods.

The more reasonable data analysis method is that the data specialist provides the basic data and the key personnel of relevant departments analyze it together. For example, the decline in conversion rate should be analyzed by the marketing department, the operation department and the commodity department to find out what causes it.

For new projects, the objective analysis method can be introduced. Target analysis is based on "the introduction cost of new customers" and "the conversion rate of loyal customers", and sets reasonable targets to judge whether the business model is feasible.

For example, a B2C website with an investment of 50 million yuan has a promotion budget of 25 million yuan, and the goal is to reach 5,000 orders per day. The definition of loyal customers is to shop once a month on average, and the sales target of 5,000 orders a day needs 6,543,800+5,000 loyal customers.

If the actual business performance shows that the introduction cost of new customers is 50 yuan and the conversion rate of loyal customers is 30%, it will reach 6,543,800+5,000 members, and it will cost 25 million promotion expenses.

According to the data analysis, when the introduction cost of new customers is higher than that of 50 yuan, and the conversion rate of loyal customers is lower than 30%, the project cannot achieve its goal. If the target is close to the actual performance data, performance can be improved by optimizing internal strength. If the data is too different, it means that the business model may not be feasible. We should adjust the business model as soon as possible and repeat the above data analysis steps in the process of trial and error.

The most important data, I think, are the introduction cost of traffic, the introduction cost of new customers and the conversion rate of loyal customers. Traffic import cost data is mainly assessed by the marketing department. Marketing department, operation department and commodity department are jointly responsible for importing cost data for new customers, while operation department and commodity department are mainly responsible for the conversion rate of loyal customers.

The extended analysis includes flow analysis, residence time, flow page and conversion rate analysis. The increase or decrease of traffic (new UV data) represents whether the promotion work of the marketing department is effective, the number of pages visited by new visitors, the conversion rate and other data, and also represents whether the promotion of the marketing department is targeted to some extent.

The introduction cost analysis of new customers is an important KPI of promotion efficiency and the promotion funds invested by each target. For example, a promotion method brought 10000 uv, 500 registration and 100 order. And this method needs 1 1,000 yuan, so the investment of each UV, registration and order in 0 yuan and 20 yuan is 1 1,000 yuan respectively. The new customer introduction fee for this promotion method is 100 yuan.

How do the data analysis results match?

The important work of the marketing department is to try different promotion methods, calculate the return on investment of each promotion, and focus on investing and optimizing the promotion method with the highest return on investment according to the data analysis results.

Improving internal strength is the basic method to optimize the introduction cost of new customers and the conversion rate of loyal customers. Internal strength includes: product structure, promotion method, website experience, logistics experience, customer return visit complaints, membership marketing, etc.

The purpose of commodity structure optimization is to understand the customer's needs through data analysis, continuously introduce and eliminate commodities, and make the commodity structure meet the customer's needs as much as possible. Establish a dimension table of goods, comprehensively consider all dimensions of goods such as price, model, shape, brand and specification, classify goods according to different dimensions, and analyze the sales volume of each category and dimension with data, so as to increase the proportion of goods with high sales dimension and streamline the proportion of goods with low sales dimension.

The process of commodity introduction and elimination is also influenced by many factors. For example, "structural goods" can't be eliminated even if the sales volume is not good, and "seasonal goods" need to consider seasonal factors.

The promotion method mainly relies on data analysis to evaluate the effect. Every time you do a theme promotion, you will create a promotion document in the ERP system, and set the promotion theme, goods and schedule. Through BI tools, this paper analyzes the changes of sales volume of promotional products, gross profit loss, normal sales volume and shopping frequency of new and old members driven by promotional activities, and comprehensively evaluates the promotion effect to guide the next promotion activities.

Website experience optimization can be expressed by a formula: UEO (User Experience Optimization) = PV/OR (Website Bounce Rate), with the purpose of reducing the customer's Bounce Rate and making shopping more convenient for customers. This is based on a full understanding of website positioning and customer characteristics, such as making the layout of the website clearer and making the shopping process smoother for customers. Through hot spot analysis, we can know the position of customer's concern and put the content of customer's concern in hot spots. Through the analysis of the pop-up rate, the recommended content is displayed on the page that customers can easily jump out, thus attracting customers to stay on the website.

The analysis of customer impression questionnaire complaint data can find customer dissatisfaction, establish complaint channels on the website, and the customer service department should pay a return visit to new and old customers. Pay a return visit to the customers who made the order but didn't submit it at last, and make statistics from UI, category, price, website experience, logistics, after-sales and other aspects, analyze which aspect affects the customer experience most, and solve them one by one according to the actual situation. Constantly optimize.

Membership marketing is to divide members into different types and conduct marketing according to their characteristics. It can be divided into: registered customers who have not placed orders, customers who have placed orders for the first time, loyal customers, high-value customers and lost customers.

For registered customers who have not placed an order, if they leave a mailbox, they should issue some large-scale coupons to attract customers to place an order at the first time and experience the service intuitively.

Customers who place an order for the first time should put some reminding gifts in the package, such as mouse pads printed with advertisements, to remind customers at any time and increase the chances of customers placing an order for the second time. Customers who place an order for the first time may not know the main selling points or advantages of our website, so they can instill this information into customers through parcels or emails. The customer service department should pay a return visit to customers who place orders for the first time to understand their feelings.

Loyal customers are customers who buy repeatedly. Through data analysis, we can understand the needs and demands of loyal customers and put forward some targeted suggestions. If there is enough gross profit space, VIP cards can be issued to loyal customers to maintain loyal customers. For loyal customers, play the role of points, recommend some points to loyal customers as gifts, and develop loyal customers into word-of-mouth promoters. If loyal customers invite new members, they should be rewarded with points.

It is necessary to conduct targeted marketing for the lost customers, understand the reasons for the loss of customers, and give preferential treatment to the lost customers. High-value customers do not often buy, but the unit price is high and the gross profit of the goods is high. If high-value goods are marketed to high-value customers like ordinary customers, it may be counterproductive to recommend them to such customers.