Editor’s lead: E-commerce data is very important and critical for the e-commerce industry, and observation data is also an indispensable daily work. The author of this article systematically introduces the indicator system of e-commerce data, and introduces the indicator system of e-commerce data from many aspects. Let’s learn it together.

In major e-commerce platforms or e-commerce companies, the most common thing is this kind of data monitoring big screen, which is used to help platforms and businesses understand the business situation in real time.
To observe data, we must first establish a complete data index system and define clearly what we want to see and how to see it.

This paper will introduce the e-commerce data index system through the definition, system construction and detailed explanation of e-commerce data.
E-commerce data refers to digital information used to record user behavior, including quantitative data of a series of behavior habits such as user registration, login, traffic, click, consumption and repurchase.
To analyze the data, we must first understand the business process of e-commerce, and analyze the user’s business path through the business process, what indicators each path needs to pay attention to, and the business loss that each path may bring. Finally, a perfect business funnel is formed.
For example, from the user’s point of view, the business process of e-commerce can include the following important nodes:
From what channels do users become registered users of e-commerce, including their own apps or other delivery channels, including WeChat, Alipay, Tik Tok and below-the-line.
Here, we need to focus on the number of registered users in each channel and the customer acquisition cost of each channel, such as CPA (single registration cost).
When the user is registered, there will be an e-commerce page to undertake, which means that the user will have browsing and clicking behaviors. At this time, the user’s preferences and potential transactions can be judged through the user’s browsing behavior.
Here, we need to focus on observing the UV, PV, browsing time, product click-through rate and so on of each page.
It should be noted that in the e-commerce position, search is a very important entrance. When users have a clear shopping intention, they are more likely to reach the goods directly through search.
Therefore, it is necessary to focus on the relevant traffic of search, including search UV, PV, and result rate.
When users reach the commodity level, they will enter the transformation link of ordering and trading.
This is the core gold process of e-commerce and an important path for user transactions.
Here, we need to focus on the quantity of orders, the number of people placing orders, the amount of orders placed, the quantity of payment orders, the number of people paying, the payment amount, the customer unit price, the gross profit margin and so on.
At the same time, we need to pay attention to the situation of after-sales orders, but when orders are closed due to after-sales or other abnormal reasons, it is also a part of the loss for the platform. Therefore, it is necessary to pay attention to the after-sales order quantity, closing quantity and closing rate.
For the platform, after acquiring a user, of course, it will be hoped that he will repurchase as much as possible and generate more GMV, that is, to improve ARPU(Average Revenue Per User) as much as possible, that is, the contribution value of a single user.
Therefore, we need to pay attention to user activity, repurchase rate, shopping cart situation and so on.
The shopping cart is actually an ordering tool, but according to the user’s habits, the goods added to the shopping cart are more likely to be purchased or repurchased later, so we can also pay attention to the usage rate of the shopping cart at this stage.

In addition to the user’s point of view, you can also build business processes and corresponding data systems from the perspective of goods, orders and stores.
Because users have a life cycle, goods also have a life cycle, from shelves, orders, transactions, to emptying inventory. A set of commodity data index system can also be established.
Based on the above data definition and business process explanation, we can first establish a set of data index system that is as perfect as possible, that is, covering all business processes and business perspectives.
In this BI system, through the indicators of different dimensions, it is built layer by layer.

Users are the foundation of the "people-goods-field" system of e-commerce. No one comes, no matter how high-quality goods and luxurious fields are, it doesn’t make any sense.
Only by operating users well can transformation be formed.
1) Basic user attributes
- Age;
- Gender;
- Provinces;
- City;
- Type: for example, the current identity of the user, such as student, white-collar worker, etc.
- Risk level: In many installment e-commerce businesses involving finance, the user’s risk level is an important attribute, which determines whether the user can obtain the quota and use installment shopping.
The basic attributes of the above users are mainly to help the platform build user portraits and analyze users’ e-commerce consumption preferences.
2) User’s trading behavior
- Days since the last transaction order;
- The amount of the last transaction order;
- Number of historical transaction orders;
- Historical transaction order amount.
Through the first two indicators, we can judge the current active state of the user and whether it is necessary to align for activation or recall.
Through the latter two indicators, we can clearly calculate the customer unit price of the user, that is, the ARPU value of the user, which is the key to measure the user value.
3) User life cycle
- Number of registered users: number of registered e-commerce platform users;
- Number of active users: the number of users who log in to the APP;
- Number of users browsing business details: the number of users browsing the product details page;
- Number of new users: the number of users with 0 historical transaction orders;
- Number of old users: the number of users with historical transaction orders greater than 0;
- Number of repurchased users: the number of users with historical transaction orders greater than 1;
- Number of silent users: the number of users who have logged in the APP for more than 30 days and less than 90 days;
- Number of lost users: the number of users who have logged into the APP for 90 days or more.
Differentiating users according to their life cycle is conducive to the hierarchical and refined operation of users. Different operation strategies are adopted for users at different stages. For example, new users can be urged to complete their first order as soon as possible through new coupons and push, and the number of silent users can be recalled through SMS and exclusive offers.
It should be noted that different platforms have different definitions of their own silent and lost users. Some platforms will judge by activity, while others will judge by transaction. This time, it is only for reference, and it needs to be formulated in combination with its own platform characteristics and business demands.
After users come in, the first pages to accept users are all levels, including APP access, home page, activity pages, business details pages and clicks on elements in each page. These are the key indicators of traffic statistics.
1)APP
- Number of APP openers;
- APP opening times;
- Each tab is exposed to uv;
- Expose PV for each tab;
- Click uv on each tab;
- Click PV for each tab;
- UV click rate of each tab = click uv of each tab/expose uv of each tab;
- The PV click rate of each tab = each tab clicks on PV/ each tab exposes PV.
In general, we define the number of people who open the APP as the number of activities, so we need to measure the activity by counting the number and times of opening the APP.
There are many bottom tabs in current apps, such as Taobao’s home page, shopping, messages, shopping cart and my Taobao.
Each tab will have a corresponding exposure and click, and the click rate can be calculated according to this.

2) Active Page
- Page exposure uv;
- Page exposure PV;
- Click uv on the page;
- Click PV on the page;
- Page UV click rate = page click UV/ page exposure UV;
- Page PV click rate = page click PV/ page exposure PV;
- Per capita exposure = page exposure PV/ page exposure UV;
- Per capita click = page click PV/ page click UV.
Activity pages include home page, activity pages at all levels, and channel pages. For the statistical traffic of pages, the most important thing is exposure and clicking, so that the traffic of an activity page can be judged.
You can also refer to the per capita click, click rate and other indicators to judge the attractiveness of a single activity page to users.
3) Active Page Element
- Element exposure UV
- Element exposure PV
- Element click UV
- Element click PV
- Element UV click rate = element click UV/ element exposure UV
- Element PV click rate = element click PV/ element exposure PV
- Per capita exposure = element exposure PV/ element exposure UV
- Per capita click = element click PV/ element click UV
Active page elements refer to the elements in the above pages, such as package station, banner, icon, feed stream, advertising space, pop-up window and floating.
This is the important data for the traffic monitoring of the active page. Only by knowing the click rate of each element can we know which content users are more interested in, and then guide the operation to design and adjust the page.
Especially for the placement of advertising space, many platforms need internal competition for the content of advertising space.
One of the main bases of competition is the click-through rate. If the content delivered has a high click-through rate, it means that the delivery value is relatively higher. Of course, this needs to be viewed in combination with the subsequent transformation.
4) Business details
- Business details page uv;
- Business details page PV;
- Add a shopping cart and click UV;
- Add a shopping cart and click PV;
- Buy now and click UV;
- Buy now, click PV.
The product details page is an important page in the payment process of ordering, which mainly records the user’s browsing and jumping to place an order.
In addition to basic browsing data, clicking on the car and buying data.
In fact, there are many modules in the business details page, such as the coupon layer, whether the user clicks to see which coupons are available and clicks to collect them.
In the actual embedding point, all the browsing and clicking of the bomb layer should have embedding points, and according to the business needs, the browsing and clicking data of the corresponding module on the page can be observed.

5) Order page
- Confirm the order page UV;
- Confirm the order page PV;
- Submit the order and click UV;
- Submit the order and click PV.
The order confirmation page is also the core page of the order payment process. When users want to place an order for payment, they must go through the order confirmation page to confirm the relevant information of the order, and then jump to payment.
Among them, page browsing data and click-to-submit order data are the most important, which are the key data indicators of the funnel of order placing process.
There are other modules that can also be properly concerned, such as the coupon layer, whether the user clicks to see which coupons are available and clicks to use them; Select the receiving address, elastic layer, etc.

6) Payment page
- Payment page uv;
- Payment page PV;
- Click uv to pay the order;
- Click PV to pay the order.
The payment page is the last page of the payment process, and it is also the key to the transaction. Only when the user finally makes a transaction will it mean the improvement of GMV. The payment page is also commonly called the cashier page.
Page browsing data and pay-per-click data are the most important, which are the key data indicators of the order process funnel.
Many platforms will also open other functions at the checkout counter, such as choosing different payment methods, supporting the choice of installment payment and choosing different installment numbers, and observing users’ payment preferences through their clicks.

In e-commerce business, search is a very important business entrance. When users have no clear purchase target, they may browse the recommended products in the feed stream.
But when users have a very clear purchase goal, search is always the first entrance for users.
1) Search the basic situation
- Search exposure uv;
- Search exposure PV;
- Search and click PV;
- Search and click PV;
- Search UV usage rate = search click UV/ search exposure UV;
- Search PV utilization rate = search click PV/ search exposure PV;
- Number of searches per capita = search click PV/ search click UV;
- Per capita search terms = (search results words+search results words)/search clicks UV.
Search portals will exist in multiple pages, usually on the home page, so that users can clearly see them, so it is necessary to record their exposure and click data and analyze the user’s use of search.
In addition, we can observe the number of searches per capita and the number of words. If users search a lot, on the one hand, it shows that their purpose of coming to the platform is clear, on the other hand, they need to pay attention to whether the relevant recommendations are not accurate enough, which leads users to use search frequently.

2) Search has results
- Search has results uv;
- Search has a result PV;
- Search for results, click on the results UV;
- Search for results and click on the results PV;
- Search results without clicking UV;
- Search results without clicking UV proportion = search results without clicking UV https://www.woshipm.com/data-analysis/search results with UV;
- Search for the amount of words with results;
- Search for the number of words with results and no clicks;
- Proportion of words with search results without clicking = words with search results without clicking/words with search results.
According to the search results, it can be divided into search with results and search without results. When the search has results, you can check the number and times of people who have search results and whether they click after the results.
If the search has results, but the proportion of no clicks is high, it may be necessary to pay attention to whether the search results are not accurate enough, and the products searched out are not what users want, so users don’t want to click.

3) Search has no results
- Search for no results uv;
- Search for no result PV;
- Proportion of people who search for no results = search for no results UV/search click UV;
- Search for fruitless words;
- Proportion of search words without results = search words without results/(search words with results+search words without results).
Compared with the search with results, the search without results needs to focus on the problem of word quantity.
If there are a lot of words with no results, and the proportion is high, it means that there is a shortage of goods or the search matching algorithm needs to be optimized.
When the proportion of users who have no results in the overall search increases, they need to pay attention to the optimization of search products immediately. If more and more users come to search but don’t get feedback, then more and more users will be lost.
In the search data, the observation idea can be:
- Pay attention to the data without results first, reduce the proportion of search without results, and ensure that users can find things;
- Then pay attention to the data with results, reduce the proportion of results without clicks, and ensure that what the user finds is what he wants;
- Finally, pay attention to the click data with results, and judge whether the things searched and clicked by users have finally formed a transformation.
Commodities mainly depend on their sales volume and profit, with sales volume paid by placing orders and profit by gross profit.
1) Basic information of commodities
- Quantity of commodities;
- New quantity of goods;
- The number of goods in stock;
- No inventory quantity of goods;
- The goods are sold in quantity;
- No sales quantity of goods;
- New rate of goods = new quantity of goods/quantity of goods;
- Sales rate of commodities = quantity of commodities sold/quantity of commodities;
- The unsalable rate of commodities = unsold quantity of commodities/quantity of commodities;
- Commodity dumping rate = quantity of goods out of stock/quantity of goods.
The quantity of goods reflects the ability of "goods" in e-commerce, and transactions can only be made if goods are available first.
But in addition, we need to pay attention to several major indicators, and whether the new rate feedback platform on commodities has continuously iteratively updated commodities.
The sales rate of commodities ensures that not only commodities are available, but also people buy them.
If the slow-moving rate of video is too high, we need to take corresponding measures to enhance the exposure of long-tail goods.
If the commodity dumping rate is high, on the one hand, it can reflect the preferences of platform users, on the other hand, it is necessary to pay attention to whether the supply chain is complete and whether iterative optimization is needed to support commodity sales.
2) Commodity sales
- Number of people placing orders;
- Number of orders placed;
- Order amount;
- Number of payers;
- Payment odd number;
- Payment amount.
The data indicators of goods ordering and payment mainly feed back the sales of goods.
3) Commodity profit
- Gross profit of self-operated goods = transaction amount-procurement cost-storage cost-logistics cost+rebate;
- Gross profit of pop products = transaction amount * deduction ratio-platform promotion expenses.
The profit of commodities can be divided into self-operated commodities and pop commodities to calculate.
For the big platform, there are both self-operated goods and shops, and the profit of the platform is calculated by deducting points.
The gross profit of self-operated goods needs to deduct the purchase cost, storage cost and logistics cost from the transaction amount of goods.
In addition, there may be cooperation with the brand, and the corresponding rebate can be obtained for selling a certain commodity, which is also part of the income.
It should be noted that only the gross profit of goods is considered here, all from the perspective of goods, such as labor costs are not taken into account.
For pop goods, that is, goods sold in stores, the commission is generally calculated by deducting the proportion of points, and at the same time, the promotion expenses spent by the platform, such as the amount of coupons issued by the platform, are deducted.
Order is the core of e-commerce system and the key to all transformation and statistics. The whole order process includes ordering, payment, completion and after-sales.
1) Place an order
- Number of orders placed;
- Number of people placing orders;
- Order amount.
2) Deal
- Number of transactions;
- Number of transactions;
- Transaction amount;
- Unit price of clinched a deal = clinched a deal amount/clinched a deal quantity;
- Transaction customer unit price = transaction amount/number of transactions;
- Turnover rate = turnover/order number;
- Transaction order conversion rate = transaction quantity/order quantity.
In addition to the basic order transaction index, it is also necessary to observe the customer unit price, which is the key to measure the user’s value, and can also understand the user’s consumption habits.
In addition, through the conversion rate of the transaction, we can check the loss of the order, and then optimize the process funnel of payment under the order.
3) Close the bill
- Quantity of customs clearance;
- Number of people closing the bill;
- Amount of customs clearance;
- Automatic closing quantity;
- Manual closing quantity.
Order closing includes that the user cancels the order voluntarily or the system cancels the order for various reasons.
In addition to paying attention to quantity, we need to pay attention to the specific reasons for closing orders, such as the classification of the reasons why users cancel orders voluntarily, whether the price is too high or the quality is not good.
It is equivalent to a user survey, so as to iterate according to user feedback.
4) After sale
- Number of after-sales orders;
- Number of after-sales orders;
- Number of successful after-sales orders;
- Number of failed after-sales orders.
After-sale forms generally include refund, return and exchange. Observing the quantity of after-sales orders and the number of people can confirm the performance quality of orders. If the quantity of after-sales orders is too high, it indicates that there are problems in the quality and service of goods.
At the same time, we should also pay attention to the results of after-sales orders and judge whether we have provided good after-sales service for users.
E-commerce marketing is mostly completed through coupons and various promotional activities, including single product direct drop activities, full discount activities, and group activities. The basic measurement indicators of different activities are similar, but there are some gaps in the corresponding indicators because of the different settings of coupons.
1) coupons
- Circulation;
- Number of recipients;
- Number of coupons received;
- Collection rate = number of coupons received/circulation;
- Number of people using coupons;
- Number of coupons used;
- Coupon rate = number of coupons used/number of coupons received;
- Number of people placing orders with coupons;
- Number of orders placed with coupons;
- The amount of the order with the coupon;
- Number of people paid by coupons;
- Pay the single amount with coupons;
- The amount paid by coupons;
- Coupon investment amount = coupon payment amount * coupon denomination;
- Coupon ROI= coupon investment amount/coupon payment amount.
The coupon collection rate and coupon utilization rate can help operators find the funnel problem, whether users do not collect coupons or place orders without coupons after receiving them.
The ROI of coupons is the key to measure the value of coupon activities. The order amount brought by coupons can be confirmed by paying the amount with coupons, and then the total input cost can be obtained by multiplying the total number of coupons consumed by the coupon denomination, and finally the input-output ratio can be calculated.
2) Activities
- Number of people placing orders at the event;
- Number of active orders;
- Activity order amount;
- Number of people paying for activities;
- Activity payment bill quantity;
- Activity payment amount;
- Amount invested in the activity;
- Activity ROI= activity input amount/activity payment amount;
- Activity indicators are similar to coupons, mainly taking ROI as a measure of activity value.
The above indicators such as traffic, orders, commodities, marketing, etc. can be used not only for platform self-operation, but also for the operation measurement of platform-settled merchants.
If there is a platform with both self-operated and pop merchants like JD.COM, you only need to apply different indicators to different objects.
Data is a reliable basis for business improvement and iteration, a powerful basis for upgrading and transformation, and a reliable basis for strategic decision-making.
It represents the user’s behavior habits and ways, and it needs to be calculated, observed and analyzed to see the conclusion.
Data plays an irreplaceable role in e-commerce and is the guarantee for its continuous development.
Only by building a complete and clear data index system can we achieve business transformation through data promotion.
The core of e-commerce data indicators includes users, traffic, search, goods, orders and marketing.
Each company and business may be slightly different, which can be adjusted according to the actual situation, and finally build a data system that matches its own business.
Author: Ball, WeChat WeChat official account: product balls.
This article was originally published by @ yo-yo. Everyone is a product manager. It is forbidden to reprint without permission.
The title map comes from Unsplash and is based on CC0 protocol.