User value segment

Different users have different value preferences on different networks. Low-priced users requesting high-priced ad sources will lead to wasted requests and time. For this reason, Waterfall settings generally set different segments based on conditions such as region and number of launches to reduce value mismatch. However, these are coarse-grained grouping dimensions. If grouping can be done at the device level, Waterfall strategies can be allocated more accurately, greatly improving the efficiency of ad filling.

1. Function Introduction

To achieve more refined operations, TopOn has built-in device-level user value estimation capabilities, allowing developers to implement strategies for setting segment based on user value at a low cost.

*If the APP needs to use the built-in user value strategy, no additional code development is required, just update the TopOn SDK v6.1.78 and above. Currently, this function is a whitelist mechanism. If you need to use it, please contact your account manager to activate it.

 

2. Create segments based on user value

To set a policy based on user value, you need to add a user value rule under the segment. The specific process is as follows:

(1) Add or edit segments on the mediation page and add user value conditions to segment rules.

(2) A segment can only set one user value interval, and you need to fill in the minimum and maximum values of the interval. If you want to set an interval less than X, you can set "0 ≤ eCPM ≤ X"; if you want to set an interval greater than X, you can set "Y ≤ eCPM ≤ 9999".

(3) User value supports two decimal places and the number range is 0~10000.

(4) You can divide the user value into multiple different user value intervals according to the user value. When the device hits multiple qualified user value groups in this placement, the waterfall flow setting of the user group with the highest priority will be used.

(5) When the historical display count of this placement on a device does not meet the TopOn requirement, this placement will use the waterfall setting of the default grouping.

For example:

If you set multiple segment intervals and priorities as follows:

Segment	User value range	Priority
User value interval 1	$100≤ eCPM ≤ $1000	Priority 1
User value interval 2	$50≤ eCPM ≤ $100	Priority 2
User value interval 3	$10 ≤ eCPM ≤ $50	Priority 3
User value range 4	$0 ≤ eCPM ≤ $10	Priority 4
Default Grouping	-	Priority 5

The results of TopOn ad placement matching segment are as follows:

• When the device's historical display count for this placement does not meet the TopOn requirement, this placement will use the " default grouping ".
• When the device reaches the TopOn display requirement for this placement, and the user value calculated by TopOn is $50, this placement will use the higher priority segment " User Value Range 2 ".

 

3. Set up waterfall flow according to user value

You can set the regular ad source price tiers for waterfall flow based on user value ranges.

If the user value range of the segment you set is " $10 ≤ eCPM ≤ $20 ", the price stratification of the regular ad source of your waterfall flow can be referred to as follows:

• The highest price tier is $25 ( about 30% higher than the maximum user value )
• The second layer is $22 (about 10% greater than the maximum user value)
• ....
• The lowest price tier is $0.01 (no floor price tier, it is recommended to add a no floor price tier to ensure that the placements are filled)

 

4. FAQ

1. How is the user value calculated here?

A: User value will be calculated based on the placement dimension. Currently, it is estimated based on the historical display price of the placement on each device. In the future, factors such as price fluctuation trends and usage duration will be added to the algorithm optimization.

2. How should I divide the user value range of my placement?

A: It is recommended to divide the data into three ranges: high, medium and low, and adjust the ranges according to the data. Initial setting suggestions:

Interval Division	Recommended interval division 1	Recommended interval division 2
Low price range	0 ~ Top 50% of revenue share ad source price	0~ecpm historical average price
Mid-price range	Top50%~80% of revenue share ad source price	The historical average price of ecpm ~ the lowest price of ad sources with a fill rate of 1%
High price range	Top 80% of revenue share ad source price ~999	The lowest price for an ad source with a fill rate of about 1% is ~9999

 

Method 1: (recommended)

Calculate the interval according to the estimated cumulative revenue percentage provided by the backend user value report

Path: [Report] - [User Engagement] - [User Value]，Select [Cumulative ratio of Est. Revenue] from the option below "Data View" View the user value data of the current ad position. Report function description: User Value Report Description.

The process of querying numbers: According to the above interval division, find the [estimated cumulative revenue share] indicator in the user value report, check the eCPM range corresponding to 50% of the cumulative share, and determine the interval.

Example: For the data in the figure below, when setting the user value flow grouping, the estimated cumulative revenue reached 45.32% in the [9,10) interval, and the low-value user interval was set to [0-10).

The value segment settings are as follows:

 

Method 2:

Here is an example of setting user value segment based on historical data:

1. First, click [full report] and observe the historical average price of ecpm for a certain ad position over a period of time, and obtain the user value of the middle and low range of ￥55, which is about $8.
2. Click [Mediation] and look for the ecpm price of the ad source under this ad slot that has a higher price and a lower fill rate but still has a chance to be filled. The user value for the mid-to-high range is ￥120, which is about $16.
3. According to the data, the user value segmengt is divided into "low price range < $8 < medium price range < $16 < high price range", and different value segmengt are set up accordingly.
4. Continue to observe the effect data after setting, and timely adjust the boundary values ​​of the user value range, increase or decrease different base price advertising sources, or increase or decrease the number of value ranges until the effect data is relatively stable.