Mobile Ads for beginners

According to the definition of Google Ads support: Mobile ads are concepts to show ads on websites or applications (& games) provided on mobile platforms – devices such as cell phones, tablets.

This is the simplest and most complete definition of Mobile Ads. That’s it! But do you clearly understand how to use it? In this series of articles, I will dig deep into a series of  basic concepts, the formats and the integration method of Mobile Ads into your application.

Part 1: Basic concepts of Mobile Ads

Currently, there are many networks providing Mobile Ads, including some popular networks such as: AdMob (Google), FAN (Facebook), Unity, Vungle, etc. Each network has some custom for itself, but generally it follows a certain “process” to make ads display.

Here is the general process:

Based on this process, we need to pay attention to the following basic parameters:

• Request (number of requests)
• Matched request (required quantity is met)
• Match rate (response rate)
• Impression
• Show rate
• Click (number of clicks – clicking of fingers ^^)
• CTR – Click through rate
• eCPM – RPM: revenue per thousand impressions

1. Request, matched request và match rate

The first thing to do when using Mobile Ads is ads request. Your application sends the request to the server of the Ad Network , this request will trigger an auction process, and the Ad network will search for ads that match your request (based on user information, device, location, etc.). If there is a suitable ad, the server will return the device. This returned ad will count as a matched request.

Match rate formula

Matched request / request ratio is called the match rate (or fill rate). Match rate (response rate) shows the ad serving capacity of the Ad network. The higher the rate is the more ads the network can deliver, which makes sure that even if you have more users, they still have enough ads to show them.

.

Factors affecting match rate

Match rate depends on several factors such as: Ad network, users (country, age, etc.), format (ad format). Depending on the market (user’s country), the match rate will have different thresholds. Tier 1 (top) markets like USA, UK, Japan, etc. will usually have higher match rates than lower tier markets like India, Brazil, Mexico.

In addition, ad formats will also affect match rate. Typically, interstitial (interstitial ads) will have a quite good match rate, probably followed by a banner and a rewarded (rewarded ads).

At the same time, different networks will have different match rates. Not only that, the definition, the ways to count requests and calculate the match rate of the networks are also different.

Specifically, networks like AdMob and FAN have specific request API (function), to proactively send requests to the server. Each call to the request function will usually be counted as one request. With this form we have a more active view of how many times the request is met and how many of them are met.

Other networks, such as Unity and Ironsource, offer automatic request mechanisms. We don’t need to send  the request function to the server, instead the SDK (ad network tool) will automatically handle the request. Whenever the ad needs to be displayed, we only call the API to check if the ad is ready or not. This mechanism makes it easier to integrate ads. However, in my personal opinion, it makes the feeling that not being proactive, controlling the request. In case, I need to show the ad but the ad is not available (no matched request), I want to actively request it again as soon as possible rather than waiting. At the same time, this mechanism also hides counting requests. For example, I don’t know how Unity currently counts requests, and Unity’s dashboard (report panel) often shows very high Fill rates (~ 98-99%). It feels like Unity only counts requests when they “almost” respond to them.

 It’s you that the last factor affects match rate. In the case of using networks with manual request mechanisms (AdMob, FAN, etc.), it is your ad request logic that can influence the match rate. (I will share some methods to manage requests in the following article). In addition, the mechanism of setting the floor price (set floor eCpm) will also be an important factor affecting the match rate.

2. Impression and Show rate

Once you have a matched request, your mission is to display it in the right place :). This step is almost error-free, 99% of cases are that after loading the ad, you only need to call the API to show and the ad will appear. However, Banner formats (banner ads) and Native will require additional processing. Banners require a large enough space to display and properly handle the UI to place the banner. Native is more complicated when you have to make the layout and handle each of its components. With Native, if not handled correctly, you can even display ads but the user cannot click.

In general, the display of the ad is still entirely your decision as well as your handling responsibility. Each successful call will create an impression (display). The ratio between the number of impressions and the number of matched requests is called the show rate.

Show rate does not reflect or affect the performance (efficiency) of an Ad Network. In fact, not many Ad Networks measure this concept. However, the show rate may reflect some issues that your application may encounter. For example, with Rewarded Ads, the too poor show rate indicates that users are not viewing your Rewarded ads, possibly due to reasons such as:

• Rewards for Rewarded ads are not of high value, does not motivate the viewers
• Design of your Rewarded ads is not good, it is difficult for users to find ads to view

3. Click, CTR and eCPM

Most of the current Ad Network calculate revenue based on click, so impression is not enough, whether users click on your ad or not that  determines your revenue. So we need to care about CTR – Click Through Rate – the “click” rate.

CTR – Click through rate

CTR reflects the average number of clicks you got per 100 impressions. The higher this number, the easier it is for users to click on your ad. CTR depends on the following factors:

• Ad format: Normally Interstitial, Rewarded will have high CTR, banner CTR is lower.
• Ad Network and the ‘value’ of advertising: different Ad Networks will have certain differences in the same format, so the CTR of each Network for each type of format may be different. At the same time the ‘value’ of the ad (attractiveness, relevance to users) also determines its CTR. The value of the ad also changes from market to market, so the CTR will also differ on the same Ad Network, in the same format but in different markets.
• Your integration: you can still control the CTR in some ways. Ad placement, frequency of displaying ads are factors under control. Same banner but put in the position on top or bottom may return different CTR, Banners with various refresh rates will create different impressions and CTRs; the same interstitial but display 2 minutes / times compared to 5 minutes / times will have different CTR. Game / App is also similar, same market, but you cannot expect that showing 10 impressions / users / day will bring the same CTR as showing only 2 impressions. Sometimes showing ads too much will not increase your clicks / revenue.

Looking at CTR can give us some insights. If the CTR is too low, you should review the ad position, the frequency of display ads (or refresh of the banner). The CTR is too high, you should also consider carefully, whether the display of your ad is confusing, causing wrong clicks from users? The misleading integration may be stopped by the Ad Network from providing ads for you.

CPC, eCpm – Advertising Value 

The final factor that determines your revenue is CPC – eCpm, which is the value of the ad itself. CPC is the cost per click – the value per click. eCpm is the average revenue per 1000 impressions. In general, these two values can be converted back and forth.

CPC is a factor you can hardly change. The CPC depends on how much the advertiser is willing to pay to buy the ad placement.Therefore, it will depend on the market and users of the application.A good Ad Network will provide the most “relevant” ads to your app: the ads that are the most “suitable” to the user and have the highest value for placement within the app. And of course, users in top markets like the UK and Australia will have higher value than Indian, Brazilian and so on.

Look at CTR – Click through rate, you can interfere in some parts of it based on how the ads are integrated: placement, how often you show the ads, etc. Showing too much ads in high density will reduce your CTR. Therefore, CTR itself may differ in the same market, format, ad network.

Combining CPC and CTR, eCPM will also vary depending on the application type, user set (market) and the way you integrate the ad. Therefore, the comparison of eCpm should only be relative. You can compare eCpm between 2 networks in the same application, meanwhile between 2 diff apps (maybe similar apps in different ads integrations) eCPm will not be completely the same, even in apps sharing the same format and market.

In addition, eCpm is the average value per 1000 impressions. So, if your data is not large enough (at least a few tens of thousands of impressions) then your eCpm is not accurate and does not reflect the performance of your network as well as your application. 

So, are there any ways to optimize eCpm? Focusing on the 2 components of eCpm are CTR and CPC, we have some methods:

• Use mediation to make Ad Networks competition and give us the highest advertising value
• Optimize ads placement and frequency as well
• Attract high value users (top market)!

In the next articles I will share about each format and how to use mediation.

(To be continued)