This document contains guidance to help you avoid making mistakes when rating Search Ads tasks. The examples are similar to tasks rated in the past.
The rating is a two-step process:
According to GL 3.1. Query Type Question, we need to select the Query Type. To do so, we must determine whether the user had a particular app in mind or was searching for any app that meets their needs. Possible ratings are Navigational, Functional, Mixed, and Unclear.
Examples:
Query | Type | Explanation |
think dirty | Navigational | The user is looking for the Think Dirty app for retrieving cosmetics and personal care products information and comparisons. |
ausgabenmanager | Mixed | The user might be looking for the Ausgaben Manager app, or similar money management apps. |
tiktok analytics | Mixed | The user might be looking for the respective Tik Tok function, or the Tik Tok Analytics app, or similar apps providing insights into their Tik Tok account(s). |
In line with GL 4. Ad Relevance, after understanding the intent of the query, we rate the relevance of the returned ad. Relevance rating options are Excellent, Good, Acceptable, and Bad.
An Excellent ad features an app with a strong relationship to the query intent. Such apps are most likely of interest to the user.
A Good ad features an app somewhat related to the query intent. Users will be quite likely interested in the app, but other apps might be more compelling.
An Acceptable ad features an app that is only slightly related to the query intent. Users would not be surprised to see the ad, but will rather not be too interested in the app.
A Bad ad features an app that is entirely unrelated to the query intent. It might surprise the user and could be detrimental to the user experience.
The user is looking for a specific fashion brand, Vero Moda. The advertised app is form another fashion brand; however, vero moda offers a wide range of apparel for women, while Hunkemöller caters to specific segments and target audiences. In this regard, they are not direct competitors. Considering this, both Good and Acceptable are acceptable ratings.
The user is looking for the Ausgaben Manager app, or similar money management apps. The advertised app provides stock exchange information and data. While both apps are related to money and finances, they focus on different aspects. There’s only a weak link between them. Users wouldn’t be surprised to see the ad, but it’s not very likely that they would be interested in the advertised app.
The query is slightly misspelled, but it’s clear that the user is looking for the pubg battle royal game, a third person multiplayer shooter. The advertised game is a building and strategy game in isometric perspective. The game mechanics and graphics are very different, appealing to a different audience. When rating games, factors to take into account are play style, presentation, and audience, GL 4.3. Game Queries.
The user is looking for the Worms series of artillery games. The advertised game is a racing game with abstract graphics. Here as well, the game mechanics and graphics are very different, appealing to a different audience. For game queries, themes are important – users are likely to be attracted to games with similar presentations to what they were looking for, even if they don’t have exactly the same gameplay. If gameplay is too different, however, the themes don’t match at all.
The user is looking for the Say&Go voice recorder. The advertised app also features a recording function, but is geared towards translating speech rather than recording and storing it. Both apps have a different scope of application; a translation app is unlikely to be of interest to users looking for a voice recorder.
The user is looking for the Kaufland app. The advertised app is a customer card wallet. While Kaufland offers a customer card to interested customers, there is no connection between the query and the advertised app.