As you read this post please note that, other than on the “Ratings” dashboard, Appbot analyzes and reports reviews that contain a text component. Learn more about the difference between ratings (stars-only) and reviews (stars plus text comment) here.
The Sentiment page is one of the different types of clustering that we offer. It’s effectively a way for you to get a quick view into what users think of your product.
You can determine which app we’re looking at by toggling between the different apps from the top left corner here.
The filter bar at the top of the screen is pretty consistent and you’ll see it on most of the pages as you go through Appbot. It allows you to filter what the page is displaying by date range, country, app version, star rating or even by keyword.
The Sentiment Timeline at the top shows you the positive, neutral and negative sentiment breakdown over time on a daily basis. The main use of this particular chart is so that you can see reactions to any changes you’ve made (particularly after you’ve released a new version). If you hover over any of the days, you’ll see for that day how many reviews were positive, neutral and negative.
Below you’ll see the Overall Sentiment gauge and Sentiment Breakdown. Overall Sentiment gives you an idea of where your app sits in the spectrum of possible sentiment grades, and the Sentiment Breakdown chart shows you what portion of the reviews were positive, neutral and negative for your selected time period.
Next, you’ll see the Sentiment Score and Suggestions section. The Sentiment Score is a report card style grade so the scale is from a D- to an A+. We look at the same sort of things that the app stores look at when determining if your app deserves a feature or should rank in the app store. These include:
- Volume of reviews for the time period.
- Trend in review volume (whether it’s trending up or down).
- Star rating of reviews that contain text.
- Trend in star rating.
- Ratio of positive, neutral and negative reviews.
- Trend in the ratio of positive, neutral and negative reviews.
Then, we provide you with some suggestions on how to improve your score ????
Below you can see charts showing your review volume and star rating over time. Hover your mouse over the graphs to isolate data points.
At the bottom here you can see the country (iOS) or language (Google Play).
The Version Breakdown is available out of the box for iOS apps. Google Play doesn’t make the version data available publicly, but if you link your Google Play Developer Console we will be able to access that data from the console and generate this report.
Countries/Languages and Versions are there to give you a breakdown of where your reviews are coming from. If you hover over any of the charts, you can see the percentage breakdown.
The Words tool finds trends and patterns in the words used in your reviews. You can see all the different word clusters we offer at the top of the chart here.
Interesting Words highlights the top 10 words used in your reviews that we think you might find most interesting.
Next to the words you can see the positive to neutral sentiment ratio, how many reviews from that time period mention it, the percentage of the total reviews that makes up and a timeline of the mentions of that word over time.
If you want to drill in and get a feeling for what the reviews that contain this word are about, you simply just click on the word and that will take you through to the reviews. As you scroll down the list you’ll see that every time the word has been mentioned it get’s highlighted for you. This is one way that customers identify bugs that aren’t super obvious if you were to read through all the reviews.
Back on the Words page, Popular Words shows you the 10 most commonly used words in your reviews for this period of time. We filter out all of the stop words so you won’t see words like ‘and’ and ‘the’ appearing here. We also stem the words, which means we take each word, look for close variants and put them together – like love, loved, lovely and loves.
- Our customers use Popular Words to identify words where the sentiment for the reviews containing that word appear to be dissonant with the rest of the reviews that they have.
- Another use case is reading down the list and looking for things that could indicate a feature request, or that relate to a specific feature.
Back on the Words page, you’ll see the next section after Popular Words is Critical Words. This is where we monitor a list of words we’ve specified that refer to performance, stability, bug and crash type issues.
Check out Trending Up, Trending Down and New Words too for further analysis.
Topics takes the concept or words and expands it out to another level. Instead of just telling you words that are appearing commonly, here what we’re doing is reading each review and making a judgement on what the review is about in terms of its themes. We read the entire review, and quite often we’ll notice a single review matches multiple topics, so it’s normal for a single review to appear under multiple topics.
The topics summary table works in a very similar way to the words table. You can see the topics have been ranked in descending order of popularity. On the right hand side you can see the sentiment of the reviews that have been matched into each topic, the number of matches, the percentage of your overall reviews that topic represents, and the same sort of trend line we were looking at in the words table.
Like words, if you want to look into the reviews making up a topic, just click on it from the list. In the list, you’ll see we’ve highlighted the sentence that lead us to classify that review into the topic.
The use case here is to help you look at a group of reviews that are similar, and help identify the sub trends that are common within the group.
Custom Topics is very similar to topics, but the technology behind it is a little bit different. When you setup a custom topic, you tell us what you want us to look for. We can monitor for keywords or phrases, and we will only look for exact matches of those words and phrases you’ve specified.
Your custom topics are exclusive to you, so you can make as many as you like and they won’t affect anyone else in you team.
To create one, press on the blue “+ Add Custom Topic” button.
Then give it a name, enter in the keywords/phrases you want to search for, the keywords/phrases you want to exclude, check if you want to scan translations too and hit “Create”.
- Phrases can be as many words as you like, but obviously the longer the phrase the less matches will be found.
- The “Excluded” acts as a blacklist or negative keyword list and will exclude reviews from the results that contain the keywords/phrases you enter here. For example you might include the keyword “watch” but exclude the keyword “apple watch” from results.
For each custom topic, you’ll see the same breakdown, the sentiment for the reviews, the number of matches, the percentage overall and the spike chart.
Once you’ve setup a few Custom Topics, clicking on any of them from the list will take you to a list of all the reviews in that custom topic for the selected time period.
A prime use case would be if you’ve got a problem with a specific feature, setting up a custom topic is an easy way to monitor how that’s doing over time.
Appbot Ratings is the page in Appbot that counts every app store rating your app has received, whether they contain text or not, whereas all of the other tools in Appbot only analyse reviews that contain a text component.
The Ratings page analyses your app store rating for iOS and Google Play apps only.
For iOS apps, you can use the dropdown at the top of the page to select which country you would like to view metrics for.
*Note: Only global metrics are available for Google Play Apps.
The Ratings Over Time chart shows how your average app store rating is changing over time. It shows the total count of ratings and star count, absolute and stacked.
*Note: While the stats panel at the top of the Ratings page updates in realtime, the Ratings Over Time graph is updated approximately once per day.
Custom Dashboards are really quick and simple to create and are simply a way for you to go through all these screens, pick out the bits of information that are useful to you, pin them onto a dashboard and arrange them any way you like.
Creating these dashboards is really simple, and takes less than a minute to set each up.
- Press on ‘Create New Dashboard’ button from the Custom Dashboards page.
- Give your new dashboard a name and press ‘Create Dashboard’
- Pick a few widgets from the left hand side to add to your new dashboard and press the ‘Add Widgets’ button. Tip: You can hover over the widgets to preview what they look like!
Congratulations, you’ve learnt how to build a custom dashboard and your new dashboard has been created! You’ll be immediately taken to your new dashboard and you’ll now see it has appeared the dashboards list.
From the Export CSVs tab, you can download a range of pre-configured reports for multiple apps at once, and for any date range you require.
To download a report, just find the one you want from the list, set the date filter to whatever you require and hit the “Download CSV” button to the right of it.