We have 2 pricing models:

‍

MAU-based (monthly active users)
You can independently calculate the cost of our service using the calculator on our website.

‍

Revenue Share
For apps with over 100,000 MAU.

We only operate on top of analytical platforms. The most common option is a connection through Firebase. We are currently integrated with Amplitude. Integration with Adjust, AppsFlyer, Mixpannel, One Signal, PushWoosh and Braze will be available soon.

Firebase
‍To work with Firebase, it needs to be connected to BigQuery (see instructions). We export events on a daily basis through an API. All the client needs to do is to create a server account (see instructions) and give us access — we need to output a JSON file and stream_id.

‍

Amplitude
‍Amplitude is capable of exporting via API — from there we export all available events on a daily basis. The client needs to provide us with the API Key and Secret Key (see instructions). We can download events via HTTP or from AWS bucket S3.

We only need 2 types of events:

‍

1. Screen transition events.

2. Conversion events for which our AI algorithms will be optimised: conversions, subscriptions, purchases, views or clicks on ads.

3. Clicks on the app’s main elements (optional)

‍

More events will improve the quality of optimisation, but these two types of events are sufficient.

You can see the general dashboard for the app on the console homepage. The Retention report can be found in the Reports section. The console generates a weekly Retention report. The custom report provides all other metrics: mailing metrics (open, send, CTR metrics), a cohort report for all target events, and daily Retention. Output format: Google Sheets (can be converted to .csv)

We work with iOS and Android.

You must have at least 6,000 MAU to get connected to nGrow.

All nGrow clients get access to a dedicated Slack channel, where Customer Success, Integration and Analytical managers connect.

We send push notifications in the language set by the user in their phone settings. For example, if you live in Germany but your phone interface is in English, you’ll receive push notifications in English.

The first results of campaigns and their impact on revenue and retention can be measured after a 2-week interval for large apps (>300k MAU) and after a month for small and medium-sized apps (<300k MAU). After this period, you can see the baseline results, which will improve with subsequent training of the system and work on creative approaches.

We have our own unique AI algorithm, which is called multi-armed bandits.

‍

Multi-armed bandits are an algorithm for shuffling push text variants in such a way as to achieve a gain for the opens, retention or conversions. The target metric is typically the click-through rate (CTR), although it can vary: for example, subscriptions, ad-impressions, in-app purchases. Algorithms automatically test combinations of notifications for different texts with varying frequency, iteratively improving and delivering more targeted push notifications and send times each time. Ultimately, the user receives push notification that has the highest probability of being opened.

‍

To describe this process in detail and in stages:

‍

User segmentation
The first stage is user segmentation, where we determine which users we want to send push notifications. You can segment by any event, property, device age, or time related conditions. For example, send users who haven’t opened the app for more than 30 days a promo code with a discount, and we’ll send users who log in to the app every day push notifications much less frequently, so as not to be intrusive.

We assign optimisation parameters to segments
The algorithm makes an assumption about the effectiveness of push notification parameters for each specific user, based on historical data on user behavior and the result of previous push notifications
‍

We choose the perfect time
‍We choose the best time of day to send a push notification based on the user’s activity and their reaction to previous notifications.

We have 2 types of push notification metrics:

‍

Mailing metrics
Sent, opened, CTR broken down by campaigns, segments, creative approaches and languages.

‍

Product metrics
Retention, cohort, incremental effect on user events, where conversion events usually occur: in-app purchases, ad-impressions, ad-clicks, subscriptions, conversion funnel. We monitor product metrics and have a control group which ensures an incremental effect from nGrow.

The control group is a group of users to which we do not send mailouts (it is set by default at 10%).

We randomly select a sample from all users so that, according to the metrics we measure, its size is statistically significant and reliable.

We carry out AA tests to ensure that the results measured are representative.

Our management of the frequency of notifications is transparent.

‍

‍Global frequency
You can specify a global frequency that will affect the whole app. For example, you can specify a global frequency if you don’t want your users to receive more than 2 push notifications a week.

‍

Individual frequency within a segment
You can set up an individual notification frequency within each segment. For example, push notifications can be sent to inactive users more often, and less often to active users, as they are already active.

We calculate standard cohort retention. We calculate weekly Retention in our console. In custom reports, we calculate daily Retention, among other things, and compare the results of the control and experimental groups.