Anagog’s on-device machine learning models identify correlations between many insights and the likelihood of user churn. This information can help to engage ‘at-risk’ users in advance with appropriate, churn-prevention campaigns. And since churn-risk users are not all the same, each micro-segment can be engaged with the appropriate content, offer and message that will keep them active and loyal.
Yes. The first 24-hours after installation are when many apps experience a user drop-off, meaning that all the budget spent on reaching out and getting users to download and install your app has gone to waste. Hooking users immediately is a high priority, but engaging them effectively on Day-1 can be tricky since you still don’t know enough about them. The Anagog SDK, as it operates locally on the device, generates a number of useful insights practically immediately, allowing you to create more efficient ‘Welcome’ campaigns that can generate engagement early on and activate new users.
Many apps struggle to streamline their onboarding processes, sweating over how to communicate with new users and take them through the basic initial steps such as signup, registration and payment. Using insights generated on Day-1, Anagog can help you engage users who are stuck in various stages of the onboarding process.
For example, instead of a generic “Can we help you?”, you can create different ‘Help’ messages based on the user’s segmentation. And instead of pushing messages when users are too busy, Anagog helps you engage users when they are more likely to be available. This personalized approach to onboarding shows customers that you care and helps them complete the journey they started when they downloaded your app in the first place.
Many marketing metrics are focused on Day-1/Day-30 retention since this is a reflection of the user acquisition campaign and the onboarding process. However, an even bigger resource drain is dormant users – those people you have successfully guided through acquisition to installation and retention, but whose activity level has dropped and no one was there to notice. Since Anagog’s insights are generated on the device and are not dependent on app activity, you still have plenty of insights to effectively target these users and engage them with relevant content based on who they really are.
For more reading about dormant users and how to go about re-engaging them, read here.
With targeted campaigns, you can ensure that the right engagements are sent to the right audience at the right time. The problem is that messages don’t always arrive. Push notifications are known to be a somewhat unreliable communication channel; poor network coverage and bandwidth constraints can block messages or delay them until they are no longer relevant or have timed out. On the other hand, inactivity may cause a user’s app to “hibernate” and therefore refuse incoming messages, and this is the main cause of failed delivery to dormant users.
Anagog overcomes these obstacles by triggering notifications from within the device, eliminating the necessity for push mechanisms and the possibility of falling victim to network failures, delays, or timeouts. Additionally, Anagog ensures that your app is not “put to sleep” due to user inactivity, giving marketers the opportunity to re-engage and invite users back with a personalized approach.
Practically every business wants a better understanding of its customers. That’s the core of modern marketing – creating personalized, relevant, contextual offers that grab the attention of consumers. Real-world insights can significantly support your sales and marketing micro-strategies. The type and tone of the message to be sent, the values and features to be highlighted, and the best time to send are just a few examples of essential information which is not available by accessing a user’s purchase history or app usage. Additionally, real-world insights are not reliant on the volume of business or level of engagement a user has had with your company/app, and you can apply them to improve your onboarding metrics, increase retention, optimize your discount strategy, generate upsales and prevent churn.
For more on the real-world benefits of real-world insights, read here.
Research repeatedly shows that over 70% of users expect personalized services and interactions from the brands and companies they purchase from. But consumers are also fairly consistent about their thoughts on how companies handle (or mishandle) their personal data. Surveys show that privacy and ethical data practices are at the top (over 85%) of consumers’ considerations when choosing who to trust and who to buy from. The challenge is that minimal privacy requirements are moving targets; regulators and mobile OS have changed them several times in the past few years and more changes are likely. Instead of settling for compliance (at best, a short-term target), you are better off aiming for true personal data privacy and anonymity that will give your users long-term security.
For more on how to provide Privacy with Personalization, read here.
Mobile Real-World Insights are insights into the behavior, preferences and routines of mobile users in the physical, as well as digital world, generated using the same data that is available to all apps, no permissions required.
By applying machine learning models that operate locally, on the device, this approach does not rely on any data collection, location tracking or app tracking. By adhering to the most ethical data practices, Real-World Insights actually help protect companies by ensuring they only need to use personal data in the course of conducting their business with the users.
Successful B2C business requires a clear perspective of customer needs and preferences, which can be challenging in today’s digital age where consumers are faceless, inscrutable and therefore hard to anticipate.
Consider a user who has purchased 12 cans of tuna.
Is it because they love their cats or because they have started a new fitness regime?
Should they receive a coupon for a flea collar or vitamins?
Anagog’s real-world insights overcome these challenges by providing the perspective to truly personalize the customer experience. They offer additional insights into the daily routines, interests and preferences of users beyond what you can gain from their digital transactions. Are they parents? Do they own a car? Do they exercise regularly or work long hours? These insights can help you provide a better service and a more relevant customer experience, thereby improving and strengthening your relationship with each and every customer.
Most apps have access to 1st-party data received from app usage, digital transactions and purchase behavior. In addition to using/accessing such “standard” 1st-party data, Anagog uses Edge AI to process data and generate real-world insights into who your users are when they are not on your app. Not only do these new insights provide you with a deeper perspective with which you can segment your audience, you also gain insights on inactive, casual or dormant users; those users that you used to know almost nothing about.
For more on reactivating dormant or inactive users, read here.
When it comes to engagement, context is king, and each type of engagement probably has its most suitable context. For example, if the goal of your engagement is an impulse buy, the best time to engage is when your product/service is available and in high demand. If the goal of your campaign is informative, you should reach out to your users when they have time to consume the content. The power of Anagog is that we enable your campaigns to engage each user in their most appropriate, individual context.
For more on the importance of getting the context right, read here.
Before Anagog, personalized marketing was typically performed by collecting and processing as much information as possible about a user (including personal data), and maintaining it in huge databases in public or private clouds. To create a campaign, marketers would then try to select the users who fit the target audience. However, additional hyper-personalization – such as micro-segmentation and contextualization – were based on a more limited set of insights.
By processing all data on the device using Edge AI, Anagog has revolutionized the campaign creation process. Now, marketers just have to define parameters such as the target audience, the context in which that audience should be engaged, and the campaign message. The list of campaigns and the campaign rules are then sent to the devices where the Anagog SDK evaluates whether the campaigns are applicable to the user of that device. Consequently, marketers do not need to know who the users are; it is enough that they know that the recipients of their campaign perfectly match the criteria of the target audience.
Recent OS policy changes and privacy concerns are impacting Facebook’s ability to track and profile users, reducing the accuracy of their targeting and attribution. Furthermore, marketing campaigns on Facebook are dependent on users’ Facebook activity, they fail to provide any new data on your customer base and they require a significant budget.
Using Edge AI, Anagog generates, locally on the device, specific real-world insights about each device user, allowing you to continuously create dozens of micro-campaigns targeting highly segmented audiences, irrespective of how active they are on social media. Additionally, Anagog provides you with a map of your audience’s aggregated interests, daily routines and preferences.
Micro-segments are criteria for segmenting an audience to target it with relevant messages, offers and information. In Anagog, the segments are ‘micro’ because marketers can apply multi-layered criteria, enabling hyper-personalization to reach the most appropriate audience with the most appropriate campaigns.
For more information on Micro-Segments, read here.
Micro-moments are instances of individual context, identified on the device by the Anagog SDK using Edge AI. They can be any combination of an action, an activity, a location, a timeframe, or an event and they are used to define the context for engagement. With Anagog’s Mobile Engagement platform, the level of control over a campaign’s micro-moments enables marketers to optimize the context of a campaign. This ensures that the campaign is received at the best possible moment by aiming for the ideal situation, frame-of-mind, and availability of the target audience.
You can define micro-moments at a fixed time (the campaign will trigger for the entire target audience at the same time) or a fixed context (the campaign will trigger at different times for each individual, when the context is right).
For more information on Micro-Moments, read here.
Anagog processes all data and performs all the segmentations on the device itself using Edge AI. Consequently, no data has to be collected for processing in the cloud in order to obtain a targeted, personalized, relevant user experience.
Definitely, you can make use of your Custom Attributes as part of the campaigns that you will create in the Engagement Console. These Custom Attributes include proprietary data points, events and geographical locations that are relevant to your business (which are the basis for Customer Micro-Segments and Custom Micro-Moments). In fact, some of the strongest Use Cases involve audiences that are created by combining real-world insights and Custom Micro-Segments and contexts that are composed of real-world situations and Custom Micro-Moments.
For more information on Custom Attributes, read here.
Edge AI is about placing machine learning algorithms near the data. In Anagog, this means placing them on the device or “on the Edge” of the network, and enabling local data processing. This delivers faster responses and more security; lowers data transmission volumes; eliminates server-side resources for data processing and storage; and facilitates the use of the richest possible data to generate much deeper insights. For example:
- Battery status and charging patterns can generate large quantities of data that are impractical to share with a cloud for millions of app users. Edge AI can process this data locally, for each individual user, to generate useful insights.
- Collecting and sending personal location data to a cloud for analysis is considered a potential breach of privacy that customers must explicitly approve. By locally processing location data, Anagog prevents any potential privacy risk while delivering value and relevance to the customer experience.
Yes, data processing on the device using Edge AI does slightly increase the memory footprint of the app. However, this is insignificant, because after processing, the data is not stored persistently on the device, but removed by a cyclic mechanism that deletes “old” data. Our experience is that the app size can grow by 3-4 MB for Android and 7-8 MB for iOS.
The SDK is active within the app. Based on the settings of the campaign micro-moments, the SDK may be listening for signals from the device even when the app is not in use (in the background). Sometimes, accessing of data while the app is in the background requires specific app permissions.
Our SDK can process the same data that any other SDK app can access, such as Device Properties, Device Settings, Network Connectivity and Sensor Data. Due to the power of Edge AI, we are able to process that data with complex models, on the device, in order to generate insights that other solutions cannot.
For more on how Edge AI makes insights generation on the device possible, read here.
We are constantly evolving our algorithms according to the reliability of the signals and our AI models. Consequently, for some insights, we may add or remove sensor data from the calculations, based on considerations such as value, battery efficiency, OS and HW improvements. To date, we have used various sensors, including accelerometer, network connectivity, barometer, gyro, GPS, NFC connections and more.
For more on how the Anagog SDK generates real-world insights, read here.
Anagog provides built-in support for marketers seeking to incorporate location-based services (LBS) in their mobile engagement strategy (for users who have granted permission to access their location data). Our rich, layered, proprietary database of POIs (Points-of-Interest) grants context to user locations and enables marketers to use location data in various ways, such as triggering campaigns in real-time, based on the location of the user, e.g., when near a relevant store or when arriving home. Since such campaigns are triggered locally, their real-time location is not available to any external system.
POI data is also important to characterize user interests and potential needs, so they can be matched with the right offers/campaigns. Recently, many apps are focusing on using LBS to bring together the advantages of the digital experience of the app with those of the in-store shopping experience, creating a hybrid customer experience. This includes Buy-Online-Pickup-In-Store (BOPIS), personalized deal-of-the-day and unique, relevant in-store brand experiences.
For more on seamlessly blending locations into your customers’ experiences, read here.
A data controller determines the purposes for which personal data is processed, and how the processing will be performed. So, if your company/organization decides ‘why’ and ‘how’ personal data should be processed, it is the data controller. The data processor is usually a third party external to the company/organization, processing personal data only on behalf of the controller.
Since the Anagog SDK operates locally, on the device, using Edge AI, it does not require any special app data permissions in order to function. There are a large number of insights and moments that are identified without the need for any specific permissions. In order to generate certain insights and moments, the app will have to request permission to access Motion/Activity, Bluetooth and/or network data. If you would like to create campaigns that rely on specific Locations, the relevant permissions should be requested and received.
Anagog does not collect any data. All the data processing is performed on the device and nothing reaches Anagog’s backend. In certain scenarios, subject to regulation, app owners may decide to upload some or all of the generated insights and merge them with additional information gathered from other sources. That decision is up to the app owner only, and Anagog is not privy to that data.
For more information about Anagog’s data practices, read here.
Today, the vast majority of consumers value and demand personalized experiences, so once you have earned their trust, they are highly likely to cooperate; however, there will always be outliers or users who are slow to accept. Therefore, personalization should be a transparent process. This has already been mandated by regulation in some countries and is swiftly becoming an industry standard. Your app should communicate the purpose and benefits of personalization to users and allow them to decline participation if they are not ready. Additionally, depending on the regulation in your country, you probably need to provide your users with the option to withdraw their data and continue to receive non-personalized services.
For more on the importance of transparency, read here
In addition to using the insights for marketing campaigns, this data can be shared with the host app, using dedicated APIs. The host app can configure which micro-segment data is requested from the SDK, and then make it available to other components of the MarTech stack, subject to privacy guidelines and regulations.
The numbers vary. Apart from global factors such as OS, OS version and country, a lot depends on how, when and why permissions are requested. In general, users are willing to provide brands/apps that they trust with permission to access and use their personal data when the right value proposition is presented to them. This requires you to gain their trust and successfully communicate the value proposition; and that they see an immediate benefit to providing their consent.
From our experience and research, under such circumstances, about 1/3 of users will be willing to provide an application with access to their location data at all times. This percentage can grow to almost 50% for access to location data while the app is in use.
Looking at permissions for push notifications, 84% of users are willing to grant permission to send them push notifications. It is important, however, not to abuse that permission – the first reason for opting out of notifications is when users receive too many, and the second reason is when the notifications are not relevant.
For more on obtaining mobile app permissions, read here.
For consumers, opting in or out depends on their perceived benefit and their confidence level in you/your app. A common mistake is to request users to agree to something before you have had the chance to make a good impression on them and before they have had a chance to really understand why you are asking this. In the past, many users gave automatic consent, but that has changed. Today, most consumers are aware that there is a tradeoff between what they agree to and the experience they receive. To increase your odds of convincing them, try to give them a chance to experience your app/service before making your request; communicate clearly why you are making the request, and give them the option to decide at a later stage.