All posts by xhtmlchop

APPTENTIVE

FAN SIGNALS – NEW INNOVATION



Could Fan Signals disrupt the NPS SaaS industry?
The ROI from this unique data and customer feedback proved to be more valuable to understanding customer lifetime value (LTV), customer retention, and customer churn prediction modeling than Net Promotor Score (NPS).

Fan Signals: Dashboard
The dashboard represents a plot chart showing customer segments through loyalty. Shifting loyalty is broken down into sentiment segments known as Fan Signals. At a high level, there are three types of customers: New Customers (CPA), Existing Customers (frequent users), and Customers who Churned (App abandonment). The business value is to connect with customers before they Churn and offer incentives (e.g., promotions, discounts, targeted ads) to bring back App loyalty.

Mike Saffitz – CTO
Quote: “Geoff reported to me (both directly and then indirectly) as part of our product team at Apptentive. He was a very talented UX designer, passionate about addressing users needs and eager to iterate amongst a variety of approaches to identify the best way to solve a particular problem. I was really impressed with Geoff’s energy and creativity– he had an infectious love of the creative build process.”

Fan Signals – Early Data Modeling
From inception to productization, the team shipped Fan Signals to all customers as a feature flipper. The POC from sales interviews received positive feedback as a new up-sell revenue stream. The left diagram shows a redash.io dashboard showing customer data before the font end work started. The insights began to show and classify consumers into segment cohorts known as Fan Signals, in which New Fans broke down, Switched Fans, and Repeat Fans, all of which correlated to customer behavior. Using AWS AI/ML to augment the feedback data was vital to feature success.

Business Results – Proprietary metrics
The tailored reports were designed to reveal more profound insights into how customers feel about the brand/product. When they reached the C-suite, they received tremendous praise. With continuous agile improvements, the 20+ page reports closed $100K+ deals.

Solution — I partnered with a data scientist to create a flexible template tailored for any customer. This unique solution was created in PowerPoint (instead of Excel or Word), and the document can be linked to an Excel data connector while keeping the visualization.

Executive Reports – At scale, the tailored reports were designed to reveal more profound insights into customers’ feelings about the brand/product before productization.

Fan Signals – MVP
To get customer feedback, I shipped the first public view for Fan Signals as proof of concept. The green represents missed opportunities, or from a customer’s perspective, the business could be losing customers or money. The purple represents new, loyal, and repeat customers to ensure brand/product success; the purple symbolizes royalty. The customers wanted to know deeper insights into when a customer shifted loyalty or sentiment towards the business.

Customer Interview
I spearheaded the first customer interview to validate the value and gauge the customer’s interest by aggregating all the feedback types. I conceptualized and iterated on various histogram chart designs and landed on a Pos. / Neg. The bar chart best represented the binary question “Do you love our brand?” (Yes/No), which customers intuitively understood. The design helped to visually solve the volume of responses and how the feedback was trending. The prototype and the number of positive customer feedback validated the POC.

The Challenge
This is where my design journey begins – working with Apptentive’s Love Dialog flagship product as the Lead UX Designer. Customers wanted to see the underlying data contributing to the responses received, represented in a single bar chart. It was the start of an innovation hypothesis that the business was willing to invest in understanding the deeper metrics.

The Beginning –
The Love Dialog started with a simple question: “Do you love our brand?” – a query that generated a 90% response rate. Based on their answers, consumers are redirected to another type of feedback interaction (survey, message center, rating R&R, etc.), which captures even more VOC feedback at scale.



Segment Builder (pt2.): Customer Targeting

The Love Dialog did well in soliciting and capturing various feedback interactions (surveys, messages, ratings, NPS), but only in one-way communication. Most feedback SaaS platforms capture feedback and never return feedback to the customers. At scale, customers would like to know that their VOC was heard.

Loyalty Builder
The product goal was to close the feedback loop. At a high level, the system is like an Entity Management, Crud system, and Query builder that allows customers to re-engage with them through automation. In customer interviews, customers told us they wanted the ability to reconnect with their customers through some interaction, promotion, or advertisement. The Game changer…

Audience Segments
The platform’s real power was identifying a Customer ID and advertiser ID and augmenting the data supplied by the customer UUID. Customers can send custom key-value pairs, Zip Codes, DMAs, Devices, Locations, memberships, loyalty, and other metadata. Then, Fan Signals augments the data and builds segment cohorts for the most accurate re-targeting.

UX/UI functionality Spec
Part of my process is documenting high-level UI functionality from a designer’s POV, covering scenarios, use cases, and high-level functionality requirements. The working doc covers a variety of product scenarios to help uncover or determine any unforeseen edge cases.

Domain Diagram
I worked closely with a systems architect to understand the system’s design; then, I created the UX/UI/IA visualization for an intuitive customer journey. The art of the discovery process is creating a wiring diagram to help visualize the data before the UX process begins—the making of Fan Signals.

BITTITAN

AI/ML ACTIVITY FEED / ENTITY MANAGEMENT



I lead three key features that are closely connected to enable customers to migrate email boxes: The key features include Device Management Agent, Entity Management, and Activity Feed. Each feature serves a specific purpose. We started with the Device Management Agent, which captures customers’ profiles and logs. We then moved to Entity Management, which provisions the end customers’ computer systems. Finally, we have the Activity Feed, which analyzes customers’ systems and utilizes AI for recommendations.

Once the customers’ data was analyzed, it was processed with ML technology to identify security issues, threats, sensitive data, compliance, and viruses. The exciting part was experimenting with cutting-edge AI/ML technologies in 2017.

The CEO’s vision was to remotely monitor customers’ systems and alert system administrators to resolve any issues that may arise. The core of the MSP Complete system was to make it easy for IT teams to deploy and manage remote automated services. The press had early access to the platform. The press had early access to the platform. View

Activity Feed (AI/ML): Using AI, the system would be a continuous proactive/reactive notification system. The customer value would be the first responder for IT professionals who could deploy a remote solution, before real threats occur.

Customer Journey Mapping – My process was designed to maximize communication with developers. I created comprehensive UI journey maps to guide developers through the feature development process. This approach helps to accelerate collaboration.

Customer Feedback – The MVP was rough by necessity, but adaptable, allowing the team to focus on core functionality features. This is agile thinking — understanding that you can’t build everything upfront, or we would never ship. Rapid development helped identify opportunities for feature improvements, which were then confirmed through customer feedback.


Discovery Phase – Collaborating with the team at a high level, quickly wiring up user flows that align development and product to understand requirements and feature initiatives.

Rough Concept Iterations – Through multiple whiteboard sessions, the dev team was able to gain a comprehensive view, which helped synthesize the product features. Through collaboration, the team understood the fundamental requirements before creating Jira stories.

Diagram Flow – I created a variety of high-level Activity Feed wire diagrams to help the team understand the data flow between different systems.



Entity Management – Users and DevicesUsers/Devices

Overview – As the lead designer, one of the key features of the managed service platform is the ability to remotely view customer systems. This enables full use of the platform’s capabilities. After you add an entity (user or computer), new services will appear in the activity feed.

Data Insight Views – IT customers can quickly see user metadata details to determine the right actions for a device and access other relevant data. Customer feedback was positive because they didn’t need to drill down further to find all the important data.



Customer Profile – Once the DMA is successfully installed on a customer’s device, the IT team can perform basic tasks on the customer’s computer. The goal was to support 100 services in the background, monitoring the health of the customer’s device. I added different panels to prevent the customer from navigating to another page.



Task Launcher – Full Modal Experience
Once customers choose a path to migrate their users, the task launcher opens in a modal experience, showing the simple steps to complete the required tasks. An example of BitTitan’s target market is universities with 40k+ users who need to migrate from Office 365 to Gmail or vice versa. The Overview is the first step, which provides clear customer guidance and a video tutorial showing how easy it is to deploy the DMA agent to all its students.



Device Management Agent (Empty State): Since I designed the deployment agent (DMA) application, creating the user invite workflow for setting up and deploying the agent from the platform is essential. The empty state lets customers choose how the DMA agent is installed on their systems.





The Beginning – Overview
A Device Management Agent (DMA) is the quickest and easiest way to remotely deploy and configure software applications on end users’ desktops or mobile devices without IT physically touching the computer. DMA helps companies of all sizes deploy new software applications remotely and can migrate mailboxes to the cloud with a few simple user tasks.

Results — The redesign increased the customer success rate by around 80%. The redesign improved customer satisfaction, and the animation reduced the number of users closing the app.

HEARTBEAT – Animation
I thought it would be a nice touch to add a little bit of eye candy (.gif animation is only supported) to let the user know the app is progressing in the configuration process. Deploying services on customers’ systems takes time. I worked with a dev (1:1) and we pushed the design by adding the animation and overhauling the entire app.

The Design Challenge
Once installed, the Agent App automatically installs services, or users can install them remotely. The App was developed as a universal shell app (C++) compatible with any OS. I was tasked with redesigning the client app to improve the customer journey. I tried to push the design, but the app’s technology limitations limited me.

Deployment Pro Agent (DMA):  Before – After
The original Deployment Pro remote client agent looked like malware, with a high customer drop-off rate because the customer thought they had installed malware.