Veritas Hybrid Cloud Solution: Sizing Framework
Designing web-based experience of sizing framework solutions to fully integrate Veritas multi-cloud and appliance products
OVERVIEW
This project aims to design a web-based sizing framework that is capable of efficiently and accurately sizing Veritas' solutions, in order to support hybrid cloud environments and multiple appliance family, as well as both internal and partner adoption.
My responsibilities include:
1) leading a small but cross-functional team to land on decisions;
2) exploring workflows for each technical section, and identify attributes in proper ways;
3) defining design functionalities by acknowledging technical constraints.
COLLABORATORS
1 PM, 1 front-end engineer, 6 back-end engineers, 2 technical writers
RESPONSIBILITIES
Only designer 1) primary & secondary research, 2) user interviews/testing, 3) concept mapping, 4) low to high-fi design
TOOLS
Figma, Balsamiq, Jira, Confluence
DURATION
March ~ August 2023 (to continue)
Background
Veritas' sizing framework solutions, aka sizing tool or planning tool, is one of the most crucial method for our appliances to achieve full solution integration and hybrid cloud support. It is currently in an Excel sheet format, which is less interactive, and unscalable. In the meantime, Veritas is the only vendor of many loyal and long-term customers who rely heavily on using our appliances but were not satisfied with existing interfaces. We don't want to lose our customers, of course. As traditional data analytics grow outdated, we want to design an extensible, scalable and flexible framework solution that enable both of our internal sales engineers and eternal customers to adopt and analyze the data for appliance recommendation.
Outcome
Introducing web-based sizing tool for Veritas appliances:
Veritas' new sizing framework (aka appliance planning tool) is a web-based product meant for direct accessibility, usability, and customized data analysis output. It supports a variety of functions in its step-by-step input, from detailed data and project information profiling, to transformations including table expansion, filtering values and etc.
The insights from sizing framework solutions can be extracted and adopted within our internal sales engineering teams and customer groups, aiming to improve workflows to be more adoptive, flexible, and intuitive.
Research
With the focus on making the new sizing tool more accessible and intuitive for internal and external customers, I closely collaborated with a technical product manager to create a research plan to document our goals and assumptions. Here are the angles we investigated: primary research interviews (6 rounds with internal users), current spreadsheet auditing, and reviewing pain points from customers on existing Spreadsheet experience.
Pains & Insights
Our team collected all the notes and synthesized into four pains and insights at this phase:
Longer running time
Users had to manually update the product whenever there's a new version, and it always takes longer then 15 seconds to proceed the running for performance
Dataflows lack efficiency
It takes users "forever" to locate relevant data values from the columns because it has lots of metrics. They need a better way to dynamically use the sizing tool to determine data quality
No error fixing
It's easy to make mistakes in spreadsheet columns, and users found it frustrating because they cannot easily fix the error inputs due to typo or other factors
Lack comprehensive guidance
Users ask, what kind of information is real, crucial, required in the sizing tool, and what if a customer doesn't have a basic knowledge toward it?
Current workflow
Ideation
Based on the initial insight direction, we began brainstorming concepts for how we could display enormous, abstract and complicated information while maintaining user understanding of core actions. With some starting points, we review the concepts with both internal and external users to iterate on functionality, feasibility and feedback.
Design challenge 1
- displaying table
The first challenge we confronted was the sizing tool table in Excel looks too messed and is not discoverable. To maintain its functionality, we designed a new table that is expandable and with distinctive columns, and replace buttons with drop-down design for users to select the best option under each attribute. After several iterations and user validation, we ensured that simplicity and visual consistency were among our core design principles.
Design challenge 2
- re-group appliance information in Storage step
Another challenge we faced was the table columns were too inefficient. Users complained about making mistakes. Instead of showing all appliance attributes in a single table, we decided to re-order and synthesize them into five appliance columns, for each it includes multiple selections in the drop-down menu. We also redesign the entry point for Safety Consideration and its relationship with Workloads, and users can access it easily.
Design challenge 3
- interpreting image policy details
Image policy details are complex to interpret and differentiate, because this step contains lots of data information and it requires accurate hierarchy. After discussing and user testing with users, we decided to design two table, one leaderboard, one secondary table to divide the attributes based on priorities.
Solution
Key persona
We created a representative persona modeled after interviewing several users. The web-based sizing framework tool helps Michael accomplish his goal of sorting out information and relationship of key attributes from datasets, and running analysis for more accurate recommendation results.
Setting up basic information
Michael has pre-populated information about the sizing project he has done in the past with traditional Excel product, and his next step is to begin with a new project using the web sizing framework. Cloning from existing project is an optional step since his customer is different from previous ones.
Defining storage location
After entering basic information of his new sizing project, the stepper moves to define the location for Customer 1's Project 1. In this new data table, he could take a deeper look at the details of the attributes, including storage location types, cloud provider, local/WAN/Cloud/Storage network, and with an option to add a new storage location to meet with complex information.
Defining storage targets
Similar to defining storage location, Michael is able to set up the storage targets based on his entries of storage locations from the previous step. What to note, safety consideration is an important feature under this step, so Michael is able to find out the details under the kebab menu.
Adding copies for image policy
Satisfied with his data, Michael navigates to this step where he can add up to two copies for each image policy. By default, we have provided an initial image data for Michael, but he can choose to review the details for this image policy by clicking "view details" under the kebab menu.
Review workloads and produce sizing output
Michael may review the workloads data where dedupe rates are located. In this step, each data is calculated based on his previous entries and the existing data for the customer. He is able to return to any previous step by clicking the stepper, yet he has to go over only one step after one, until the system running the output.
Iteration
Page vs. Modal
We decided to use page rather than modal design for the ultimate creation flow because it allows user to do distinct actions by adding rows directly into the table, and therefore it's less disruptive and can provide more space to accommodate more steps in a big sizing use case.
What does success look like?
01
Increased adoption and consensus among Veritas sales engineers, lead and architects in their appliance deal.
02
Neutral or greater top-line metrics on performance and recommendation results for the new sizing tool.
To continue 🚀
Our new web-based sizing framework, originally serving as a rocket to support Veritas appliances, now has an extremely exciting opportunity joint our flagship appliance, NetInsights Console to integrate with our SaaS product, Alta™ cloud platform in order to optimize application performance, availability and resiliency.
Our next step, in which we are working on, is to explore AI-driven design opportunities for embedding sizing framework into System Health Insights to monitor systems health and detects potential issues, and creating proactive remediation.