VTS provides a platform for brokers and lease managers to help them with managing commercial real estate properties by streamlining processes and centralizing data. It enables landlords and brokers to collaborate more efficiently and make more informed decisions, allowing them to close deals for commercial real estate leases with more information and increased efficiency. Our stakeholders wanted to explore different ways they could surface this critical information to their users using conventional interfaces, with a hypothesis that the lower friction would help their executive

VTS was looking to host a live demo of their new voice skill at their annual conference, Accelerate. This gave us a hard deadline and heightened the need for a working solution in a short time frame of 5 weeks. For the first two weeks, the team consisted of myself as research and conversation designer and a product manager, and in the last three weeks, we had two software developers join us.

To help us understand VTS's product offering our team members all made VTS accounts and spent our first week exploring the platform. What we found was an extensive service offering of digital marketing, online leasing, property and tenant management, and data analytics and market insights. We also spend this first-week conducting stakeholder interviews.

One of the customer segments that our stakeholders defined for us were people who were either higher-level managers or executives of CRE brokerages. They weren't engaging with the platform. and in general, across all segments, VTS was interested in if a voice interface could help them increase engagement.

The VTS team provided deep end-user knowledge, which was supplemented with our 1:1 interviews with customers. In the following week I used these insights to frame several opportunity areas and using these as prompts, I facilitated an ideation session with our team to generate dozens of high-level concepts — after some synthesis and evaluation, I prioritized these into the 4 top concepts that I matured further and shared with our stakeholders.

After receiving feedback from our stakeholders and we converged on two of our concepts for further validation testing. This included sample dialogue writing, wizard of oz prototyping and lightweight user validation testing. These tests were done over a conference call with our interview participants from our earlier research, a small group of VTS users who fit our management/executive customer segment. We synthesized this early feedback and used it to help us define the requirements for our MVP—something we could release in a controlled alpha with a select group of VTS customers.

The two concepts that we developed into Alexa skills both provide different business insights we had identified through our research to be of most importance and most impactful for VTS customers. They are separate Skills at the moment to help us simplify the flows and also to help us evaluate them separately, we think that in a future Beta release we might join them into a single product, but a small suite of related skills is also interesting to consider.

To express our agent's executive assistant personality I used Amazon's speech style SSML tags. The styles use a set of text-to-speech voices that use different variations of intonation, emphasis, pausing, and other techniques to match the speech to the content.

The following is an example of the UX flow for a customer using VTS Analyst to help them review their high-priority expiring leases.

Since the initial end goal of delivering a working demo for Accelerate 2020 was no longer relevant, VTS decided to extend the project from 5 to 7 weeks based on the opportunity and time the team could take to polish the VTS Analyst feature and build out an MVP for another concept called VTS Tenant Analyst.

After the decision was made to go forward with building out two skills, I began in-depth Conversational User Interface (CUI) design and usability testing while an engineering pair focused on fundamentals, such as account linking, followed by core skill implementation once the CUI design had been finalized.

Despite a tight deadline, we were able to learn a lot very quickly, generate many novel solutions for customers, and deliver two Alexa skills, all within 7 weeks. These skills have been released to 26 high-value accounts to help them out during the market uncertainty brought on by the COVID-19 pandemic. VTS plans on using the data they acquire around product desirability to inform them of the next steps on where they should focus their feature build-out efforts in the future.

VTS has used the two skills as a valuable tool in their sales pipeline, utilizing it for live demos with new prospects. Following these demos, they have sent out VTS-branded Echo Dots to prospects enabled with work-from-home kits. Bala Pitchandi, SVP of Engineering at VTS, said,

We’ve since learned that our approach to the extraction of data and business insights and communication of these through natural language generation has actually informed the core VTS dashboard product, with business insights now being surfaced within the main application using natural language.