Securitas Technology · Client project · 2022

Secu-Rita

A conversational support assistant that lets Securitas Technology’s customers handle common requests themselves, instead of waiting 15 to 20 minutes on the phone. Our research and recommendations shaped the next update of Securitas’s support chatbot.

Type

UX research, conversation design, usability testing

Duration

9–10 months (2022)

Team

Senior digital marketing manager, UX product manager and 3 UX designers, including me

Three phones showing Secu-Rita conversations: the welcome options, suggested locations with a date picker, and a ticket status

Overview

About Securitas Technology

Securitas Technology provides access control, remote monitoring and intrusion systems for commercial and residential spaces. Its most-used management app, HQ, lets security managers monitor their buildings remotely, check system status and respond to notifications.

My role

I worked across the project: planning the research, the literature review and competitive analysis, user interviews, facilitating affinity mapping and analyzing CallRail call data, then user and chatbot flows, prototyping, interface design and usability testing.

The challenge

After the onset of Covid-19, support calls to Securitas surged. Customers waited longer, more of them hung up, and the customer care team was overloaded. Securitas already ran a support chatbot, but customers still picked up the phone for most of their questions.

600%increase in customer support calls
15–20 minaverage wait for a response
20%of calls abandoned
4 of 6interviewees preferred a chatbot

The brief

“After the pandemic, our customer service experience suffered from the surge in support calls and a shortage of staff. We want to give customers a quick, hassle-free alternative.”

Meet Secu-Rita

A conversational assistant that is quick, human and helpful, built around the tasks customers called about most.

Secu-Rita greeting a customer and offering help

Human conversations, not robotic replies

Short, precise replies with a human touch, written to acknowledge the customer’s problem before solving it.

Secu-Rita recognizing what the customer needs

Learns what customers need

The more questions it handles, the better it recognizes what a customer is asking and where to send them.

Secu-Rita confirming before taking an action

Confirms before it acts

Before carrying out any task, Rita checks with the customer, which builds trust and catches mistakes early.

Secu-Rita handling a request without a phone call

Skip the support line

Common tasks like putting a system in test mode, creating a service ticket or checking a ticket’s status happen right in the chat, with no wait.

How the solution balances business goals and customer needs

How the solution balances business goals and customer needs

How we arrived at the solution

The process looks straightforward, but each method was chosen to work within real constraints and ambiguity, balancing business goals with user needs.

Our process from research to testing

Research planning

Setting the research goals

We used a UX project canvas to plan the research with Securitas and agree on three aims:

The UX project canvas used to plan the research

We kept the project canvas updated with Securitas throughout the project

Research planning details

Domain analysis

Who are the users, and how do they manage their security systems?

The users are security managers, safety and compliance managers, and admins who monitor large systems: fire alarms, burglar alarms and security cameras, often across many locations. They do this in HQ, Securitas’s app for monitoring, checking status and testing systems.

Call data

Why customers were calling

CallRail data showed the keywords customers called about most. We listened to recordings of customer calls to validate those keywords and understand the context behind them.

CallRail data showing the most common call topics

User interviews

Talking to customers and stakeholders

6customer interviews
2stakeholder interviews
User interview overview

Analysis

What did the users say?

As a team, we coded the interviews into themes, then used affinity mapping to turn quotes and observations into insights about users’ frustrations, needs and goals.

Affinity mapping of interview dataThemes from the interview analysis

What insights did we distill?

Key insights from the research

Define: two scenarios to focus on

The interviews and call recordings pointed to two scenarios that drove a large share of calls and were well suited to a chatbot. We built a persona and start-to-finish user flows around them.

Test mode

Putting a security system into test mode, a routine task customers regularly had to call in for.

Service dispatch

Creating a service ticket and checking its status, without waiting on the phone for an agent.

Persona

Age

40

Education

Security engineering

Occupation

Security analyst

Location

Seattle

Company

SecureUs

“I strive to make sure my company is fully secure, and we trust Securitas with that service.”

Core needs

  • Quick and hassle-free resolution of issues
  • Connecting to the right contact or customer service team for a specific problem
  • Being up to date on the status of issues through alerts, texts or the app

Frustrations

  • Very high wait time on the 1800 number
  • Limited alternatives for customer service
  • The current chatbot isn’t comprehensive and doesn’t provide actionable solutions
  • No easy access to information needed during the support call

Prototype and test

We built a working conversational prototype of both flows in VoiceFlow, so people could have a real conversation with the bot rather than click through screens.

Test mode

V

Hi! Welcome to Securitas 👋
How can I help you today?

Put my system in test mode for 3 hours

V

Please enter location at which you want to start test mode.

Texas

V

Putting system at Texas on test for 3 hours.
Thank you! Have a great day. How helpful was this experience for you?

😞😐🙂😀
Type a message...➤

The bot needs a duration and location to put a system on test, asks for anything missing, then acts and asks for feedback.

Service dispatch

Service Dispatch at Texas

V

Do you have an existing ticket?

Yes

V

Please enter the service ticket number

1827EKs3

V

What do you want to do next?

Increase Priority on my ticketCheck status of my ticket

Check status of my ticket

V

Your ticket is due to be serviced next week on 11/01

Type a message...➤

The bot asks about an existing ticket, then lets the user raise its priority or check its status. Creating a new ticket needs a passcode.

Think-aloud testing

We ran think-aloud sessions with 5 internal Securitas users, asking how the bot’s language came across and whether each conversation flowed as they expected. Each person completed three tasks:

1. Put a system in test mode

2. Check a ticket’s status

3. Create a new ticket

What we changed after testing

We built the final version in Figma, which also let us handle edge cases that VoiceFlow couldn’t.

The final prototype

The final experience, refined after testing and built in Figma, with each screen showing one design decision.

Hi! Welcome to Securitas Technology 👋

Here are few things I can help you with

Put system on testCreate a new service ticketCheck status of an existing ticketOther
Type something➤

Descriptive / Predefined Input Method

Hi! Welcome to Securitas 👋

How may I help you today?

Put system on test

Sure, which of the following locations would you like to choose?

#1132 University Boulevard, Indianapolis#112 Lockefield Ct, Indianapolis#113 Navy Pier, downtown, Indianapolis
Indianapolis|

Auto Suggest Location from HQ Account

Put system on test

Sure, which of the following locations would you like to choose?

#1132 University Boulevard, Indianapolis

On what date would you like to put your system on test at #1132 University Boulevard, Indianapolis?

Date11/11/2022 📅

At what time do you want to start this test on the 11th of November?

02 : 00 PM05 : 00 PM

• • •

Type something

In App Widgets for Date and Time

Putting #1132 University Boulevard, Indianapolis on test mode from 2PM to 5PM on 11th November 2022.

Confirm

Test Mode will be activated as per your request. I will be sending you an email confirmation shortly.

Is there anything else I can help you with?

No

Do you think I did a good job helping you today?

👎

Sorry to hear that. What could I have done better?

I think it could be more interactive|

Taking Negative Responses into Account

How may I help you today?

YesNo

I can help you with that! Please select your problem category.

System software malfunction

Got it! What information would you like to provide to the service technician regarding your issue?

My alarm keeps alerting for fire at all times.

I have made a note of that. Is there any other detail you need to add?

Type something

Empathetic & Appropriate Affirmations to Enable Trust

My alarm keeps alerting for fire at all times.

Sorry to hear that. Is there any other detail you need to add?

No

Alright, which of your locations is facing this issue?

#1132 University Boulevard, Indianapolis

You and the on-site contact for this location, Jay Stuart, will be contacted for updates related to this ticket.

Is there anyone else you would like to add for this ticket?

YesNo
Type something

Smart Bot Responses based on Existing Knowledge

Hi! Welcome to Securitas 👋

How may I help you today?

Check status of an existing ticket

Can you tell me your ticket ID?
You may find it in the confirmation email you received after raising the ticket.

TIX20063

Ticket Status: Assigned to technician.
Ticket Category: System Software Malfunction.
Technician Name: Alex Russo.
Date of Ticket Creation: 11/27/22.

You will be contacted by the assigned technician for further assistance.

If you need urgent help with this ticket, please call us at 1800-xxx-xxx.

Type something

Detailed Information in Response to User Queries

Alright, which of your locations is facing this issue?

#1132 University Boulevard, Indianapolis

You and the on-site contact for this location, Jay Stuart, will be contacted for updates related to this ticket.

Is there anyone else you would like to add for this ticket?

No

I am raising a ticket at #1132 University Boulevard, Indianapolis for a System Software Malfunction. Is there any additional information you need to provide?

YesNo
Type something

Confirmation Prior to Finalizing Bot Action

Outcome

Impact

Our research and recommendations shaped the next update of Securitas’s support chatbot

Recognition

Won the Best Research and Design Project Award

What we handed over

Why customers call

The top call reasons, from CallRail data and recorded calls

Interview insights

Findings from 6 customer and 2 stakeholder interviews

Tested conversation flows

Validated flows for test mode and service dispatch

Client feedback

“

The research was good validation for what we had already collected. This work and the data are a great foundation to base our future conversational projects on. Great job!

CN

Connor Nash

Stakeholder and Design Manager, Securitas Technology

“

Working with your team through this project has been a wonderful experience. You really came through with the final deliverables, and I’m proud of your work.

OW

Olivier Willems

Project Manager

KEEP EXPLORING

More Projects

Here are a few more of my case studies.