Hellobot
An innovative bot solution based on voice technologies to optimize any kind of business.Details
- Scope Branding, design, development, QA
- Innovation RASA algorithms
- Technologies .NET, C++
- Business sector Various branches
- iteo Team 1 x BA
1 x UX designer
4 x backend developers
1 x QA specialist
2 x POs
1 x UI designer
Timeline
Product Workshop
Bot creation process
POC
Project’s highlights
Hellobot can be a virtual consultant, an advanced answering machine or a digital call center employee. After integration with a client’s database it’s able to collect, update, and analyze all real-time information.
Our study includes a case of a Lellek Group company – the largest authorised dealer of automotive brands such as Porsche, Audi, Volkswagen, Skoda, and Seat in Silesia and Opole Voivodeship
Business needs & goals
The client wanted to automate processes connected with contacting their customers via phone channel. Implemented alterations were supposed to reduce the costs related to the man-hours lost by a car service when its clients didn’t show up at a prearranged time. The overall project came down to creating a voicebot to optimize and automate three phone channel based processes formerly fulfilled by the employees of car showrooms and services.
They included:
- Confirming appointed service visits
- Informing about a new car purchased in a showroom being ready to collect
- Customers’ satisfaction surveys concerning customer support in showrooms and services
All these processes were provided by assistants and advisers of Skoda Lellek showroom – the procedures were labour-intensive and nearly fully manual. Thanks to using a voicebot and shi!ing to automation, employees could dedicate their time to customers who visit the showroom in person.
Target audience
We analyzed existing Lellek Group’s customers together with the
client and defined two main groups of target audience for Hellobot.
The first group includes customers and businesses who own cars purchased in one of the showrooms and use one of Lellek Group’s services. Whereas the second group consists of regional employees exploiting fleet cars bought in one of Lellek Group’s showrooms and companies using one of its services.
Functionality
Voicebot’s key performance indicators for Lellek Group were:
- Reminding a customer of a prearranged visit, collecting a declaration of arriving on a set date, accepting a directive of canceling a visit or fixing a new date, and eliminating service’s lost man-hours.
- Vehicle’s status in the service (incoming calls) including information about a car being ready to collect (outgoing calls), and a number of informed customers. The amount of customer’s declarations concerning the collection date and passing information that a car is ready to collect were an additional indicator.
- Customer’s satisfaction questionnaire – getting answers for all the questions in the survey.
Technologies
The project was based on the .NET framework.
We used it due to its high efficiency and great performance.
The technology is supported by Microso” which guarantees
constant updates and fixes. What’s most important, it allows a plenitude of possibilities – whatever you’ll up to, it’ll help you achieve your goal. In addition to its advanced solutions, it’s also reliable, stable, and lets you work faster and more productively.
Model of collaboration
Together we created a unique value proposal and a project vision based on Lellek Group’s mission.
Our collaboration started with workshops that lasted two days. We performed a complex analysis of the company, paying special attention to the possibility of optimization and automation of all its processes.
Next step was fitting the solution with a well defined target audience and its needs. We also pinned down the bot’s personality and conversation scenarios for each of the automated processes and target groups.
After the processes were written down and mapped, the client chose three of them which could be operated by a voicebot. At this point we’ve also determined additional processes dedicated for automation, KPIs, and expected benefits.
Innovation
Thanks to artificial intelligence algorithms (RASA), the bot constantly learns and becomes even more human-like.
It’s able to identify a client’s intention while holding a conversation. Additionally, if it calls a client and there’s no answer, it calls again in a defined time interval. A client can call back, too.
To let the client keep up with all the processes performed by Hellobot, it sends daily email reports with a status of realized calls and details about every record from data basis to an employee responsible for the process in the company.
What’s more, we frequently update its knowledge base, build and complement phrases concerning Lellek’s clients’ responses. We are also able to modify a conversation’s parameters such as: questions order, speed of talking, conversation scenarios depending on obtained answers, calls schedule, and a manner of handling incomprehensible phrases. Such approach allows us to increase the efficiency of a conversation and maximize the benefits of using a voicebot in Lellek company
Challenges & solutions
On the grounds of performed workshops, we prepared a POC (Proof of Concept) project defining a range of a voicebot MVP functionality.
Chosen business process had to comply with a
number of criteria – provide the client with a business value, confirm effectiveness of the system and agility of its implementation, and reduce the necessity of integration with client’s systems.
We used a predefined Hellobot module which was adjusted to Lellek Group’s individual requirements and necessities.
Matching the needs of POC, the bot was implemented in a Bot as Service model, assuring complete infrastructure on Hellobot’s side and requiring only one integration – with the client’s phone number. In the future we plan an integration with the dealer’s CRM, an end-to-end automation, and an automatic edit of the service calendar.
Quality assurance
We’ve performed a series of tests to enhance and optimize the voicebot solution.
The first phase included internal testing conducted before the implementation. Its purpose was to work out the best scenario fulfilling all the assumptions and to check if the bot understands it well.
Then, we moved on to the second phase of tests – this time along with the client. The process involved some cosmetic changes optimizing the dialogues in order to increase customer’s satisfaction and reduce the amount of unsuccessful conversations.
Outcome
First tests conducted in one of the company’s locations prove that such a solution is effective and has a positive influence on the brand’s image and clients’ satisfaction.
Main noticeable business advantages for Lellek Group are:
- Positive opinions of Skoda clients who had an opportunity to talk with the voicebot
- Limiting the number of lost calls and potentially lost clients
- Increasing the amount of time spent by the employees on more innovative and less repeatable tasks
- More time for face-to-face contacts with clients visiting Lellek Group’s showrooms
- Promoting Lellek Group and Skoda brand as pioneers of innovation on the car dealers and !SO services market
- Ensuring consistency and accurateness of confirming, informing, and surveying in Lellek Group
- Optimizing the processes and increasing the availability of services which influences clients’ satisfaction and results in easier, more efficient communication