Chatbot: a unique and helpful retailers’ tool that saves 150 000$

I’m Igor Zaitsev. Today I represent a company that allows clients to reduce monthly marketing costs up to 150 000$. And the Telegram Chatbot we designed and produced is the only tool that helps to achieve such rates.

Once an international retailer being one of our clients ordered a Telegram Chatbot development. As aresult,it allowed this retailer to save significantly due to an automated and free mailing system via Telegram targeted on its customers.

If compared, the same SMS mailing average per-user cost is about 0,07$ (text + an image). Moreover, our Telegram Chatbot considers user’s/clients’ hierarchy. This means that all the sent info is truly needed by recipients. Traditional channels do not offer such an option.

At the same time, today this very Chatbot is used by more than 40 thousand users daily for several mailings at once. This way the company saves more than 150 000$ per month.

On how we managed to achieve such results

Stage №1. Development stage and our first users

In 2016 one of our clients ordered a Chatbot development. The following features and options were a must:

● Newsletters;

● Polling;

● NewssortingbyCategory;

●Transitions analysis and news categories statistics filtering.

Right after the development process was finished all the info related to this bot instantly spread through social networks and other clients’ communication channels. Just a couple of months and our bot has drawn attention and involved several thousand users.

Stage №2: Additional options’development

In 2017 we held a massive Event that promoted a branding pics and images option integrated into our Chatbot that doubled its target audience in a couple of days.

All the participants uploaded pics taken during the Eventwhile out bot put down additional layers covered with branded elements and sent a new image back. People shared those with their friends, colleagues, and the bot that helped this miracle come true itself.

Stage №3. Integrating the ability to view Bonuses

In 2017 we held a massive Event that promoted a branding pics and images option integrated into our Chatbot that doubled its target audience in a couple of days.

All the participants uploaded pics taken during the Eventwhile out bot put down additional layers covered with branded elements and sent a new image back. People shared those with their friends, colleagues, and the bot that helped this miracle come true itself.

Stage №4. Smart replies: collecting info

Throughout 2019 we collected users’ requests not related to tech support that allowed us to generate a dataset needed for further machine learning. At the same time, customers generated content needed for a feedback in each request category.

Stage №5: «Smart helper»:development

In 2020 we started to train our Chatbot through machine learning. Also, acollected dataset and neural networks able to process requests were used. Moreover, we implemented «YandexSpeechKit» — speech recognition technology able to send bot voice messages and receive relevant replies. As a result, the bot’s daily target audience reached 500–1500 users since «Smart helper» was announced.

Conclusion

Chatbot sare of great help to modern companies. Saving on marketing costs is not their only one advantage. They allow increasing average check and clients’ loyalty. And this very case proves it perfectly.