Skip to main content [aditude-amp id="stickyleaderboard" targeting='{"env":"staging","page_type":"article","post_id":2057782,"post_type":"guest","post_chan":"none","tags":null,"ai":false,"category":"none","all_categories":"bots,business,entrepreneur,","session":"C"}']
Guest

How this Telegram chatbot improves business efficiency

Image Credit: Shutterstock

At ContentRobin, one of the instruments making our marketing agency more efficient is our Telegram chatbot.

Interaction with potential and existing clients is one of the most labor-intensive activities for most agencies. The success of any project depends on how well you handle requests and respond to questions that arise during the work process.

[aditude-amp id="flyingcarpet" targeting='{"env":"staging","page_type":"article","post_id":2057782,"post_type":"guest","post_chan":"none","tags":null,"ai":false,"category":"none","all_categories":"bots,business,entrepreneur,","session":"C"}']

As we continued to attract new clients, the same issues kept coming up time and time again — and often from people within the same company. We realized that we needed a “knowledge warehouse” that could help current and potential clients with the basics. At first, our blog took on that role. We posted informative articles that answered frequently asked questions like When is a good time to post?, How to write clickable headlines, and even How to deal with writer’s block.

That fixed the problem for a while, but as we grew, our needs did too. Our blog didn’t only feature content marketing articles, it also contained pieces about our vision for client interaction and business. After a few years, we had compiled more than 60 articles, which made it difficult to find relevant information, even though we could still answer questions with direct links. But frankly, even our staff found it hard to get our hands on the relevant articles, so we decided to audit our content and create a “knowledge warehouse” with answers to the questions that really matter to clients.

AI Weekly

The must-read newsletter for AI and Big Data industry written by Khari Johnson, Kyle Wiggers, and Seth Colaner.

Included with VentureBeat Insider and VentureBeat VIP memberships.

At this point, we realized that a robot could complete this task better than we ever could, so we created our marketer bot. As we developed the program, it became clear that the bot could do more than we originally planned.

Tips for fixing simple issues

Our first idea was to create a database of common questions and answers for our clients to free up our staff to handle more pressing tasks.

There are two approaches to accomplishing this. The first is through “smart” software that can analyze user queries and learn over time how to resolve more questions. But such a chatbot is quite an investment and requires a lot of advanced programming knowledge. The second option was simpler (and cheaper): Use conversation scripts.

To get started with these scripts, we compiled a list of popular questions about content marketing like: “What should we write about?” “Does the timing of posts affect their popularity?” and “How do we work with negative comments?”

Above: A simple interface that gets answers to typical content marketing questions

At the moment, only our new clients get the pleasure of playing with our bot, but the number of questions sent to actual staff has already fallen by 20 percent.

Automation for better usability

Of course, we know that the bot on its own is not enough, and this system has much more potential that needs to be explored. But our bot can already provide answers to frequently asked questions, as well as analyzing the quality of content and giving recommendations.

[aditude-amp id="medium1" targeting='{"env":"staging","page_type":"article","post_id":2057782,"post_type":"guest","post_chan":"none","tags":null,"ai":false,"category":"none","all_categories":"bots,business,entrepreneur,","session":"C"}']

We started building a bilingual bot (English – Russian) from the outset, so we had to integrate a system to analyse both languages. Fortunately, the platform allows third-party services to connect to their bots through open APIs.

Ultimately, our marketer bot is able to check the quality of Russian and English articles (using the IBM Tone Analyzer for the latter).

Above: Our bot can analyse articles using the IBM Tone Analyzer API

Routing incoming requests with precision

After reading tips and getting to work creating content, clients can still find themselves in need of expert advice. Here, the chatbot can cut costs by sending a request directly to the right employee.

Even a simple form with relevant questions has a lot of value for a company. It helps decide what issues the client wants to solve, what budget they have, and so on. The answers then determine which employee is best suited to work with the customer.

[aditude-amp id="medium2" targeting='{"env":"staging","page_type":"article","post_id":2057782,"post_type":"guest","post_chan":"none","tags":null,"ai":false,"category":"none","all_categories":"bots,business,entrepreneur,","session":"C"}']

Of course, no one likes filling in forms and questionnaires, especially those with a lot of mandatory fields, but that’s easy to fix: Just get rid of all those troublesome fields.

From our experience, bots can benefit agencies. Businesses can even automate certain routine tasks and facilitate client access to a knowledge database. This is an excellent way to cut costs and free up time for staff to accomplish more valuable tasks.

VentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Learn More