Chatbots have become one of the hottest trends in digital marketing and customer service. As part of our Made @ HubSpot content series, we‘re pulling back the curtain on how the HubSpot team has been experimenting with chatbots to create more helpful, human experiences for our audience.
Our key insight: chatbots work best when they‘re tailored to specific audience needs. By personalizing our chatbots for different segments, we were able to significantly improve the customer experience and deliver more qualified leads to our sales team. Here‘s how we did it.
The Backstory: Fixing Chatbot Friction
Our chatbot journey began back in 2019 when we noticed a problem with our existing chatbot experience. At the time, roughly 15% of the chats initiated on our website were from current customers seeking technical support. The issue? Our sales-focused human chat team wasn‘t equipped to handle these in-depth support requests.
To solve this, we implemented a dedicated "support bot" that could identify customers and quickly route them to helpful resources like our knowledge base, support team, or community forum. The results of this first experiment were promising – our support bot drove an 80% increase in sales team efficiency by allowing reps to focus on conversations with qualified prospects.
However, as we dug into the data, we realized we were missing a bigger opportunity. While support chats made up 15% of the volume, the vast majority – nearly 70% – were from prospects looking to learn more about HubSpot‘s offerings. Our support-only chatbot wasn‘t serving this key audience well.
Personalizing the Chatbot Experience
To better serve the needs of both audiences, we went back to the drawing board and designed two distinct chatbot experiences:
- The "support bot" for customers, focused on resolving technical issues and questions
- A "lead bot" for prospects and leads, serving up product info, pricing details and conversion opportunities
Here‘s a look at our new personalized lead bot in action:
We rolled out an A/B test on our main product pages, sending half of visitors to the original "support bot" experience and half to our new "lead bot" variant. We tracked two core metrics:
- Sales team efficiency (number of meetings booked divided by total chats)
- Number of sign-ups for our free CRM
After 30 days, the results came in and validated our hypothesis that personalized chatbots would outperform a one-size-fits-all approach:
- For the visitor/lead segment, the "lead bot" drove a 70% increase in sales efficiency and a 7% lift in free CRM sign-ups compared to the "support bot" control
- For customers, the "support bot" performed 52% better when shown only to that segment compared to all visitors
In other words, by matching the chatbot experience to visitor intent, we were able to connect prospects with the right sales rep faster while still giving customers the support they needed. The "lead bot" engaged visitors with helpful product info and only escalated those with strong buying signals to our sales team.
Implementing Your Own Personalized Chatbots
Inspired to create your own high-performing chatbots? Here‘s our 4-step advice for implementing audience-specific chatbots in 2023:
-
Analyze your audience: Mine your chat data, CRM and web analytics to understand the distinct audiences coming to your site. Look for patterns in the types of questions asked, pages visited and conversion actions taken by each segment.
-
Define conversation tracks: For each audience, outline the primary intents and map out personalized conversation flows to address their needs. Identify relevant resources, next steps and target KPIs for each chatbot experience.
-
Choose the right tech: Select a chatbot platform that integrates with your existing marketing and sales tech stack. This will allow you to tap into customer data for more precise targeting and chatbot triggering. HubSpot‘s chatbot builder, for example, can show different chat flows based on CRM list membership.
-
Test and iterate: Like any good marketing program, your chatbot strategy should be constantly evolving. A/B test different approaches, track performance and optimize over time. Experiment with elements like copy, visual design, targeting criteria and offers to find the winning formula.
While technology is a key piece of the puzzle, it‘s equally important to take a human approach to chatbot development. Write in a natural, friendly tone and use context to personalize the experience. Set clear expectations about what your chatbot can and can‘t do. And always give users an easy out to connect with a human agent when needed.
The Business Case for Chatbots
Still not convinced chatbots are worth the investment? Consider these eye-opening chatbot statistics:
- 68% of consumers prefer chatbots for quick communication with brands (Source: Salesforce)
- $11 billion in annual cost savings from chatbots by 2023 (Source: Juniper Research)
- 85% of customer interactions handled without human agents by 2021 (Source: Gartner)
The business case for chatbots is clear. When done right, chatbots can significantly boost customer satisfaction, accelerate sales cycles and slash service costs. According to IBM, companies that use chatbots can expect to see results like:
- 99% improvement in response time
- 30% better conversion rate
- 10x increase in interaction volume
With advances in natural language processing and machine learning, chatbots are becoming more capable by the day. Platforms like Google‘s Dialogflow and IBM Watson allow developers to build sophisticated conversational apps that understand intent and context.
Evolving Your Conversational Strategy
Of course, chatbots are just one piece of a larger shift toward conversational marketing and service. Forward-thinking brands are weaving chatbots into an omni-channel strategy that spans chat, messaging apps, voice assistants, and other emerging touch points.
The key is providing a consistent, cohesive experience across channels while still optimizing for the unique strengths of each medium. A mobile messaging chatbot, for instance, should be more visual and transactional, while a voice app requires robust natural language understanding.
As you build out your conversational strategy, here are a few best practices to keep in mind:
-
Put customers first: Always start with the customer need and work backwards to the technology. Don‘t build a chatbot just for the sake of it.
-
Keep it simple: Constrain the scope of your chatbot to handle a narrow set of use cases really well rather than trying to do too much.
-
Design for failure: Have clear escalation paths and fallback responses for when your chatbot gets stumped. Be transparent about limitations.
-
Measure and optimize: Establish the right metrics for your goals and continually fine tune your chatbot based on real interaction data.
As the old saying goes, the best interface is no interface. The businesses that thrive in the coming years will be those who can achieve the perfect blend of human and machine service, delivering relevant information through natural conversation.
At HubSpot, our chatbot experiments have shown us the power of this approach firsthand. By personalizing our bots for visitor intent, we‘ve been able to provide better experiences at scale while still keeping the human touch where it matters most.
The chatbot revolution is here – and it‘s not too late to get on board. By starting small and pursuing a true audience-first approach, you can begin realizing the benefits of conversational marketing today while future-proofing your business for the hyper-personalized world to come.
So what are you waiting for? It‘s time to start putting your own chatbots to the test and see how they can transform the way you attract, convert and delight your customers. The bots may be coming, but there will always be a place for helpful, human-centric brands. Just ask HubSpot.
