For the last 15 years, we’ve been helping some of the world’s top brands sell through Online Chat. In that time, we learned what it takes to make online Chat for sales a resounding success.
Everytime we attend a conference or meet with prospects, we get asked the same questions by marketers already using live chat for sales and facing major challenges as they scale:
(1) How do I make sure my sales team isn’t bogged down by unproductive chats?
(2) We deployed chat and chatbots across the site and we are seeing moderate results. How can we quickly increase our conversion rates?
(3) How do I reduce my Chat abandonment rate?
For years, marketers have been struggling with these questions - and there was only so much that existing technology and expertise could do to answer them. That is, until we realized the missing piece was the ability to predict which visitors will convert if engaged and which ones won’t.
(1) How do I make sure my sales team isn’t bogged down by unproductive chats?
We all know that a Live human agent converts at a higher rate than a bot. However, you cannot afford to have every web visitor interact with a Live human agent (plus, not every web visitor should get a Live agent).
The trick is knowing who will lead to revenue through chat. If we knew that, we could say:
- Web visitors with a low revenue potential should be handled by chat bots, we don’t need to be using costly live resources on these visitors.
- Web visitors with a medium revenue potential should be connected to a Live agent immediately to maximize conversion (speed really matters).
- Web visitors with the highest revenue potential seem obvious - they should also go to a Live agent. However, in some cases, those visitors should be left alone to self-serve, where putting Chat in the way of their experience could actually reduce the chance of conversion.
The challenge of knowing all of this increases many times over when you have thousands of web visitors simultaneously on your site, and most of them are anonymous. How do you then determine who-is-who and who-should-get-what, all in real time? Currently, this is almost impossible to do with traditional chat tools.
Without a way to proactively predict in real-time the revenue potential locked in each and every online chat, Live sales agents have been handling all Live chats, leading to lots of unhappiness: wasted time, low sales conversions, and lots of revenue left on the table.
What you need is a real-time method to precisely score and rank web visitors based on their conversion intent. Then, based on the ranking, you would align the best experience for each web visitor: a bot, live human agents, or self-serve.
To solve this problem at scale, MarketLinc built and launched Lift AI, an easy-to-deploy machine-learning application designed specifically to add a layer of intelligence to chat tools, chatbots, and your other martech solutions.
Lift AI really helps marketers when they start to run into challenges from the sheer volume of chats. Right out-of-the box, Lift AI is able to:
- Score all web visitors based on their conversion intent in real-time (even if they're anonymous)
- Immediately surface high conversion score visitors to Live Sales agents
- Instantly filter out low conversion score visitors to bots or self-serve (totally bypassing the Sales team)
As a result, Lift AI delivers much higher conversion rates and revenue figures as it sends web visitors who are predicted to convert straight to live agents, with the rest sent to bots. Immediately, visitors are matched to the best Chat treatment possible based on your resources.
(2) We deployed chat and chatbots across the site and we are seeing moderate results. How can we quickly increase our conversion rates?
In order to convert, you need to be dealing with qualified visitors. These are people who we know are likely to convert. After years of refining our targeting with traditional tools, we saw that up to 55% of visitors engaged with would still be unqualified. Bots helped to mitigate this slightly, but they still put up a barrier in the customer experience, or missed golden opportunities for the Sales team.
With that in mind, here are a some of best practices we have learned that will help you achieve higher conversion rates and more qualified visitors:
- To get the best results and customer experience, Live Agents need to be part of your chat experience - bots alone do not deliver the same results or experience (yet).
- Minimize wait times and “no agent available now” messages as they significantly reduce your conversion rate.
- Forecast your chat volumes to ensure you have the right level of agent coverage at the right times to ensure you are getting to all the visitors with conversion intent.
- Optimize invitation messages and bot dialogue flows to drive non-sales Chats to the right informational content or support team.
To help quickly improve your results from all the above points, you can introduce a layer of Lift AI intelligence to understand which web visitors are predicted to convert so they can be connected with your live agents immediately, while the rest are filtered out and sent to bots or self serve. In fact, Lift AI can cut the unqualified chats in half almost immediately, which results in increased conversion rates, and a more productive sales team.
(3) How do I reduce my Chat abandonment rate?
There are 3 key ways to reduce visitors abandoning their Chat experience:
- Prioritize visitors that show the highest level of interest or intent for human engagement, then engage them quickly before they become frustrated. On average, research shows that customers are waiting 2 minutes and 40 seconds, which every marketer will know is simply too long.
- Refine your Conversational Design. Chat is popular among visitors because it alleviates the “sales pressure” of a call or other experience, so make sure your conversations aren’t putting that pressure back on. For example, don’t ask for email upfront. This is far too common and results in companies losing at least half of their sales opportunities almost immediately.
- Improve operational processes to ensure live agents are available as and when needed (especially when visitors click the Live Chat option)
At scale, this is where an artificial intelligence tool such as Lift AI is critical to success. Visitors are segmented by their score, so you can quickly see how many conversion-ready visitors there are on any given day, then plan your operations and live agents around it. Simultaneously, those visitors can be connected to live agents immediately, instead of waiting to be served and ultimately abandoning their experience.
Today, Lift AI’s expertise is open to all digital marketers who need to improve the performance of their chat solutions and generate significant incremental revenue from the traffic already on their site.
