Gone are the days of one-size-fits-all sales funnels. Instead, e-commerce marketers must rely on personalization and customer engagement to cut through the noise in a customer’s inbox and turn clicks into sales.
Thankfully, advances in artificial intelligence have made it easier than ever for small business owners to target individual customers with messaging that converts. From sophisticated email algorithms, to chatbots and app notifications, AI can help you track and respond directly to customer behavior insights.
Customers are starting to expect more personalization from online retailers, too. According to the Kissmetrics blog, 79% of customers prefer automated interactions with live chatbots over receiving emails. Not only is it easier – and faster – to solve a problem via live chat, but chat messages are also highly personalized to that customer’s particular needs.
And when 68% of e-commerce customers abandon their carts before checking out, you need ways to increase engagement, solve problems, and personalize product recommendations until you seal the deal.
Because artificial intelligence can gather and analyze customer data lightning-fast, incorporating AI into your sales and customer retention strategies will yield results – but only if you’re ready to adapt.
Here’s everything you need to know about how to leverage artificial intelligence in your e-commerce to capture new business and engage repeat customers:
Step 1: Get Personal with Your Customers
You’re most likely already used to using customer data – like demographics, previous purchases, or location – to direct your sales team. AI offers even more robust personalization, allowing you to access predictive data about customer behaviors to fuel product recommendations. This three-pronged approach to customer personalization can yield powerful results:
Personalization: You’re probably already using some form of personalization in your marketing and outreach with customers. Go beyond incorporating names in email communications or sending timed reminder messages; instead, leverage artificial intelligence tools to:
- Time marketing messages and target consumers to close more sales. Do you know when your best customers are online? How often do they make purchases? How likely are they to respond to a well-timed discount or offer? Predictive analytics can help you target customers based on their habits, a tactic that leads to more successful sales, higher overall profits, and better customer retention. For example, when Envelopes.com started sending emails to customers who abandoned their carts, they cut abandonment by 40% across their site. Now that’s a well-timed message.
- Suggest content. Do you have a customer who always clicks through to your blog content but hasn’t yet made a purchase? By tracking engagement, you can use customer data to suggest additional content, from articles and blog posts to meaningful ads, that will help you drive sales. Tools like Marketo make it easy to match your best content to the right visitor at the right time – to turn a reader into a customer.
- Send push notifications to mobile users. Knowing when your best customers do their online shopping means you can send optimally-timed push notifications with discount codes and personalized sales offers. Swrve and Twilio offer easy SMS marketing – triggered by a customer’s location and past browsing history – that will make an impact on your ROI.
- Create behavior-driven user personas. It’s time to put all that customer data to use. By generating specific user personas that are informed by customer behaviors, you can more effectively personalize your marketing efforts. Know your customers’ pain points – and how these challenges change depending on the type of customer you’re dealing with. Then, track specific behaviors, like open rates, to trigger the right message at the right time.
Customer Prediction: With more robust data analysis, your team can make sophisticated predictions about what your customers want and need – and time their outreach to be more effective. Use this data to:
- Make reasonable predictions about future purchases. Track customer preferences and behavior to gain insights and suggest additional products your customers might like. Amazon’s been doing this – successfully – for a long time. Just think about how they suggest pairing their products with similar items to entice customers to add more to their carts. Try Tableau, a full-service sales analytics program that can help you analyze and predict customer behavior to better coordinate across your team.
- Design customized sales strategies. How often does your sales team coordinate with your marketing team to share insights? With more relevant data from AI tools, your sales team can personalize pitches to individual consumers to turn leads into sales. Customer data might suggest that particular discounts, products, or marketing messages work more effectively for certain customers – which can trigger the kinds of messages they receive in the future. Tools like Einstein from Salesforce or Apollo from Base can help your sales team make sense of the insights they’re receiving and integrate these insights into your CRM.
- Drive search results. Sometimes it helps when the search engine for a business speaks the customer’s language. According to BrainSINS, improvements in “semantic search” can transform a query like “women’s black blazers under $100” into a page full of product recommendations – rather than an error message. Not only will this make navigating your products easier for customers, but you can use search term data to continuously optimize how search works on your site, too. Think better SEO, better click-through rates, and more insights into the needs of your customers.
Product Recommendations: Make your predictive data even more personal by suggesting products your customers will enjoy. Integrate customer data to:
- Help customers find the right product. The AI capabilities of Google DeepMind will transform search, as well as the ability to process large amounts of data quickly, over the next five to 10 years. According to Information Age, the powerful artificial intelligence technology that powers predictive searches used by Google can provide pathways to more sales for e-commerce businesses, too. This means your e-commerce platform can be designed to help customers find exactly what they’re looking for – or the next best product to fit their needs. According to Entrepreneur, visual search capabilities will also help customers communicate their needs to retailers, driving sales. Neiman Marcus and Urban Outfitters both use Slyce to help customers find products they’ve seen on the street or in catalogs, transforming engagement with the brand into a transaction.
- Personalize recommendations. Past behavior is a great indicator of future wants and needs, something behemoth retailer Amazon understands well. Their product recommendations draw on your past behavior, as well as demographic and geographic insights from other buyers like you. According to MarTech Advisor, these recommendations drive as much as 35% of Amazon’s sales.
Use a service like Barilliance to generate product recommendations, or build your own with Amazon’s recently-released DSSTNE.
Case Study #1: Warby Parker
What they did: When Warby Parker burst onto the e-commerce scene in 2010, they offered customers a new way to shop for trendy eyewear. But, because the company didn’t have a brick-and-mortar storefront at the time, they needed to come up with a way for customers to try on – and invest in – their product.
Warby’s innovative Home Try-On Program allowed customers to select and try on five pairs of glasses – for free – in their own homes. “Unlike most e-commerce sites our basket size is very small,” explains Carl Anderson, Warby Parker’s director of data science. “Customers wouldn’t normally purchase a pair of glasses frequently, as they would groceries.”
Always data-savvy, Anderson and his team understood that the data from their Home Try-On Program offered crucial marketing insights to:
- Build a recommendation engine. The Home Try-On Program is a “really great dataset because it is reasonably large and you can look at the covariances among the five frames plus what [customers] subsequently purchased and build a recommender based on basket analysis,” says Anderson.
- A/B test marketing messages. Anderson uses regular A/B testing to help make decisions about the data points that fuel his recommendation engines and marketing funnels.
- Predict demand and purchasing. Use predictive behaviors to help optimize stock, purchasing, sales predictions, and messaging.
Your takeaway: Putting robust analytics in place to track customer data can help you predict customer behavior, drive sales, and increase customer satisfaction and loyalty.
- Happy customers fuel word-of-mouth Warby Parker’s social media users regularly post pictures of their try-ons online, asking for feedback. How can you use referrals to fuel your data collection – or data collection to fuel referrals?According to Pixel Union, social sites like Facebook and Twitter will soon move beyond helping businesses drive word-of-mouth. Advances in AI will soon make it possible to use social networks as places where consumers can purchase products, too.
- Predictive engines can help drive high conversion rates – even with niche products. “Warby Parker sells a monocle and it has an extremely high conversion rate,” says Anderson. “Most people who order this in their Home Try-On boxes end up purchasing it… we had to tweak our basket analysis algorithm specifically to account for it.”Which of your own products have high conversion rates? How are these reflected in your recommendation services?
Step 2: Provide Constant Customer Service
In a competitive field like e-commerce, excellent customer service is a given – and it’s often what stands between you and a lost sale. Use AI to free up your best customer service associates for more high-level work, automate segments of your customer support channels, decrease customer wait times, and guide customers through the sales process.
Virtual buying assistants: With powerful AI and machine learning, brands can develop predictive software that funnels customers towards a sale based on their particular needs. Use virtual assistants to:
- Improve real-time customer service. From answering basic questions about products to making product suggestions, virtual assistants anticipate customer needs and trigger sales strategies based on customer behavior. Outdoor retailer The North Face developed an AI helper using IBM’s Watson technology, which is pre-trained to handle basic customer questions. The virtual assistant asked customers about the conditions where they might wear a particular item of clothing – like a jacket – and then scanned a product database from The North Face to offer suggestions in real time. No longer in beta testing, Watson is now available to other retailers interested in implementing similar solutions.
- Reduce service costs. By handling common customer inquiries, virtual agents free up your customer support staff to handle more complex problems. This can eliminate some of the need for customer support by phone and increase efficiency with automated responses. At the Red Stag Fulfillment blog, Jake Rheude suggests that customer service chatbots may even be able to integrate with CRM to answer inquiries about orders, including tracking shipped packages. ZenDesk and Groove provide your support team with easy automation, chat tools, and customer support ticketing to improve how problems move up – or down – your customer support chain.
Customer service chatbots: Eliminate wait periods for customers and respond to questions or problems quickly and more effectively. Use bots to:
- Program basic customer support answers. Most customers have easily-answerable questions about products, shipments, or other data you already track via your CRM. With developments in AI, chatbots can now provide highly personalized answers to these common questions, freeing up your support staff to tackle more sensitive issues.
Some brands are even using messenger bots to make product suggestions, like Whole Foods, which uses Facebook Messenger to send hungry customers recipe ideas, or Sephora, which quizzes customers about their makeup preferences.
Use Facebook Messenger’s development tools to make a splash with a branded service, or ChattyPeople, which offers an easy solution for programming your own customer service chatbot without getting bogged down in software development.
- Trigger staff responses. A good chatbot program should ping dedicated, specialized staff who can handle more advanced problems. Use analytics to help train your customer support staff to be be more effective – and to anticipate common customer needs with helpful content.
One of ZenDesk’s advanced features uses customer satisfaction data to compare and improve the efficiency of your team members and analyze customer needs. These insights can provide managers with meaningful performance review tools – not to mention powerful sales and marketing data.
Case Study #2: My Starbucks Barista
What they did: Last year, Starbucks launched their virtual assistant, My Starbucks Barista. The app was developed using language recognition software, a subset of artificial intelligence that powers many virtual assistants like Amazon’s Alexa and Apple’s Siri.
According to AdWeek, “The feature… allows users to place their order by tapping a button and talking to the virtual barista. The bot then pings the order to a nearby store where an employee makes the drink. Users can then pick it up within minutes, bypassing stores’ long lines to the cash register.”
Of course, this virtual assistant has as many perks for Starbucks as it does for consumers:
- Reducing lines. By cutting down on wait times for customers, the My Starbucks Barista app adds instant value to the customer’s life – and earns Starbucks customer loyalty.
- Increasing mobile payments. Allowing customers to order and pay for drinks using mobile technology harnesses the 20% of customers who make weekly mobile payments at their favorite restaurants.
- Collecting customer data. The real benefit for Starbucks is all of the customer data the app collects, including drink preferences, frequency of visits, and lifetime customer value. All of this data can be used to market to and increase visits from repeat customers.
- Improving recommendation engines. When you have more data, you can create more sophisticated recommendation engines based on customer preferences, which drive sales.
Your takeaway: Artificial intelligence can ease customer pain points, improve and automate basic customer service needs, and provide your company with more data to predict behavior and drive consumption.
- Where might automation help you increase the efficiency of your customer support staff?
- What can customer service data tell you about customer satisfaction or customer loyalty in your business?
Step 3: Accelerate Company-Wide Decision-Making
Despite working in a data-rich field, more than half of marketers don’t use robust analytics on a day-to-day basis to measure the results of their campaigns or improve targeting. Don’t let this happen to you and your team. Big data tools can provide valuable insights into your biggest pain points, help higher-ups make informed decisions more quickly, and make yours the most competitive business in your industry.
Big data analysis: With big data capabilities should come smarter, faster, targeted decision-making about sales, growth potential, and marketing tactics. Although it can be easy to succumb to ‘data overwhelm’, resist the urge to apply a scattershot approach. Instead:
- Focus strategy on problem areas. How can you improve cart abandonment? Which offers work best on mobile customers? Which incentives or services have the best track record of converting new business into loyal customers?
- Share data across teams. Improve responsiveness and next-phase growth by actually cross-sharing your data. How often do sales and marketing talk with one another about the data they see?
An analytics tool like Domo can help teams get on the same page. Collect, analyze, and share data and insights from all of your cloud-based services – with multiple people, in real time.
- Calculate ROI and readjust. Do you have an accurate picture of which services have reduced inefficiencies, increased conversions, and boosted profitability? Big data should help you better anticipate areas of improvement and growth, so you can scrap the ideas that aren’t working – faster.
Competitor research: Plenty of big data tools allow you to compare and contrast how your business stacks up against your competitors. Use big data analytics to:
- Increase market share. According to Forbes, 66% of businesses say that competitors will use big data to increase market share. Whether you think big and use competitor data to help you provide a service no one else has mastered, or convince customers who are patronizing the competition to come back to the fold, understanding what your competitors are up to can help you drive marketing efforts and lead conversion.
- Monitor pricing patterns. Regular analysis of pricing patterns in your industry will help you stay competitive – and it’s never been easier to keep tabs on your competition across the web. Use a service like io to scrape important data, like competitor product pages, so you can analyze trends in your industry and stay ahead of the curve.
Case Study #3: PASSUR Aerospace
What they did: Since 2001, PASSUR Aerospace has collected sophisticated data points on the flights that come in and out of major airports, in order to calculate more accurate flight arrival times for airlines. Their service, RightETA, has major clients, including airlines like Delta, American Airlines, and Southwest, in addition to hubs like JFK and Newark Liberty.
Most clients need help eliminating the gap between stated arrival times on departure boards and actual landing times. According to Harvard Business Review, even a five- or ten-minute gap in arrival times can cause major inefficiencies – which leads to millions of dollars lost.
By using big data, PASSUR offers their clients the ability to:
- Identify specific problems – and offer targeted solutions. With large datasets for multiple airports across the country, PASSUR can help airlines target problems at specific locations, suggest changes, and implement up-to-the-minute plans for addressing issues as they arise.
- Make informed decisions more quickly. PASSUR is able to do this because they have datasets that span – in some cases – decades. “This allows sophisticated analysis and pattern matching,” explain Andrew McAfee and Erik Brynjolfsson at Harvard Business Review. “RightETA essentially works by asking itself, ‘What happened all the previous times a plane approached this airport under these conditions? When did it actually land?’”
- Increase efficiency. If crews – and passengers – have accurate schedules, no one wastes any time waiting around to do their job or get to their next location.
Your takeaway: Big data analysis – especially of datasets over time – can and should impact the decisions your company makes about products, services, and areas of growth.
- Which areas of your business need to be more efficient or cost-effective?
- How responsive is your staff to problems of efficiency? To taking advantage of new areas for growth in your industry?
Do you use insights from AI software to make company-wide decisions? Tell us how data and analytics powers your e-commerce business in the comments below: