To be completely upfront, this post details how you can use Hive to solve the cold start problem.
Two weeks ago, I purchased a handful of mystery novels from Amazon. Last week, I received an email from Amazon offering $10 off clothing and accessories. Having never bought clothing or accessories from Amazon, this offer was not very enticing (especially with the six irrelevant product recommendations included in the email). If this offer was for $10 off books, I would have dropped what I was doing and started browsing!
Although Amazon hasn’t quite figured it out yet, as ecommerce evolves, many product recommendation tools have become quite clever. Recommending relevant products to your customers is one of the best ways to increase repeat customers, and is definitely worth the effort. You can base recommendations on browsing behaviour, past purchase behaviour, or similarity to customers who’ve purchased similar products.
All of these methods are known to successfully increase sales, especially when compared to generic product recommendations not based on user data. Not only did Amazon’s untargeted email not succeed in creating a sale, it made me question Amazon. Why do they think I’m interested in men’s shoes? Why is there a picture of a belt above Handbags & Wallets? They have plenty of data on me - why aren’t they using it?
Outside of product recommendations, we know some simple email campaign targeting can go a long way: MailChimp’s latest investigation into segmentation reports 14% higher opens for segmented campaigns. It makes sense - knowing more about your customers naturally helps you talk to them about what they care about.
But what do you do when you have no information on your customers, other than a name and an email address? This is the definition of the cold start problem: having a data profile on your potential customers that lacks the depth necessary to accurately predict what they’d be interested in buying. And it’s affecting a ton of ecommerce stores. According to Retention Science, most ecommerce companies have a cold start problem with 45% of their user base.
Having a newsletter signup on your store is a great way to collect potential customer leads who want to hear from you, but if you don’t want to hassle them with long forms and never-ending fields, you’re collecting limited data. How do we send these potential customers great content without any insight into their characteristics or behaviours?
This is part of the reason we built Hive. When you import your customer data (however sparse or rich), we automatically use public and proprietary data sources to enrich your contacts further with names, demographics, locations, social media profiles, interests, and more. You can see what a full contact profile in Hive looks like here: What is a contact profile?
It might seem a little too good to be true, but it works. This gives you a huge leg up on the cold start issue, and helps you send great emails right out of the gate. You can intelligently segment users you had almost no data on before and send relevant emails based on gender, age, location, and even interests inferred from social behaviour.