This ties in to today's topic of behavioural marketing. Behavioural marketing is a version of hyper-targeted marketing that is individually tailored to the unique preferences of a person. The preferences are found by tracking their behaviour and then determining what messaging will be the most effective in reaching them at a given time.
Even if you aren't a big company, you still have a lot of data. Every individual person who comes to your site got there in a specific way. If you can see their past behaviour you can reach them with relevant messaging. You would be able to send targeted discounts or messages that can capture visitors based on their browsing tendencies and where they might be on the buying journey.
What should I be tracking?
The more visitors and traffic you can get, the more data you can collect. There are a few areas you'll want to watch to collect the best information:
You can look at previous purchases from your users and use that to send tailored recommendations, upsells, discounts and other types of marketing messaging. If you notice a customer buys a specific item pretty regularly you can send them info on that item, or discounts if they buy in bulk. The opportunities are endless with purchase behaviour.
Track what parts of the website people click to. This helps determine when to push notifications, popups and other reminders. If they're visiting for a bit maybe popup a discount code to encourage them to make a purchase. If you notice they've come back a few times you can highlight your free shipping to perhaps encourage them to make that purchase. For returning visitors you can look at data from their past visit to compare it to this visit and direct messaging at them.
Knowing where your customers are from will help you with marketing barriers. If your shoppers are largely German you might consider offering better shipping to Europe. Or having your site translated to offer a German version. You can also target visitors based on events or specific things happening in their area. An app like Bandsintown can do this and send you alerts to buy tickets when your favourite artists are touring near you.
What are people looking for?
If you can identify what visitors are searching for you can streamline your content to them. A user searching for cat toys isn't going to care about a popup advertising a big sale for dog food.
What data is useful?
Any data you can get from your customers is useful. Look at what you're trying to do or sell and decide which things you'll want to look for. You can always collect all data and then look at only one specific portion. Test different things to see what gives you the best return.