I attended a few sessions on personas. I didn't get it all, but I did have some thoughts that I tried to record. Here are some questions to ask:
- Who is the shopper? Examples include
- First time buyer
- Repeat customer with specific frequency
- Loyalty program member
- What task is the shopper trying to accomplish? Examples include
- Replenish - buy a product they've bought before
- Accessorize - buy products related to what they've bought before
- Research - find information on specific product
- Browse - just killing time
- Leave - didn't intend to get here)
- What do you know about the shopper? You can find information in
a lot of ways
- Session data - referer, connection speed, IP address, etc.
- Checkout data - name, address, etc.
- Survey data - ask the customer!
- Ratings and reviews by this shopper
- Email responses
- Past purchases
- Clickstream - where has the visitor been on this visit and past visits
- Analytics data - What percentage of your visitors are repeat visitors? That's one way to measure if you're giving customers relevant products and services. How long do visitors stay?
Using a persona depends on defining tactics for how to engage the customer for various combinations of the above. For each combination try to understand (analytics) how they behave on the site. When something goes wrong (like shopper not getting what they're after), how and where can you intervene?
Personalization should be used to overcome the paradox of choice in ecommerce where longtail economics give more product selection than any offline store can provide. Some simple things:
- Change the homepage for repeat customers vs. new shoppers
- Change the homepage for product search engine browsers
- Add a loyalty box for loyalty program members
- Add a "best sellers" box
- Reorder product search results according to a merchandising strategy
- Add a "click-to-chat" button up when shopper puts a high ticket item in their shopping cart
Here's an example merchandising strategy for product search results:
- Give the newest styles higher priority for shoppers in 'fashion shopper' category
- Give the on-sale merchandise higher priority for shoppers in the 'bargain shopper' category.
- Give in stock priority over back order
- Finally order by price
54% of shoppers notice recommendations and 72% of those find them helpful.
Recommendation engines (customers who bought this item, liked these items...) are a good way of automating some personalization tasks. New recommendation engines personalize the recommendation (customers like me who bought this item, liked these items...). Real time, based on shopper behavior at the moment, not static models.
Speakers seemed to blow off privacy concerns even in the face of direct questioning about it using logic that went something like "we're going to shove products at you on our site it might as well be more relevant."