Digital (Bear) Hugging
In my latest column I reviewed Indonesia’s potential to turn into a regional ecommerce powerhouse. Following a very successful first ClickZ live conference in Jakarta last month, and after talking with many local players, it looks like we are still on the road to far far away- the business fundamentals are there so Indonesia is on the right path but it is a long path. Groupon’s Indonesia CEO said that beyond the promising statistics (market size, internet users, smartphone adoptions, etc.) from investment banks and consultants there are still challenges to be solved (specified in my article).
The night before the event and after having had a couple of drinks with friends I sat down to tweak the presentation (typical me – both the presentation tweaking the night before and the beers…). I wanted to talk about reaching out to customers and leads in different channels and wanted to title the presentation “contextualization” which stands for: content, context and situation related. Forrester first revealed the concept of contextualization in the paper “Advance To Next-Generation Personalization”. In a nutshell, they argue that contextualization goes a step beyond standard personalization by adding:
- Customer data (who the customer is)
- Historical data (what the customer did in the past)
- Situational data (what the customer is doing now)
First response, bravo, I like the thought behind the theory. Spot on.
But do you really have a surplus $499 (come on Forrester, you’re not fooling anyone you actually want $500!) to spend on the report, which is bursting with fancy buzzwords. If not then read on…
Back to my story. So after a sleepless night, the presentation changed its title to “Digital Hugging”, which is a simplified version of contextualization. Digital hugging is essentially the ability to surround the client/prospect with love and affection across the core digital channels, while providing them with a relevant experience in the context and time that’s most relevant to them.
For a portion of digital love and affection, follow this simple recipe:
- 1 Recommendations engine
- 1 Email system
- 1 plug-in from your web analytics to your Database
- 1 Ad-exchange
- 1 database
- A lot of customers
- Website with traffic
- 1 plug-in for the purchase history of your clients
- Take the scripts of the recommendation engine and implement on your website (a simple copy and paste).
- Allow the machine learning algorithm 2-4 weeks to learn your user’s behavior.
- While on medium heat, ensure the plug-in from your web analytics/recommendation engine is connected to your customer ID on your database. It means that you will have the possibility to identify the web behavior and email behavior of your clients and prospects.
- Close the lead, and let it work its magic for 2-3 weeks.
- Take the plug in for the historical data, and link historical purchase data to the customer ID (be it an email address, first name/last name, customer #).
- Now your database contains: email behavior, web behavior, and also historical transactions (hint: you can add also social data if you use companies like Gigya for instance).
- While the gourmet dish is cooking, implement the recommendation engine widgets on your staging website, test them against other content you think can do better until the flavor of the algorithm is perfect. Recommended widgets: related (content related to what the users are browsing on the product level – complements a set of products to the product they view. E.G – related to camera are: memory, cleaning instruments, and a bag), Personal (locate it on your main page, so the user receives content that’s relevant to them), also bought (people who bought this product will also buy that product), and cart (allow the user to add easily recommended products to the cart, and by this increasing the cart value).
- Let’s rock n’ roll. When everything works move the widgets to your working website.
- Engage customers based on their situation: dropped carts? Send an email with recommended products to the user based on their purchase and web behavior. No response to email? Retarget customers with dynamic content recommendation (attention: not static banners, but recommended products abased on historical purchase data and browsing history). Customers not visiting the website? Follow them with email recommendations and or retargeting and try to bring them back to the products they like.
- Identify more situations based on channels, locations and behavior and use data, channels and content to engage your customers.
- Measure your customer engagement and CLTV (customer life time value). You’re looking for an increase in the metrics.
I hope the above recipe for simplified contextualization has been of some use. At first glance, it sounds complicated but it’s really not that difficult. Technologies who simplify the process, are available in the market and becoming easily accessible to SMB’s. BTW: a company who just gets it, and implementing this concept well is Tripadvisor. Kudos.