The combination of first -party and artificial intelligence behavior can transform outgoing marketing electronic trading.
The aim called “individualization AI” is to create a personalized shopping experience adapted to prefee, behavior and history of the purchase of an individual.
Perfect shipping
“Internally, we strive for” Perfect Drying “when 100 percent of people who get a message click or join and no one are formed,” said Alex Campbell, the main innovative official and co -founder in Vibes, a mobile marketing platform.
Campbell discussed the potential for individualization AI (AI-I), rich communication services and mobile marketing in the retail sector when describing this 100% commitment, 0% scenario logging out.
Emmerce traders could edit this definition, but the perfect sending is when sending messages currently meets the need for a shopper.
The expectation of the trade
Shoppers who have signed up for e-mail, text or pushing messages want to be perpetrators.
“Every year we conduct a customer’s survey … and we always ask the question as,” What would make you decide? “Two years ago, we first heard,” You don’t send me enough news, “Campbell said.
The interviewed people signed the acceptance of mobile marketing. They wanted to receive relevant and timely product announcements and discount offers.
Ai-i can help.
First Party’s data
AI-I is possible because online stores can collect the first history of data purchase, browsing behavior, dat-inz relying on cookies or third-party providers.
People cannot come through all data. Even rules and automation would try to detect individual preferences in real time.
However, the AI layer can be used even courageous Deployment of messages.
Regardless of segments
Usually segmented traders segment shoppers around normal behavior. For example, a wine trader may have a segment for “Value Wine Shoppers” or “Premium Wine Collectors”.
AI-I creates segments of one, for example, a customer who buys red wine under $ 20, prefers RHône varieties, responds to Friday’s sending and often redeems mobile offers.
Folding perfect sending is much easier with one segment.
Say that Mercihant wine implements AI-I. This tool can broadcast Rich Communication Services (RCS) shopping messages and has access to the product’s behavior and customers’ customers.
Testing can lead to perfect shipping.
AI sends a RCS message containing a carousel with a product. (RCS has a feature of a similar application.) The message contains two offers: (i) Argentine Malbec for $ 18, as recommended by AI data -based, and (ii) Portugue Red Blend for $ 17, which was to introduce new wines to this buyer.
The shoppers ran over, taps, visited the web, clicked the “Malbecs under $ 20” filter and finally make a purchase. AI adds data from these touch points to the customer’s profile, records a purchase below $ 20, or adds a note for testing a copy around the value.
Every new message is an experience and will bring AI-I to distinguish what the shoppers want and when.
This process is nothing new. Data scientists could describe this as “individualized multivariating tests” or “context bandit”. It is an established way of identifying individual preferences.
What is different is speed and scale AI.
Detailed process
For a hypothetical wine trade, AI-I would use the initial settings for more detailed data collection, data normalization and integration.
However, once it is launched, the AI-I tool would probably follow a simple workflow for each new customer.
- Segmentation base. Start with wide categories of wine based on initial purchase, such as red or white, sparkling or sturdy and top or value.
- Early commitment. For example, start sending messages and songs when shoppers click Bordeaux for $ 40, ignores Rosé, but buy Malbec for $ 15.
- Individual testing. Generate messages specific to shoppers. Each of them is an experience. Offer in Bordeaux for $ 35 or Syrah for $ 18. Continue to monitor the commitment and behavior. Repeat.
- Specify the profile. The AI-I system has identified over time as likely, “The customer is likely to buy the price if the price is below $ 20 and the variety is bold red.”
- Balance with discovery. Introducing Wild Card wine every few – perhaps Spanish white or sparkling wine – to expand the knowledge of the system about the customer and prevent marketing fatigue.
- Feedback. All clicks, purchases and logging out feed on the AI model, both for individuals and to improve the overall system.
With each iteration, AI-I is approaching perfect sending.