Act-On Gets Adaptive
Act-On Software, the Portland, OR-based integrated marketing platform, today announced “Adaptive Journeys” — essentially the addition of an AI-powered customer journey mapping layer to its marketing automation capabilities. “Customers can have a unique journey, rather than just being on one big generic track,” says CMO Michelle Huff.
The aim, Huff explains, is to go beyond personas: “We don’t like being stereotyped. And we aren’t always buying.” Huff points out the need to better understand the individual person through tracking, scoring, and learning: “All interactions mean something,” she says.
Huff offers an analogy with cloud-based mapping technologies like Waze, the GPS-based geographical navigation app. Apps like Waze:
- Are built on a cloud architecture
- Layer additional datasets (eg traffic conditions) over the basic data
- Identify your location
- Improve the user experience (eg by helping them avoid traffic jams.
Adaptive Journeys, likewise:
- Is built on a cloud architecture
- Adds extra data layers (eg email open rates, CRM data, etc)
- Takes account of the consumer’s location
- Sends the best message at the best time, throughout an ideal journey
That’s the intention, anyway, and the technology takes advantage of machine learning capabilities which, Huff says, were developed by Act-On in-house. There was no independent announcement about Act-On’s AI plans, Huff explains, “because marketers don’t really care until they can see something concrete.”
In the past, Huff says, when trying to figure out when best to send an email, marketers would try a general blast at a specific time; look at the results; do some guesswork; then graduate to some A/B testing. With AI, there’s continual, real-time, automatic optimization based on large-scale response data. It becomes possible to send an email when engagement is predicted; and a 1% increase in engagement can be hugely impactful on results.
Asked about how Adaptive Journeys differentiates itself from the broad spectrum of AI-based personalization tools appearing on the market, Huff says that the Act-On approach will have strong appeal to mid-sized brands who had previously thought machine learning and personalization at scale were only available to enterprises on the scale of Microsoft and Google. “For mid-sized companies with a small to decent-sized marketing team, it was difficult to think [we] can do this.”
Another advantage for Act-On is that it was built originally on NoSQL cloud architecture, designed to manage very large quantities of distributed and unstructured data, and is therefore well-positioned to house machine learning capabilities without fundamental re-design or re-building.
The 2017-18 road-map for Act-On is to push this new functionality further in the direction of predictive engagement.
Specifics of the release include:
- Adaptive Segmentation: the option to create lists of engaged contacts based on data CRM systems, web forms, or behavior tracking, the segments being automatically optimized as new data becomes available (additional sources will be added in the future)
- Adaptive Forms, now supporting conditional follow-up questions, as well as presenting or hiding, questions based on individual responses
- Adaptive Sending, predicting the best time to send a message to an intended recipient based on past behaviors and actions
- Adaptive Scoring, automated across segments, industries, and buyers, and
- Adaptive Channels, automatically selecting the optimal channel for messages.