Leveraging the power of Adobe’s built-in AI

At this year’s Adobe Summit EMEA, as anticipated AI was a consistent theme that ran through keynote talks, workshops, and product demonstrations. Adobe’s CEO, Shantanu Narayen, talked about how AI and generative AI represents a new technology frontier and the power it provides to be a co-pilot in our day to day lives.  

For businesses it provides greater opportunity for automation, productivity, and personalisation. It also enables brands to engage with their customers in more meaningful ways, helping bridge from the audience segment to each of us as individuals.  

With Sensei, Adobe have been infusing AI into products for some time and innovations demonstrated during Summit showed how generative AI can be our collaborative co-pilot allowing us all to work faster and smarter.  

Those of us who were lucky enough to be at Summit were wowed by the innovation of Firefly, a technology that allows even those of us who are less creative to create beautiful images and designs with features such as generative fill.  

AI is not only enhancing our creative potential; it’s allowing us to understand our customers and deliver relevant personalised experiences at scale. We’ve been harnessing the built-in capabilities of Adobe technologies to help us in our everyday tasks. In turn, we’ve helped our clients better understand and connect with their customers.  

Adobe Analytics 

Anomaly detection is a statistical method of determining, in comparison to historical data, if a metric has significantly changed. Anomaly examples include a drastic drop in average order value or spikes/drops in landing page views. This is a great feature to see if there is something malicious affecting your data, as well as a quick look method to see if your latest campaign is driving an increase of traffic. Alerts can be set-up to notify teams if the same behaviour happens again.  

Contribution Analysis is where the true value can be seen when an anomaly has been detected. It discovers hidden patterns and insights on the root cause of anomalies in your data. Contribution Analysis looks at patterns of a dataset to see how statistical anomalies or correlations are present in your data. It breaks down contributions to anomalies in seconds driving vast efficiencies for data analytics teams.   

Adobe Target  

Auto-allocate. The traditional approach of calculating a winner in an A/B test is to calculate the number of visitors that must participate in the test before concluding the results are meaningful. One potential downside to this approach is that an underperforming experience is continued to be served until the required traffic volumes are reached.  

Auto-allocate, powered by the machine learning and AI of Adobe Sensei, determines the winning experience faster, allowing it to be served more often. More visitors seeing the best performing experience earlier is particularly beneficial when campaigns are time-sensitive, such as the peak Christmas period for online stores. We’re utilising Auto-allocate for a global navigation bar test that we’re about to push live for one of our clients with the aim of improving user experience and increasing conversions. 

Auto-target. Contrary to traditional A/B or multivariate testing approaches, the purpose of Auto-target isn’t to determine an ultimate winning experience. Leveraging AI and the profile information it knows about each visitor, it serves the best experience. It seeks to deliver the right winning experience for each visitor. As the algorithm is constantly learning, the experiences delivered can improve over time.  

Adobe Real-Time CDP  

Adobe Real-Time CDP gathers known and unknown customer data. It builds customer profiles, updated in real-time allowing brands to segment audiences and offers for personalised experiences. 

Customer AI, a built-in feature that predicts what customers are likely to do - and why – at an individual level. This is particularly powerful for predicting conversion and churn, allowing brands to deliver personalised messages accordingly. For financial services, it can predict if someone is likely to take out a loan, while for online stores it can predict a customer’s propensity to purchase.  

Predictive lead and account scoring (B2B and B2P editions) offer an AI-powered feature to help businesses predict the likelihood of leads and accounts converting and advancing through the various opportunity stages. Aggregating individual behaviours onto account level and automating the qualification process allows businesses to prioritise high-value accounts to target.  

The ambition to deliver multi-device and channel personalised experiences isn’t new, however a collaborative effort between brilliant human minds and AI infused technology now truly allows this to be realised.