See it and forget it
While simple AB testing allows Swisscom to improve experiences, Adobe Sensei enabled capabilities in Adobe Target allows Swisscom to improve returns and speeds as well. For Swisscom’s mobile product pages, all versions of a smartphone or other device are displayed on a single aggregate page. Customers then choose the exact colour, storage and other specifications desired for the device. Swisscom decided to test which options should be selected as the default.
Typically tests distribute visitors equally between the test options to discover which option performs better. With real-time analysis from Adobe Analytics, digital teams can monitor progress of the tests to spot trends and manually adjust distribution or options if needed.
The Auto-Allocate capability, powered by Adobe Sensei, automatically adjusts traffic based on performance to maximise returns. If conversion rises when customers are presented with the silver version of the phone instead of the black version, for example, then Auto-Allocate will start funneling more customers towards the silver version. Conversely, a poor-performing option will see less traffic. This overcomes the issue with manual AB testing where companies cannot capitalise on the winning experience while the test is still running. As a result, Swisscom benefits from increased conversion and revenue.
“With Adobe Sensei enabled capabilities, we’re automatically improving returns without needing to manually monitor and adjust performance,” says Mériel. “We can set it and forget it, as Adobe Sensei will take care of the rest.”
Personalising recommendations automatically
Customers are used to seeing recommendations while shopping or reading articles, but attempting to generate these recommendations manually can be a time-consuming process. The Personalised Recommendations capability, powered by Adobe Sensei, automatically curates content by identifying relationships between content, generating links and crosslinking relevant pages based on user activity and internal algorithms. This allows Swisscom to create cross-sell and upsell opportunities without the high time cost.
Swisscom also uses Personalised Recommendations in less obvious ways to personalise customer experiences. On the help and home pages, Personalised Recommendations reorders the topics for each customer to highlight desirable topics and provide a better experience. For customers viewing the home page on a mobile phone, Personalised Recommendations will detect over 90% of the different devices in order to provide a unique “My Device World” experience with accessories, help topics and other information related to the customer’s particular phone model. By visiting or deep-linking to the “My Device World” experience directly, Adobe Target in conjunction with Adobe Experience Manager Assets will automatically detect over 200 different devices. This smartphone-centric and unique “My Device World” experience is highly relevant as it is targeting at the digital doppelganger level and generating uplifts of almost 500% compared to the previous more manual approach.
“Personalisation is an important part of developing a fantastic customer experience, but it can be time consuming,” says Mériel. “By automating recommendations using the intelligence of Adobe Sensei, we can provide even more personalised experiences while spending our time building new tests and experiences.”
Using machine learning to develop new opportunities
Auto-Target is the latest Adobe Sensei enabled capability being used at Swisscom. Much like Auto-Allocate, Auto-Target automatically identifies top performing test options and adjusts audiences to maximise returns. But Auto-Target adds audience segmentation to identify the highest performing test option for each segment. For example, a test might show that most visitors prefer a news link labelled as “Headlines”, except for young audiences, who respond better to the phrase “Breaking News.” Auto-Target would maximise returns by sending more traffic from young audiences to the “Breaking News” test option, but adjust the traffic towards “Headlines” for all other audiences.
Swisscom has started to use Auto-Target on pages such as the Sports page on its news and entertainment website to capture sports fans’ attention by personalising the headlines of their favourite sport. By segmenting audiences based on the products and services that the customer already has, Swisscom can use Auto-Target to identify relevant upselling and cross-selling opportunities for each customer. The company has applied Auto-Target to hero banners on the Home Page and in the Customer Centre. As a result, the smarter algorithmic approach in Auto-Target, powered by Adobe Sensei has generated an uplift of up to 40% as compared to random experiences by customers.
“Auto-Target is the most exciting feature out there,” says Mériel. “Once we set up an experience, Adobe Sensei will take over, using its machine learning to develop highly personalised algorithms that adjust over time.”
Looking into future innovations
Swisscom plans to continue expanding its Adobe platform, using Automated Personalisation, another Adobe Sensei capability in Adobe Target, to predict customer behaviour. The company also plans to expand its current usage of Adobe Campaign, part of Adobe Marketing Cloud, to serve as a central hub to orchestrate campaigns across all online and off-line channels.
“As a company, we’re a strong believer in the power that artificial intelligence and machine learning can bring to a business,” says Mériel. “With Adobe Sensei, we’re bringing speed, intelligence and scalability to our digital operations so that we can create deeper levels of more meaningful and more valuable experience for our customers.”