Skip navigation
Like what you’re reading?

Increase 5G upsell and customer engagement with contextual recommendations

22 percent of smartphone users who own a 5G-ready smartphone are still using a 4G subscription. Machine learning (ML) and artificial intelligence (AI) in BSS can leverage the behavior of early adopters of 5G to influence the buying pattern of potential subscribers, securing return on 5G investments.

Senior Solution Architect

woman using mobile for payment

Senior Solution Architect

Senior Solution Architect

5G - a missed upsell opportunity

Smartphone users surveyed in the Ericsson ConsumerLab & IndustryLab’s  Five ways to a better 5G study indicated  they are willing to pay 10 percent more on average for 5G plans that offer enhanced mobile broadband access. Furthermore, they are willing to pay even more for plans with bundled innovative digital services. Globally, the inclusion of relevant use cases on a 5G plan increased the 5G premium by a further 20 to30 percent. Suffice to say, there’s a risk of missed revenue opportunities when there are 4G / 5G plan and device mismatches.

The study also highlights that the lack of clarity around 5G is one of the major reasons behind slow adoption. Complex marketing rhetoric combined with technical jargon like AI, hyper cloud, mixed-reality, machine-to-machine, robotics, etc., creates a cloud of confusion. A knowledge gap around service and device compatibility adds to the problem. Among the respondents from the United States (US), 14 percent of smartphone users who claim they are on 5G were using a 4G smartphone and 12 percent who own a 5G smartphone say they have a 4G subscription. Better awareness and understanding of its tangible benefits can accelerate interest around 5G. The report points out that improving consumer perception could increase the likelihood of 5G adoption by at least a factor of five, enabling faster growth.


Old wine in a new bottle?

Enticing subscribers to consume newly launched products is not new to communications service providers (CSPs). To highlight the launch of 4G, CSPs used to send SMS promotions to activate the service, often with free trials. The only challenge with a 4G upgrade was the need for a SIM change that required manual action from the consumers. For 5G a SIM upgrade is no longer required, but the 5G upgrade process calls for an additional action on the consumer side – switching to a 5G-enabled handset. Subscribers that already have 5G-ready phones but are still using 4G subscriptions is an obvious and immediate segment to target.

Pursuing the subscribers likely to upgrade to 5G has its own challenges. These subscribers are more mature, and their expectations are very focused. CSPs must address knowledge gaps and communicate with consumers in terms that resonate with them when cultivating a relationship for upsell.  As an example, not every subscriber with high data consumption fits into the stereotype of avid gamer. An immersive gaming pack may not be relevant for all. It is important to understand the consumers’ purpose of usage rather than the usage itself. Maybe the best offer is an immersive gaming pack, or maybe an exclusive streaming service will prove more enticing. Another factor to keep in mind while selecting the subjects is the current 5G network availability. A CSP may find potential targets, yet they dwell in the places where the CSP has yet to upgrade coverage to 5G. All these factors must be considered in tandem to determine the best offer to target subscribers for maximum success.


Engage customers with contextual targeting

If a CSP launches a 5G product that supports guaranteed bandwidth for 4K streaming, it should be tailored to high streamers within a 5G coverage area that are using 5G enabled devices but still with a 4G subscription. A smarter way of doing this is to cluster the base to reach out to this finite group of subscribers. Streaming data volume, monthly data spend and frequency of add-on purchases are some of the attributes that can be used by AI for clustering.

Once the target group is finalized, promotions can be activated. A limited duration promotional trial can be run across all digital touchpoints possible, including social media to maximize the viewership and conversion. The trial could offer the acceptor a free access to the service for a pre-defined validity period culminating in renewal of service through subscription.

On acceptance each subscriber should be tracked till the end of the free period to see who renews the subscription. The candidates who renew the paid subscription, across all digital touchpoints, constitute the influencer group. With these learnings, the behavior of the influencers can be absorbed and used to create the next sets of product recommendations to the general base.


Engage customers with contextual recommendations

Recommendation context can be of two types: product context and subscriber profile.

  1. Next best offer (NBO) applies both a product context and subscriber profile for a recommendation. An AI algorithm correlates purchased events with the buyers’ profile, to find out candidates with higher purchase propensity for a specific product.
  2. Recommendation based on similar interest relies only on products and recommends based on similarity and association of the product with reference to the rest of the products in the portfolio.

The NBO engine learns from the 5G product purchase events made by the influencer group. As a result, it can identify the ideal characteristics of a subscriber with higher purchase propensity. It now can recommend the product to a subscriber having similar features. Subscribers with similar behavioral traits, irrespective of the fact that they own a 5G device or not, will start getting such recommendations. This effectively extends beyond the target segment we started with and with time covers a large portion of sub-base.

Recommendations based on similar interest are influenced by how many times the 5G streaming product has been purchased. As more and more subscribers opt to renew the subscription, the AI algorithm puts higher weightage on the 5G streaming product. The product starts taking priority in the recommended item list. As an alternative, a CSP can manipulate the similarity and association of the 5G product with similar 4G products and lift the rank of recommendation of the product. In either case, the 5G streaming product should start appearing as upsell when a similar product is viewed, and as cross-sell recommendations when an associated product is added to the cart.

Both product characteristic based recommendations and subscriber profile-based recommendations need continuous feedback for retraining purposes. The algorithm re-aligns itself according to what it learns from the feedback received from all digital channels through which the recommendations have been made. Iterative training significantly improves the conviction of the recommendation.


Enabling 5G monetization through increased subscriber count

Intelligent recommendation is a part of the Digital BSS - Ericsson portfolio, within the Digital Experience Platform’s customer intelligence capabilities, that enables CSPs to orchestrate customer engagement with contextual recommendation via some simple steps.

Upsell 5G subscriptions with AI based contextual recommendation using the Digital Experience Platform’s customer intelligence capabilities

Figure 1: Upsell 5G subscriptions with AI based contextual recommendation using the Digital Experience Platform’s customer intelligence capabilities


Machine learning capabilities that underpin customer intelligence span a range of different intelligence capabilities: intelligent segmentation, recommendation based on similar interest and next best offer. In addition, marketing journeys round out the intelligent recommendation methodology. In the context of encouraging 5G subscription upgrades, intelligent segmentation helps to identify the next set of likely adopters of 5G in an iterative approach, while the marketing journey engages with subscribers owning a compatible device across omnichannel digital touchpoints such as WhatsApp, Email or social media. The recommendation methodology capitalizes on the purchase conversions to promote 5G products further.

Overview of the Intelligent Recommendation platform [indicative]

Figure 2: Overview of the Intelligent Recommendation platform [indicative]



A hypothetical example demonstrating how to boost 5G adoption

Paul, a marketing manager, is using Ericsson’s Digital Experience Platform to grow his 5G subscriber base. There are three specific stages Paul goes through:

  • Stage one: He strategizes with DXP’s intelligent segmentation and identifies a potential influencer, Maggie.
  • Stage two: He opts to promote a 5G product by engaging with Maggie over social media using DXP’s marketing journey.
  • Stage three: Paul monetizes his 5G product and extends the recommendations to Dave and Sam. He does this by influencing DXP's next best offer and upsell/cross-sell recommendations with Maggie’s purchase event. The outcome of the engagements with Dave and Sam are consumed by the AI engine to recalibrate the subsequent recommendation iterations. With time, the conviction of the recommendations increases, effectively boosting the 5G mass-adoption rate.


A CSP can improve the conversion rate of contextual recommendations by using the ML capabilities of DXP-CI strategically as demonstrated above. In a full blown 5G eco-system, the success of a use case will heavily depend on how a CSP can commercialize the digital services offered by the CSP and its partner players. In such an ecosystem, intelligent recommendation methodology can be a real game changer.

Contact us for a demonstration


Learn more

How contextual marketing can help you monetize 5G use cases

Five ways to a better 5G - Ericsson ConsumerLab & IndustryLab

Read more about telecom BSS

Read more about 5G monetization

Getting creative with 5G business models

The Ericsson Blog

Like what you’re reading? Please sign up for email updates on your favorite topics.

Subscribe now

At the Ericsson Blog, we provide insight to make complex ideas on technology, innovation and business simple.