July 7th, 2008

The Power Of Personalized Multi Step Marketing

By Lior Lavy

The traditional mass-marketing “one-size-fits-all” approach is no longer sustainable in today’s competitive environment. Services and bundles of products being offered leave a lot to be desired when it comes to customization and personalization. This lack of flexibility is the main reason why many of the messages and promotions delivered fail to convert a user to adopt a new service or purchase a new content item.

The key challenge for all marketers in this sector, if they are to thrive in this environment, is to employ a multi-step approach that maximizes relevance for each customer through the dynamic adaptation of personalized marketing offers.

Bringing Personalization To Life

Everyone is talking about mobile personalization these days but it’s not as if operators can simply press a button and have it. In this post we are going to reveal for the first time the scope of the full process involved in bringing a leading tier 1 European mobile operator to significant revenue-generating results through multi-step marketing. Having the ability to create and execute a detailed, targeted multi-step strategy enabled this operator to make the leap from a simple mass marketing approach to a truly personalized strategy, which customers feel is specially tailored to their needs and preferences.

Why multi-stage you may ask? Because in order to create a true personalized and effective approach, you need to be able to lead subscribers up a value path based on their responses, or lack of response to previous offers. A multi step approach also enables fine tuning the promotion based on certain success criteria. In April 2007, we launched the first two promotions for downloading games and ringtones, and exposed customers to targeted messages on the mobile portal homepage. Additional phases involved scaling up operations to include MP3 and video downloads, chat and adult content.

At each phase of the process, the plan was basically the same, just taken to a higher level each time. An initial promotion phase would expose the user to special-terms content (free/demo/discounted) to provide us with an opportunity to learn what the user was interested in (if at all). Additional stages further down the line included promotions of discounts such as: Buy One Ringtone and Get One Free, or Every Second Ringtone At 50% Off.

Another type of activity included discovery of new services through recommendations, based on the content the user had shown interest in. For example, after choosing from a selection of free singles, the user is then offered to purchase singles in the Top 20, and whoever expresses an interest is then offered 5 singles for 3€ . Following the discounted bundle offer, any user who expressed an interest in this type of content through browsing the offer – regardless of whether they purchased or not – is then offered music video clips. The usage based recommendation is then reinforced with ongoing segmentation where specific segments of low/regular/high users are presented with enticing content offerings. Activity is reviewed and updated on a daily basis so that a low-user who starts buying for example, is automatically switched to a more relevant segment, and non-users who start downloading become low-users.

So as you can see in the illustration below, all three elements, Recommendations, Discounts and Segments, are all working together to efficiently produce an evolved marketing plan at each new phase. A bi-weekly synchronization with the operator’s product managers, real-time CRM managers and Pontis’ targeted marketing experts produces a comprehensive analysis covering:

Unique Recruits – new customers becoming engaged in the multi-step personalization process.

Usage Increase – not only if more products were bought in a specific segment but whether more people in the segment bought products.

Revenue Increase – giving things away for free may increase usage but may not necessarily have a good impact on revenue, and it was important for us to make sure that revenues never go down.

Click Ratio Increase – measuring the amount of clicks on specific banners to see if creatives are working.

After processing all this information decisions are made regarding the next phase of discounts, recommendations and segmentation (see below diagram).

It is clear that the scientific methods employed have a significant impact on operator revenues because over the first two months we increased the acquisition of bundles by 50%(!) From there we got to work on the video clips, which doubled the amount of clips being purchased. Don’t forget that all this began simply by exposing the user to content they may not even have been aware of through the free ringtone offer. From there the customer is guided up a multi-stage “value path” that is based on their usage, preferences and history.

So here is a recap of how the proposal evolves simply by the customer coming to the mobile portal four times after the initial free promotion offer, and based on the fact that they reacted to the offers there:

Recommendation: Get New Single For 1€ (at Ringtone price)

Recommendation: Buy 5 Singles bundle for 3€ (better value offer)

Recommendation: View video clips for 3€ each

Just like when you come into a store to buy clothes, the more the saleslady is in tune with what you need, the more you will buy. But when it comes to mobile operators, the increased value is exponential over time. Using these techniques led to a 100% increase in the purchase of video clips, and we believe people would very inclined to buy more of this type of content if they were aware of its availability.

In a typical Ringtone scenario shown below, non-downloaders were offered the first Ringtone at 1€ (from a special collection). Responders who have become low users are then exposed to a onetime offer of 1 Ringtone + 1 Ringtone at 100% off. And those who responded to the BOGO offer are then offered 1 + 1€ on a recurring basis and are automatically switched to the high-users segment. The result was a 17% increase in usage and 7% increase in overall revenue for Ringtones.

A few additional tips about how we achieve these results make this story complete. All statistics are based on a genuine control group that reflects the true behavior of users before Pontis was introduced. The control group in this case was provided by the mobile operator’s Business Intelligence department, which is completely in tune with external affairs such as holidays and special occasions.

Beyond the very real control group covered, measurements take two important elements into consideration:

The click ratio – if for example, there are no clicks on a certain banner it is best to replace it instead of taking up precious real estate.

Defining priorities among various products – unlike a supermarket that promotes one offer to all customers coming into the store, a mobile operator can maximize the click ratio at the entrance to their business by taking the connection between various products into account and combining that with the various segments to receive a very sharp picture of priorities. This would look something like: Ringtones for High Users – 1^st Priority, Singles for Low Users – 2^nd Priority, and so on.

So the bottom line is that by taking the service, segment and priority order into account a mobile operator ensures users seeing banners that are based on their usage history, responsiveness and real-time behavior. What more could you ask for?