Loyalty, a Journey or a Destination – How to Build and Execute Successful Loyalty Programs

Marietta:
Welcome to Velti’s webinar, broadcasted from The Netherlands & Greece. After many great comments from YOU, we are here with our 5^th webinar. Today’s webinar is about: Business Intelligence & Business Analytics. I’m GLAD you’re all watching. As you may know: DATA = the new gold! So why is Data the new gold?

We will answer this in today’s webinar. We have divided the webinar into the following parts: I will tell you more about VELTI, who we are and what we do. Then we will tune in to Velti Greece, where my colleague Professor Vasilis Vassalos, Velti’s Head of BI and Data Science will share more insights on the topic: Business Intelligence and Business Analytics and how it can help a business improve and thrive. We also host Mr  Orhan Celik [Cloud Solutions Architect AI & Machine Learning] and he is working Microsoft.

If you have any questions, please do not hesitate to type them in below, in the questionnaire and we will come back to it during the Webinar and otherwise after the Webinar in writing. My name is: Marietta Robbe Groskamp, Country Director for the Netherlands.

So, let’s start with a brief overview of our company. Velti was founded in the year 2000 has more than 20 years of experiences in the Mobile Marketing area. The solutions that we offer can be split into: CVM, Marketing Automation, Gamification, Loyalty, Permission Marketing and Gamified Campaigns. Today, we are going to discuss the last one – the Gamified Campaigns, which is a marketing tool that our clients extensively use in order to engage and reward their customers. The Velti family is in total 130 employees and growing and 50% of the all Velti employees are software engineers which means that we are on the latest trends and technologies. The main office of Velti are in Athens (where you also find the HQ and the development centre), Dusseldorf and Dubai. We also have 5 additional sales offices in Europe, Middle East, Africa, and Latin America. In the next slide you see an overview of Velti’s footprint and this is something we are extremely proud of.  Here you see an overview of all the customers we are working with. > more than 300 clients representing 70 countries, from Europe, Middle East, Africa and Latin America. Clients such as Deutsche Telekom, Vodafone to Etisalat and Pepsi Co and many more. If you would like to know more about a specific case study, please do not hesitate to contact us.

So let’s move to our main part of the webinar and let’s tune in to Velti Greece where my colleague: Professor Vasilis Vassalos, Velti’s Head of BI and Data Science will share more insights on the topic: Business Intelligence and Business Analytics.

Kalimera Vasilis, how are you? Greeting from The Netherlands.

Vasilis:
Goede dag, Marietta. The sun is shining today in Athens, a perfect day to talk about Business Intelligence!

Marietta:
So, Vasilis, what is BI and why is it important for every business?

Vasilis: 
BI is the eyes and ears of every business, of every project.

If you can’t measure something, you can’t improve it. BI allows a business to set and measure goals, to plan better through better forecasting, to use data for strategy, to generate insights, to monitor different aspects of the business. It’s a key component to success.

The philosopher Daniel Dennett coined the term “Intuition Pump” to describe a thought experiment used to illuminate specific aspects of a problem. Business analysis provides information, appropriately presented, visualized, used to illuminate specific aspects of a business. This may show problems, opportunities, trends, effects of actions we took or of ommissions. It facilitates and accelerates decision-making on strategy and tactics. And that’s why it’s necessary for every business.

Velti’s job is to run at a high level Customer Value Management efforts for our clients, to run highly engaging Loyalty programs, to employ Gamification effectively in these efforts, and to run successful Value Added Services with our worldwide Telecoms partners. To be able to achieve high quality in all 40+ projects we run concurrently across 3 continents and 30 clients, touching tens of millions of customers on a daily basis, we rely heavily on Business Intelligence and Analytics to achieve our goals. We want to share with our viewers a few ways we do that that have significant positive impact on our business.

So, how does the BI process work?

Step 1 is to get data from Data from source systems, integrate them and load them into a data warehouse or other analytics repository

Step 2 is to organize Data into analytics data models or OLAP cubes to prepare them for analysis.

Then follows the key step of running analytical queries or building models from the data to answer business questions. The results are built into data visualizations, dashboards, reports and the like, in order for the final step, to use the information and insight for decision making and strategic planning. While often the decision making is a human’s job, sometimes the conclusions are fed into automated workflows that take the Best Next action.

Marietta:
This looks complicated! Let’s break it down! What are some specific ways BI can help supercharge a business?

Vasilis:
Let me start with the bread and butter of Business Intelligence, business monitoring.

In order to understand the key aspects of a business or a project, we create Key Performance Indicators, KPIs, and reports showing their evolution. We create filters, graphs, tables, dashboards so that different stakeholders can view the aspects that affect them and are affected by them, daily, or in real time. Reports should provide appropriate feedback to all levels and functions of the company, from the operations specialist to the salesperson to the Chief executive.

This particular dashboard, built for one of our loyalty programs, shows the evolution of a number of KPIs over time, including participation rate and redemption rate. Ops and QA check some of these KPIs like Codes Sent, to ensure normal campaign operation, while Customer Success check Redemption and Participation Rate to measure effectiveness. In order to keep upper management informed, it is important provide a summary view that includes the most important KPIs. For our detail oriented executives and clients, we provide for our VAS offerings a summary showing on a daily basis subscriptions, billing rate, revenues from different sources, etc. Another key issue to monitor in every online business ( basically, after COVID, in every business) is recruitment. For our VAS services we monitor closely daily recruitment and quality for each channel for each month. Sales teams and Customer success use this to decide on business strategy, marketing budget allocation, and overall service optimization.

Marietta:
This was interesting, I like how it applies to every business, no matter the size or the industry! What other way do you use to help Velti succeed?

Vasilis:
The second way I want to highlight is data-driven business modeling. Almost all successful businesses use Business Models, and Business Analytics can help make them more detailed and accurate.  

So why is a BM important?

It is the tool that enlightens the future of the companies and reduces the outcomes of uncertainty.

Contributes to strategic decision making using all the relevant historical data.

Combines all the assumptions of business teams and returns answers and results.

Quantifies the profitability and sustainability of business projects in the mid term period.

Provides a clear view of the performance of any project at any time.

To do all that, it forecasts using AI and statistical methods coupled with business assumptions of key business drivers and business requirements.

At Velti, 21 years of activity give us the experience and knowledge to create and adapt our business models in every situation and focus on the KPIs that drive each project. In order to evaluate the investment on digital markets we use specific metrics. For subscription value-added services, for example Redemption Rate, Billing Rate, Unsubscribe Rate and days of tenure and more are all applied in our models. Revenue per user per tenure day is distilled and applied into ARIMA models in order to have a forecast of the current performance. Digital and Customer Success teams use the Business Model to optimize, configure and adapt their strategy, and to pick up early on trends.

Marietta:
Ok, so BI is critical for business planning, and with business monitoring. Both important activities. I come from Marketing. Anything specific BI can contribute to Marketing initiatives?

Vasilis:
I am so glad you asked that. Indeed, Business Analytics can aid with customer segmentation. It can provide detailed data to run “traditional” segmentation analyses like RFM, that I will explain in a minute, and also it can help run more data-driven decision making analyses, using machine learning techniques such as clustering for more fine grained customer segmentation.  Customer segmentation is the process of organizing customers into groups (or clusters) based on common characteristics. This allows Adjustment of sales and marketing campaigns to each customer group, leading to personalized sales & marketing strategy This results in Improved customer experience, Drives revenue, and fosters customer loyalty to the brand. It can tease apart subtle differences in the customer base that marketing actions can use to better address customer needs.

There are three “canonical” types of customer segmentation:

Value-based: customers are stratified based on their current and expected value for the company

Value-at-risk: customers separated by value and estimated churn scores

Behavioral: customers grouped by behavior, e.g. spending habits. This can provide the more detailed picture, and the most information that marketing and sales can use to address segments.

For Velti and Velti clients, segmentation leads to targeting each segment with specially scheduled and specially created marketing content and product features. Let me first mention a very common way to do segmentation, called RFM Segmentation, from Recency, Frequency and Monetarty value. Historical data are leveraged to put each customer in a quintile in each of the three metrics.  This way, for each customer we have a three (3) digit score, ranging from 111 to 555, resulting in a total of 125 combinations of RFM values. Collective marketing experience then leads us to assign customers with specific ranges of scores to predefined categories, shown here. For example, Can’t lose customers have good frequency and monetary value but bad recency, while Promising have good recency , and average frequency. The RFM analysis results in a map such as shown on the screen and can help give a fairly detailed picture of the customer base. But nowhere near as detailed as we can get by using Machine Learning techniques such as clustering for segmentation.

Clustering uses a lot of (relevant) characteristics, variables, to map customers into a high dimentional space and then find which points, representing customers, are closer together. Different algorithms can be used to find such groupings efficiently, our BI team has significant expertise in this. Business Results are indeed sensitive on finding high quality groupings. The effect of using segmentation to drive our marketing efforts has been significant, in every case we employed it. The challenge of course is to get our clients to share enough information about the customers to enable segmentation.