To Buy or to Build Your Real-time Personalization Platform?

Posted on: September 6, 2019 | Posted by: Aurobindo Sarkar
Explore benefits and downfalls of building a platform in-house versus buying from a trusted partner.

The Pros  and Cons of Constructing a Real-time Personalization Platform Internally versus Partnering with an External Provider

These days, achieving ‘personalization’ in customer engagement is one of the most important organizational priorities for driving revenue growth. However, while in the past, mass personalization was sufficient to engage customers, today winning customers‘ attention relies significantly on 1:1 interactions that are relevant and meaningful to the customer, in that moment. This is not only about getting their names right in the messaging but also differentiating consumers on a human scale, and intelligently engaging every customer based on their wants and needs.  

Given that a majority of businesses recognize the importance of personalization for both customer acquisition and retention, the primary question that arises is: How to deliver this next generation of intelligent customer engagement? Should an enterprise construct its own platform internally or use a third-party personalization platform?

The case for building …

The main benefit of using internal teams for designing and constructing an intelligent customer engagement platform is having the ability to fully customize the platform (to match the very specific requirements of an organization). For example, if you want to combine a highly unique set of proprietary fields, or customer attributes and transactions, to real-time customer data then it would seem that a custom-built platform is the best way to go. In most cases, when an enterprise builds a personalization platform from scratch, they can respond to specific or unique requirements of their business users.

What does it take to build?

While internally constructing a solution can provide the maximum level of customization, some considerations should be kept in mind. For instance, building an intelligent personalization platform from scratch and training it to understand how historical, environmental, and real-time data inputs correlate requires expert understanding of how to integrate data residing in silos with real-time streaming data to get a 360° customer view. In addition, the use of machine learning (ML) models is critical when it comes to personalizing at the one-to-one level, at scale. These ML algorithms can combine many different sources of data, draw insights about an individual, and then select the most relevant experience to deliver. While a good personalization platform will make it easy for a marketer to leverage and even customize ML algorithms, the development of the underlying algorithms requires a team of sophisticated data science engineers. 

Accomplishing all of this is certainly possible, but it takes significant investments to design and build a robust scalable real-time platform that can flexibly support a wide variety of business use cases (consistently, across all customer touch-points). Putting together all of the solution’s components is also a technically challenging task requiring specialized skills. Additionally, new and / or continuously evolving business use cases can result in a prolonged implementation timeline, not to mention on-going maintenance and refinements to keep up with new feature requests and use cases can compound.  In most cases, such a significant development activity will take away core business and technical resources from their primary jobs for substantial time periods. 

The case for buying …

Conversely, by using or subscribing to an existing solution, enterprises can alleviate the risk of starting from scratch while onboarding 1:1 personalization capabilities. Business users can test their customer engagement ideas using out-of-the-box functionality, and in most cases they will be able to rapidly configure and deploy their personalization use cases in production. Custom-built solutions can take months or even years to complete, depending on the scale of data, parameters to configure, experience of the team, and roadblocks encountered. When build is prolonged, these platforms risk being outdated even before they are first deployed—costing enterprises wasted time and resources for only partially successful programs. However, solution providers, whose core business is customer engagement technology, tend to have an up-to-date platform that can meet most of the business’s needs. The enterprise is far more likely to gain competitive advantage, reduce risk and drive more benefits using the services of a proven, innovative, easy-to-use, easy-to-maintain personalization platform. 

How to choose a solution provider?

In general, the criteria to select a solution provider must encompass the following needs: 

  • Requirements Alignment – How well does the product meet your business requirements? Are your key objectives met with the product? Are major customizations required? 
  • Technical – Does the product meet your scalability, availability and security requirements? Can new sources of data be integrated easily? Is there flexibility to integrate in-house ML models, and custom filters/business rules?
  • Financial – Is there a specific return on investment (ROI) objective to be achieved? What is the total cost of ownership including human capital, software, and hardware?
  • Maintenance – What is the effort and cost of on-going maintenance and support with respect to human capital, hardware, and software?

The bottom line is that personalization is a top priority for businesses and the stakes are high to get it right. According to a recent Gartner survey, businesses are at risk of losing 38% of their customers because of poor personalization. They further suggest that to personalize effectively, businesses must hire and develop key competencies for personalization including:

  • The ability to understand consumers’ micro-behaviors; 
  • Know individual consumer preferences for trigger-based messages; 
  • Produce content that can be versioned to meet a wide variety of consumer profiles and triggers;
  • Learn to mine nontraditional data and apply creativity to analytics; and 
  • Share decision making with other disciplines

At the end of the day, given the revenue generation potential, the business’ preference is crystal clear  – they want their personalization solution now…and not months or years from now. 

© 2019 ZineOne. All rights reserved. | Privacy Policy