Aligning business with data strategy: “convergent” vs. “divergent” data approaches

In every go-to-market strategy we often begin with a hypothesis around the value proposition of the offering in the chosen market segment alongside competing products, target customers and their acquisition strategy, risks and rewards, etc. The validation of the accuracy of this hypothesis, however, only comes later. Rightly, the proof is in the pudding metaphor applies. Except that if the hypothesis is baked with poor quality information, it is one expensive pudding to be staring at.

The internet revolution has certainly made it possible to turn around traditional business models by enabling maximum reach in the shortest possible time with data in real time; all at a much lower cost base. However, the question is, should internet based companies, irrespective of the sector and product maturity stage, race to stake its claim to the great bounty that digital democratization has laid bare? Or should they pause and think clearly to better understand the sector they want to disrupt, and then using a combination of approaches, collect product specific data that best reflects their risks and rewards before going full throttle?

Before going further, we would like to explain what we mean by “convergent” and “divergent” data strategies.  While “divergent” data strategy focuses on a carefully selected dataset at the initial stages, followed by a broader data acquisition strategy later on, the “convergent” strategy is one where the aim is to reach out to the widest possible audience to capture as much information as possible to identify a niche on which to focus later on.

To support this reflection, we are choosing two sector specific examples: Fintech and Classifieds.

In the world of marketplace lending (sometimes still referred to as peer to peer lending), especially those dealing with unsecured consumer loans of a certain value and term, a “divergent” approach certainly holds merit.

In alternative finance, data is the fuel powering the core of this fully online, almost instantaneous loan-approving model. The acquisition of a good training dataset to enable automated scoring and rating on the platform is, however, a no mean feat! In theory, for developing a robust scoring model, deploying every possible tool in online marketing for generating and qualifying leads across risk classes, would be an invaluable data strategy. But, in practice, we know that passing fraud checks, video KYCs (Know Your Customer, as required by regulations in Germany) and certain threshold criteria, doesn’t necessarily guarantee a pre-determined probability-of-default in that risk class or lower loss-given-defaults, as a matter of fact. What is needed are continuous, persistent efforts in aligning business KPIs with data and marketing strategy to attract the right pool of customers on either side: borrowers and investors.

Despite the wealth of data held in databases with credit rating agencies, it is not all-inclusive in terms of coverage. Listing third part originations, though another strategy often used by online platforms to boost their loan inventory, is also not without risk as mapping across risk classes is tricky and inconclusive. In the end, too many defaults result in the platform being deemed as risky by investors; whereas with fewer defaults, the product becomes unattractive in terms of ROI as a fixed income asset. Under these circumstances, the best data acquisition strategy should be based on developing selected online and offline marketing channels for targeting and acquiring tightly constrained customer segments identified through sound market research. Converging to build a position of strength with selected customer base, and well defined checks and processes, is much needed to gain credibility and confidence of investors before diverging.

For “convergent” data strategy, one should certainly deep dive into the Classifieds business (or other e-commerce businesses).

Classifieds marketplaces, where used goods are sold by individuals or professionals to other individuals, are established following a very basic playbook: first growing to secure itself in position of market leader (literally crushing the competition) and then starting monetization. In parallel, the growth phase allows to collect very valuable data to understand the user base (sellers, buyers, professionals) and eventually identify niche opportunities/markets. Dedicated platforms (often referred to as segments or verticals) can then branch out from the main platform using learnings from the originally collected dataset. Obvious examples are second-hand cars or real-estate platforms, but more innovative platforms can emerge elsewhere in fashion, agriculture, etc. The data strategy thereafter is focused on collecting relevant-only information on the market/niche in question.

Once again, the data strategy is aligned with the business strategy for maximizing the efficiency of the data collected, and therefore support business growth.

Leave a Reply

Your email address will not be published.