As cookies crumble and privacy laws intensify, DMPs are falling out of favor

By Alex Mcllvenny, country manager UK at zeotap

We’ve come to the end of a cycle in adtech, where past technologies will give way to privacy-compliant solutions. Data protection laws, such as GDPR and CCPA, are creating numerous challenges. However, while some adtech and martech players left the European marketplace in reaction to GDPR, many remained, and this hasn’t stunted the speed at which adtech and martech are advancing.

Tomorrow’s challenge is the ability for brands to move agilely across a complex adtech and martech landscape, out-maneuver their competitors and offer hyper-personalized experiences, all wrapped with data security and privacy compliance to win over customers.

Technologies of the past

For over a decade, to build up their stack, marketers have attempted to grasp growing customer-data complexity through a series of often ineffective or incomplete marketing technologies. Data management platforms (DMPs) were once seen as the holy grail of marketing. However, these technologies primarily focus on third-party data — which is being challenged due to Google’s latest announcement to drop third-party cookies in the coming years — and are designed specifically for advertisers and agencies to improve ad targeting. Additionally, the customer segmentation that’s available only offers marketers a chance to segment based on historical data — offering little insight into future behaviors.

Customer data platforms (CDPs) promise to offer a revamped and easier way to implement and achieve a single customer view from first-party data. However, brand marketers have recently voiced their opinions about how CDPs lack crucial capabilities for identity resolution, data hygiene and cross-channel orchestration — as shown in a recent Forrester report.

The limitations of both DMPs and CDPs mean brands are often still unable to make sense of who their customers are and what they’re doing. This general lack of understanding when it comes to how a customer interacts with a brand, and how their interactions with many other brands can affect their relationship, leaves them with little in the way of actionable customer intelligence. Since customer intelligence is central to smart marketing decisions, where do we go from here?

Turning data into intelligence

The marketing technologies of the past have focused on improved data capabilities, but have left out a much needed extra layer of intelligence that’s essential in making smarter, real-time, marketing decisions.

Customer intelligence will dominate the new decade, as brands become more sophisticated at developing their machine-learning models and internal algorithms. These can better extract insights, knowledge or intelligence about their customers, their future behaviors and the next best marketing action to perform. Intelligent solutions are being built and improved every day, from customer intelligence platforms, such as Oracle, SAS or IBM, to data clean rooms, but how can brand marketers combine them all and gear up for smarter marketing in the new decade?

Here are some steps to follow:

1.Bring your first-party data together and unify it under one single technology — it’s critical when forming single customer views to consolidate as much data as possible.

2. Resolve your current customer identity issues and fully understand their interactions with your own brand across channels (find out more on resolving identity resolution here).

3. Now that identities are resolved, segment customers based on historical data from the last decade, or take it one step further with the following steps.

4. Blend your first-party data with additional external data for a 360-degree single customer view that goes beyond your own interactions with them. Start looking into data clean rooms so that your data and this third-party data are not cross-contaminated and the whole process is GDPR-compliant.

5. Apply your own propensity models/algorithms to further and better understand how your customers would likely behave in the near future.

6. Re-cluster your customers according to future-based behavior and decide which cluster is worth additional investment and marketing resources.

7. Finally, create customer-centric strategies that best fit with each cluster and that are fully personalized based on real customer intelligence.

8. Analyze the return on investment from your marketing.

Potential real life scenario: Reducing churn in banking 

A credit card holder has three months left before their card expires. The customer is frequently on the go and has noticed the additional travel perks that a competitor offers. The customer’s current bank knows she is a high-spender who is quite disengaged from the brand.

The company’s internal propensity models (algorithms that calculate the likelihood of a customer’s propensity to perform an action) show that they are likely to lose this customer to the competing card. The customer is indeed looking for a new bank that complements their travel lifestyle, but the current bank cannot develop this specific customer intelligence without external input, unless they know other touchpoints and details about the customer’s frequent engagement with other brands, such as Booking.com, Uber and Skyscanner.com.

In this scenario, the current bank is relying on incomplete data or a gut-feeling based on first-hand interactions the customer has had with the bank. Lacking customer intelligence means the renewal email is likely to be overlooked, while the customer focuses on finding a new credit card with a better-aligned provider.

For the current bank, the other choice is using data-blends and analytics. For example, the bank could implement a data clean room to securely enrich and run analytics on their own data with high-quality data from a number of external, GDPR compliant, sources. This step ensures they have an accurate single customer view that combines first-party and third-party data. From the resulting insights, the bank discovers that the customer is a frequent flyer.

After applying their internal machine learning models to better predict how the customer will behave in the future, it becomes clear that a tailored email for travel discounts redeemable at their top-interest apps is the best approach to pique their interest. On receiving the targeted direct email, the customer opens it and redeems it.

The bank also discovers through machine learning-led customer insights that many similar customers also enjoy using a popular ride-sharing service as their app of choice when traveling. The bank then creates a temporary partnership with the service and offers customers a 40 percent discount on their next trip. The result: The customer feels well catered-to by the brand.

As a brand, there are two paths to take in the new decade. Rely on historical first-party data or tailor your strategy based on real customer intelligence and accurate behavioral future forecasts. Which path will you choose?

https://staging.digiday.com/?p=356928
Ad rendering preventing in staging

Ad position: web_bfu