Consumer-centric marketing in a cookie-less world
It’s not an understatement to say that data has become a crucial part of marketing & media planning, fully integrated into the current workflows. However, new privacy regulations around personal data and its use are coming (or are already here!), and many brands and marketeers are wondering how they can still understand their audiences and stay relevant while also being privacy compliant.
Google’s announcement of third-party cookie deprecation is pushing this even further (although browsers such as Firefox or Safari had already removed these tracking solutions in the past), but you may be wondering: why is this having such a large impact? One of the main reasons is that companies believe that cookies are essential to effectively reach their audiences, and that while first-party data is valuable, it’s still far from offering the same value.
We do not agree with this view, and believe that there are options that can bridge this gap by improving on the analytic capabilities surrounding both the collection and use of all available data.
How can we do this?
We encourage brands to not lose sight of what’s important, take a step back, and reflect on the most important factor: relevancy. Traditional media approaches based on psychology and sociology, coupled with the latest advances in technology/data in the MarTech landscape, can provide better insights into audience’s behaviour and interest. This, in turn, can lead to a unified and holistic solution for all brands.
Backed by Data Science capabilities, this approach can greatly improve consumers’ interactions with the brand across all stages of their customer journey. From our perspective, some important points that brands should bear in mind are:
- Focus on building better first-party data
- Unlock the power of “hyper-local” actions
- Understand the most receptive context
- Create real-time lead-scoring models
- Don’t buy into one-size-fits-all solutions
- Integrate your short-term solutions with a long-term business plan
Focus on building better first-party data
One of the most important points that should be considered is the intelligent enrichment of first-party data stack. Intelligent in the sense that data has traditionally been collected in a reckless manner, trying to gather every available data point in order to not miss anything.
Nowadays, this strategy must instead be focused on identifying and capturing only relevant information. What constitutes relevant information depends on each company’s audience and strategy, but in the end relies on understanding what their brand represents and how their audience behaves and interacts with them.
Otherwise, Data Science and Artificial Intelligence solutions run the risk of becoming another “black box” where multiple data sources are thrown into the mix without first checking if they have actual value for the company’s purposes. The risk is precisely in that these superfluous data points can affect the result in unexpected ways!
Once the first-party data is correctly designed, advertisers should then turn towards how to leverage this data to, for instance, activate “lookalikes” when cookies are gone. Multiple alternatives are present today, but one such solution is Data Clean Rooms: a secure and private solution to store and analyse relevant anonymous user information, and further enrich it with additional data sources. Check the following article written by my colleague Jesús Templado for further context and information.
Unlock the power of “hyper-local” actions
Most current solutions in the market are reliant on IDs or identifiers for users based on some encryption and anonymisation process; for example, based on hashed emails or other unique user identifiers. However, Data Science can enable further analysis detail by focusing on location and geographical information instead.
Although direct geolocation data is strictly regulated by GDPR, the use of complementary data at a high granular level can help improve the first party data stack and obtain valuable insights to drive the marketing strategy. One such example is Google Search data, which can be analysed using Data Science algorithms such as clustering or regression models to identify regional trends, build keyword universes to understand competitor actions, and make targeted campaigns down to the city level.
Understand the most receptive context
Couple traditional adserving strategies and methodologies with the right medium to present your message. This is exactly what content relevancy through contextual marketing provides.
It uses Artificial Intelligence (AI) models to identify and index the content of different websites and place relevant ads on websites with similar content. This is done through a combination of Natural Language Processing (NLP) to analyse the text content of the webpage and the search results, Computer Vision (CV) models to identify the image and video content, clustering models to group and identify the topicality segments for webpages, and so on.
In short, the usefulness of context can help drive a message to users when they are most receptive, and can help increase the efficiency of specific steps in a marketing campaign.
Create real-time lead-scoring models
User intent is defined as the set of signals and actions a user performs and that leave some footprint that can be measured, which in turn showcase the user’s inclination towards a certain brand, category or product. These signals are not only impending purchase intention, but also “softer” intention signals; these include reading blog posts, subscribing to a newsletter, following the brand’s LinkedIn page, or browsing several product pages from a specific product category, for instance.
User intent can help multiple areas inside an organisation, from helping with Account-Based Marketing (ABM) and Sales Departments’ lead qualification processes to identifying lower intent users and guiding them to discover and engage more with the brand.
One application of intent in marketing is the real-time intent modelling of website visitors, assigning an intent score towards their possible conversion. Since most of the web traffic received on a webpage is anonymous, this solution can help drive more conversions on previously unknown users. This solution uses web analytics data, the user’s real-time web interaction, and other relevant data sources to build an AI model capable of qualifying each new user at each point of their website experience.
Don’t buy into one-size-fits-all solutions
More often than not, tech companies tend to replicate their solutions across clients, independently of the context surrounding their business, operations and specific characteristics. This is usually due to difficult requirements on the provider side and higher operational expenses, but can hurt the overall solution in the end.
A simple example just to illustrate this could be an AI-powered attribution model. Imagine such a model was developed for a car manufacturer, presenting the most relevant factors influencing their media investment, the ROAS per channel and so on as to successfully portray and influence their strategy decisions. Could this model be reused, with the new datasets available, for a chocolate manufacturer? Aside from the obvious difference in the product category, the underlying connotations (price, planned vs impulse buying, different media spends, different target audiences…) already show that this would not reflect the reality of the brand.
Whenever possible, it should be ensured that data solutions implemented in a company are tailored specifically to suit the brand’s needs, in order to avoid headaches in the future.
Integrate your short-term solutions with a long-term business plan
Data projects are often seen as tactical approaches to solve one impending need or digitalise a specific process. However, these “one-off” solutions often cause more problems than they solve, since:
- They might not be aligned with the company strategy
- They can be disconnected from other one-off projects or internal processes
- They usually lack context on the root of the challenge, focusing on a consequence and not answering the important questions
Data is just one more asset that can help brands stay ahead of the market, and therefore must be aligned with the company strategy and processes. Defining a long-term roadmap of actions and subdividing it into specific projects and tasks keeps all solutions aligned with the brand objectives and strategy.
In order to achieve this, an internal culture of change and data-driven decisions must be fostered inside the company, as adoption and belief in these solutions is key to their usefulness.
We hope these examples have shown several options that advertisers can take advantage of to stay relevant and ahead of the market. Although the landscape may seem daunting, allying yourself with a trusted partner that can help bridge the gap from vision to execution can be of great help moving forward.
BEDROCK is a Data Consultancy company with ample experience in the media & marketing field, accompanying clients in their process to becoming data-driven while making it understandable, easy and useful in every step of the way.