Measurement Source of Truth
Welcome to our 2nd guest blog post. This time we’ve asked Maor Sadra, CEO of and Co-Founder at INCRMNTAL, to dive into the realm of marketing data, emphasizing the importance of measuring value over merely tracking users. INCRMNTAL is an always-on incrementality measurement platform, used by gaming, fintech, eCommerce, and other advertisers globally. We hope you enjoy this.
In today's data-driven world, marketers have assumed a role akin to day traders. However, the landscape of marketing has been reshaped significantly by privacy restrictions. Marketers now strive to unearth a solitary wellspring of authenticity within an increasingly fragmented environment.
In the 1950s and 60s, marketers were limited to primarily two methodologies for measuring their marketing efforts: Incrementality tests (which involved suspending advertising in one region and comparing it to a region where advertising continued) and Marketing Mix Modeling (a method that organized sales, advertising, product pricing, and other variables to identify correlations for more effective planning).
During this era, marketing was conducted through a handful of mediums, and the advertising process was relatively cyclical and almost static.
Today’s marketers will often advertise across multiple mediums, channels, publishers, and campaigns, utilizing methods such as algorithmic optimizations, dynamic creatives, and programmatic advertising.
The new generation of marketers turned to user-level attribution as a source of truth. The challenges of real time attribution were perceived as a cost of doing business. Only a few sophisticated advertisers continued conducting incrementality experiments using hold-outs, or planned experiments as a means to add some controlling by validating the actual incremental value of certain channels.
Over the past several years, lawmakers, regulators, and platforms began making drastic moves to reduce access to user-level data. These steps were slow but predicted a pattern where no user level data can be accessed without users’ consent.
This change in paradigm disrupted how marketers measured performance. If previously advertisers were able to track users ad engagements, building attribution models such as Multi-Touch attribution, First-Touch attribution, or Last-Touch attribution, the elimination of user-level data put an end to these methodologies of measurement.
With user-level attribution as a source of truth gone, marketers started looking for a new source of truth; however, methods such as incrementality experiments, and MMM, invented almost 60 years ago, were simply not applicable for today’s world.
Incrementality Measurement
Incrementality measurement acts as a source of truth to understanding the VALUE of ad spend.
Incrementality testing used to rely on a planned experiment. For example:
- I will keep my radio ads live in Cincinnati, while pausing it in Miami. I will compare the difference between the performance in these two cities.
- I will pause TV for a period of 2 weeks, and reactivate my TV spend for the next 2 weeks.
The concept of incrementality, relying on planned experiments, was theoretically effective in the 1950s. However, in today's dynamic marketing landscape, advertisers often operate on a larger scale, utilizing multiple digital channels where audience overlap and the influence of external variables have a profound impact on performance.
Pausing a single channel frequently yields minimal discernible effects, as conversions may merely be attributed to another channel, thereby misleading the marketer about the true value of that channel.
In today's context, the notion of a regional hold-out lacks practicality. While advertisers theoretically have the capability to run highly targeted ads, a simple check of your location based on your IP address will likely reveal inaccuracies. Consequently, pausing one of many campaigns merely provides an opportunity for other campaigns to fill the void.
Micro-Experiments pose a new way forward
Many marketers implement a plethora of changes. Consider your own team for a moment. How often do you refresh creatives, adjust bids, expand campaigns, downsize campaigns, initiate new ones, suspend old ones, or introduce new channels?
In essence, every small marketing adjustment you make constitutes a micro-experiment. Leveraging your own micro-experiments in conjunction with your unique seasonal patterns enables a platform like INCRMNTAL to accurately gauge the genuine value of anything you wish to measure. This approach integrates all your previous marketing endeavors with your performance trends throughout the seasons.
From Micro to Macro
MMM acts as the source of truth for distributing budget amongst mediums, and channels.
Marketing mix modeling has traditionally proven highly effective for retail advertisers, particularly those operating at scale and lacking direct access to consumer-level data. Retail brands frequently employed marketing mix modeling to allocate their advertising budget across various mediums such as Radio, Print, Billboards, and TV.
Through MMM, they gained insights into which medium contributed most to sales, providing economists with valuable data for future planning.
However, MMM was not originally designed to accommodate the complexities of a DSP with thousands of publishers, where the inventory composition changes daily. It also wasn't tailored for algorithmic bidding and dynamic creatives.
Hence, MMM should be viewed as a directional compass primarily useful for guiding budget allocation among channels, albeit within certain constraints and limitations.
Single Source of Truth ?
The notion of having a single source of truth in marketing is but an illusion. User-level attribution may have purported to be such a source, but the presence of attribution fraud, indications of excessive attribution, and persistent unresolved discrepancies among data sources have debunked this claim.
Incrementality, the ultimate goal in marketing, signifies that continuous progress is being achieved. It implies that marketers are generating outcomes that the company would not have realized without their marketing endeavors.
While Marketing Mix Modeling (MMM) stands as a remarkable tool for allocation and scenario planning, it can only furnish a high-level perspective, susceptible to distortion by the intricacies of digital inventories.
In truth, your own unattributed data serves as the sole authentic source of veracity. The practice of marketing measurement should never have been as simplistic as tracking the last click of a user, for most users traverse a marketing funnel encompassing awareness, interest, desire, and action. Attempting to attribute this entire funnel to a single engagement has led to marketing disasters, including overspending, misallocation of budgets, diminished efficiencies, and more dire consequences.
The core focus of both incrementality and MMM lies in evaluating the value generated by your marketing activities.
About INCRMNTAL:
We assumed that today’s marketers are not like 1950s marketers. And for that reason – we should not try and apply old methodologies dressed in new clothes, but redefine incrementality measurement for today’s marketer. The result was a platform that offered a new view of measurement. A platform that can measure the true lift in value, rather than a theorized approach, that would only survive in a laboratory.
INCRMNTAL’s platform is used by companies such as Binance, SEGA, AutoScout24, eToro, Hopper, Gamehouse, and many more
About Maor Sadra:
Maor is the CEO at INCRMNTAL, and is a long term adtech veteran with over 20 years experience, having spent a decade in web advertising, and later mobile. Previously was Managing Director at inneractive (acquired by Digital Turbine), and CEO at Applift GmbH
Maor is well known as an opinionated thought leader in the area of programmatic advertising, marketing measurement, and marketing analytics.
Maor co-founded INCRMNTAL with his former colleague and friend Moti Tal, previously CTO at inneractive (acquired by Digital Turbine) and CTO/co-founder at Simplaex (acquired by Index Exchange).