When search engine marketing (SEM) first appeared some 20 years ago, most major brands delegated ownership of this burgeoning channel to specialist agency partners. Over time, as search budgets grew (alongside fees paid to those agencies) and self-serve tools became easier to manage, many brands made the decision to take SEM in-house.
Today, we see a similar trend with programmatic. What started as a shiny new plaything 10 years ago has morphed into the dominant channel for buying digital media and audiences across all screens. While the impulse of most brands might be to take programmatic in-house as they have done with search, it might not be right for every buyer.
The following five questions aim to help brand marketers evaluating whether or not in-housing programmatic is the best strategy for them.
1. Does valuable or sensitive data need to be protected?
Marketers today collect an abundance of data about their customers from a variety of sources including site interactions, loyalty programs, online and offline transactions and campaign analytics. When aggregated and structured in actionable ways, these data sets become valuable signals for use in programmatic media buying and optimization.
If this data is highly sensitive, or if the organization has very strict policies around how and where customer data can be activated, making it available in shared media buying platforms could risk data leakage or exposure.
By bringing programmatic buying technology deeper into the organization, brands have more control over how they activate their data and with which partners, if any, they share it. Rather than uploading sensitive information to a third-party ecosystem, brands can instead inform the bidder about the user segment or user value, and influence buying logic without risking data security.
When the buyer controls the programmatic logic, they can activate data in a way that best suits their specific needs – building custom audiences, using it in unique buying algorithms, enriching bid requests with unique first-party signals, and finding look-alike users at scale, for example. For owners of valuable data, in-housing widens their horizons for its activation.
2. Are better optimization options needed?
Every business is unique, which means every marketer has their own priorities, goals, and success criteria for evaluating the efficiency and effectiveness of advertising efforts. Most major media buying platforms, however, only offer a limited number of buying tactics, with these optimized to a common set of digital metrics, such as campaign pacing, clicks, downloads, installs, or the price of media.
If a brand’s advertising success is evaluated against outcomes more specific to its business than can be captured with clicks or post-click conversions, having more influence and control over the algorithms that are used to buy and optimize media could make a world of difference. This is another case where bringing programmatic in-house to exert more control over how the buying technology and intelligence behaves might make sense.
For example, it might be known that users who spend at least 30 seconds on the homepage convert twice as well as those who spend less time, or site visitors who hit four pages in a specific order come back twice as frequently as those with different viewing behaviours. Deploying bespoke algorithms that drive to those unique metrics will result in considerably better outcomes.
Alternatively, the ultimate goal might be measuring the incremental lift from overall advertising efforts. While there are solutions in the market today that help marketers understand attribution and incrementality, they rely on shared methodologies and common assumptions. Being able to design and deploy a framework based on its own success drivers ensures that the outcomes reflect what is valuable to that specific business.
3. Is there a direct relationship with media owners?
Media buying is all about finding the right audience, at the right time, in the right setting. And most media buyers already have a good sense of what inventory performs best for them and work to build strong relationships with those publishers to secure exclusive access to the supply or preferential pricing. With the imminent deprecation of the third-party cookie and heightened focus on user consent and privacy, solidifying those direct relationships takes on even greater importance.
In-housing programmatic offers media buyers an easy and direct way to strengthen those publisher connections in order to unlock unique value and drive greater efficiency, transparency, and performance. For example, media buyers may decide they want to strike direct contractual relationships with supply partners, setting their own payment terms and securing access to specific supply at certain times at fixed rates. They might also want to arrange first-party data sharing or matching agreements that allow buyers to find and target known audiences in a privacy-safe and consent-based way. The greater the number of intermediaries standing between the buyer and seller, the more difficult these special arrangements are to agree and manage.
4. Does your business require solutions that are more flexible?
Support of business-specific models and company infrastructure typically requires access to user interface or API customisation; while this can’t typically be supported by standard market solutions, it can be implemented in-house. Flexibility requirements can include: the need to control the privacy settings of different data sets and advertisers; integration with specific data providers, intelligence vendors or user identity services; and the creation and management of customized user identification metrics or business-specific parameters. These requirements all indicate that in-housing may make sense for the business.
In addition, media buyers may also need to report on metrics that are relevant to the business-specific KPIs across all their systems. Typically, the programmatic provider’s reporting data needs to be matched with an advertiser’s or agency’s internal first-party reports so that the media buyer can analyse the results and improve future campaign setup. Merging various formats of reports is typically a complex exercise that slows down the analytics process and decreases its efficiency. With an in-house set-up, the buyer controls both the bidder reporting data and first-party data and can ensure consistent reporting (supporting the combination of Google Analytics metrics in the in-house bidder reporting for example) .
5. Does the status-quo programmatic pricing model work for the business?
Along with transparency, the potential for cost savings is one of the main drivers for brands considering taking programmatic in-house. Most major demand side platforms (DSPs) charge buyers based on a percentage of media spend with additional fees tacked on for data usage, optimization, AI services, and so on. That means, if the fee is 15%, and the media spend is $100,000, the buyer will pay a minimum fee of $15,000 to the technology platform for using its system and services – regardless of how much media is being bought.
This model works against marketers who are buying high-value, high-CPM media like connected TV (CTV) or digital video. For these buyers, setting fees to the volume of media purchased makes more sense, and drives far greater efficiencies. Taking programmatic in-house allows media buyers to better control their costs because they decide on the pricing model that works best for their business. It also provides more commercial flexibility so brands can establish and maintain direct contractual and payment relationships with strategic partners (whether publishers, data providers, or verification services), providing yet another level of transparency and cost control.
There are some strong arguments for in-housing programmatic media buying, but also some good reasons why it is not always the best option. Determining which route to take requires businesses to understand their advertising set-up and goals and the potential gains versus trade-offs of the two options.