The way we buy digital ads has changed fundamentally.
In the past, you purchased digital "real estate" or "inventory" to promote your products or services.
Now, you are buying "audiences."
This migration to digital advertising places a much greater emphasis on the technological mechanics behind the process - these mechanics are complicated and are based on real-time data signals sent to the various algorithms that serve specific creative to specific users in fractions of a second.
There are many conversations in the industry around these systems, but most of those conversations seem to be broken.
The Major platforms, including Google and Facebook, recycle basic definitions for programmatic media buying to educate their consumers, while vendor and supply-side platforms (SSPs) produce content full of advertising jargon and biased software pitches that seem to be educational guides.
To effectively navigate this complex and highly fragmented digital advertising ecosystem, one must break through the marketing jargon to gain insight into how these platforms operate.
A brand that buys media at scale will require a thorough understanding of service models, platform access, and procurement friction.
Creating revenue through programmatic media buying
The short message here is to focus on execution, not technology.
Programmatic media buying can effectively create revenue growth at scale.
It is important to understand that programmatic media buying is not a traditional automated machine but can be very automated when you have an appropriate access model in place, a solid data infrastructure, and disciplined operations.
You must expect friction at some stage.
The minimum spends required to purchase inventory, the limitations to have this inventory attributed back to the individual purchase, and strict specifications for creative will create the pace for your campaigns.
The most intelligent brands do not simply seek out the lowest cost of technology, but instead seek to find the most suitable operational fit to maximize their internal resources and risk tolerance.
Your company needs a programmatic advertising agency
Buying digital media at scale means access to Demand-Side Platforms (DSPs), Supply-Side Platforms (SSPs), and many fragmented data exchanges.
This is often where the execution fails.
The operational overhead created by daily trafficking, brand safety checks, and bid optimization overwhelms these companies, leaving them little time for real-time bidding, which is a very demanding industry.
Conversely, a growing number of companies choose to leverage specialized programmatic advertising services instead of doing everything in-house.
Collaboration with programmatic specialists allows companies to benefit from the expertise of other companies in order to bridge the gap between high-level strategy and the granular execution of their campaigns.
The workflow benefits of working with external programmatic experts go beyond just access to their software.
Working with external teams allows you to access a proven workflow model.
Partnerships with external teams also result in the expertise gained through having a proven model that includes a track record of successfully negotiating with programmatic media sellers on minimum spend thresholds, managing aggressive frequency capping, and creating campaign names to promote greater understanding in report generation.
The operational reality of campaign management
When reviewing polished vendor slide decks, it is rare to see the day-to-day effort behind campaign management.
Media buying is a systematic approach to set up and iterate your media campaigns.
The term "media buying" defines the work done to prepare campaign materials, and with the amount of time required in setting up campaigns, it is impossible to do so without a lot of manual effort.
Data setup and data onboarding
Before a single bid is placed, the first step of setting up the audience for your campaign is to gather and structure the relevant audience data.
Audience data consists of three major components:
- First-party Customer Relationship Management (CRM) lists
- Privacy compliance based on the current local regulations
- Defining the audience signals needed for your campaign
A failure to gather and onboard accurate audience data causes the organization of supply paths based on audience parameters that are not defined.
As a result, the Demand-Side Platform (DSP) will spend its budget roughly across many low-value placements when the audience parameters are not defined.
Trafficking and quality control
Making one mistake while setting up your ad operations could lead to an enormous financial loss.
The trafficking process for your campaigns includes uploading dynamic creatives, mapping conversion pixels, and precisely configuring line items within the DSP.
Without exception, organizations must develop a rigorous quality control checklist; a simple omission of a single brand safety exclusion term could cause the ad to appear next to or in proximity to advertising or news of controversial topics.
The learning stage of a programmatic media buy is the process of testing and discovering which of the available media and audience segments generate conversions, while spending a stipulated budget.
During this stage, it is important to continually improve returns on advertising spend (ROAS) by reducing underperforming inventory, adjusting how much to bid on a device level and moving money between different campaigns that are running concurrently.
Media buying methodologies
The type of media you purchase from an advertising technology provider determines how much you will pay for your media purchase.
The programmatic industry has three primary ways that companies buy media:
Self-serve platform licensing
With this method, you purchase a Demand Side Platform (DSP) from a provider, sign a contract with the provider and then create your campaigns on the provider's system.
The upside of this method is that you have full transparency into how much you are paying for the platform's service fees and how much it costs to place your bids.
The downside is that you need to have experienced programmatic traders in-house to manage the build-out and optimization process.
The learning curve can be long, and if your company does not meet the minimum spend requirements set by the platform, you may lose your access to the software.
Managed services
With a Managed Service approach, you give a budget to a managed service vendor or a Technology Provider, and they will completely run the programmatic buys for your business.
This method does not require any in-house expertise for buying media programmatically.
The downside of this approach is that it is extremely difficult to understand what you are actually paying for, since Managed Services often include the pricing for the Media Cost (the actual cost to purchase ads), the Data Fee (the cost for the data you will use) and Technology Markup (the fee for the third-party vendor's technology).
Therefore, when you are trying to audit what you paid for the ad inventory from the vendor, and what you paid for the services provided by the vendor, you may find this to be nearly impossible to identify.
Agency execution
In this approach, you engage a third-party agency to assist in the building and execution of a DSP on your behalf.
The Agency provides you with the personnel needed to manage the DSP, trains you on how to use it.
As a result of working with a larger, agency partner, you have access to their collective purchasing power and existing technology partnerships.
The downside to creating a relationship with an agency is that you will typically pay for either an ongoing retainer (often with a standard agency fee) or a commission based on a pre-determined percentage of the ad spend.
The costly truth and risks of programmatic buying
The primary risk with programmatic buying is the risk of systemic failure. To mitigate the risk of systemic failure, it is essential to identify where things can go wrong.
The costs involved in the learning curve
In order to establish and build baselines through machine learning algorithms, a programmatic buyer must generate enough volume to establish these baselines.
As a result, if a programmatic buyer has a budget that is relatively small for testing purposes or if their targeted audience is too small geographically, the demand-side platform (DSP) will remain in the learning phase.
A programmatic buyer will pay higher-than-market costs for random impressions, without generating any statistically significant conversion data.
Attributing the ambiguity of user interactions
As a result of the way the growing open internet, generally, and the increasing popularity of mobile, have created an increasingly complex process by which to track user interactions with brands on the open web.
Consequently, the cookieless approach has not yet matured sufficiently across the advertising ecosystem to be able to rely solely on last-click attribution.
As such, a programmatic buyer will likely have their programmatic campaigns display as a disaster if reliance solely on last-click attribution models is the only option available to them.
As a result, savvy marketers utilise blended measurement models and incrementality testing to determine the most accurate measure of their programmatic advertising success.
Sparse inventory and fraud
The open exchange is flooded with made-for-advertising (MFA) websites specifically created to host paid advertisement(s).
Without stringent brand safety filtering, advanced software solutions for pre-bid fraud detection, and tightly-circumscribed private marketplace (PMP) contracts, a programmatic buyer can expect to see their budget leaking out in every direction, with the constant threat of non-human traffic and invisible inventory on the open web.
The evolution of the media landscape
The ad ecosystem is rapidly evolving from using desktop display banners as the primary channel for advertising to utilising new channels to absorb the majority of ad spend.
CTV is quickly becoming the most dominant segment in the advertising market
Instead of purchasing traditional linear television advertisements upfront, advertisers place real-time bids for streaming television inventory.
Digital audience-targeting techniques are implemented in a linear way and delivered via high-impact visuals; thus, measurable performance data is also available.
Retail media networks
Retailers are now utilizing their vast amounts of first-party data to create their retail media networks by monetizing this information.
As a result, consumer brands can identify and target their customer base programmatically based on actual purchasing behavior instead of inferred browsing behavior.
The ability to pinpoint individual shoppers with precision is unmatched.
The use of AI and autonomous bidding
Artificial Intelligence (AI) is taking on an increasingly greater role when managing Demand-Side Platforms (DSPs).
Bid shading algorithms can identify the lowest possible winning bid for a given impression, thus reducing the amount of money spent by advertisers.
Predictive AI can be utilized to model the missing data regarding conversions within privacy-first contexts; even when direct tracking is blocked, predictive AI enables advertisers to continue optimizing their ad campaigns.
When programmatic advertising should not be purchased
While programmatic ads have scale, they may not be the right solution for every advertiser.
Here are two examples of business models that should not purchase programmatic ads:
Ultra-Niche B2B targeted marketing
If your total target market consists of exactly 400 executive-level decision-makers, broad programmatic (algorithmic) buying does not work.
The best results will occur through Account-Based Marketing (ABM) that relies on customized Publisher Direct Deals and highly-targeted LinkedIn campaigns instead of employing a more generalized approach on the open web.
Shoestring budget for media purchases
Demand Side Platforms (DSPs) are designed for companies with large media budgets, and thus they are not suitable for advertisers with lower resources.
When starting with a low monthly media allocation (less than $2,000) you should use traditional media channels until your budgets allow for enough funding to properly support the algorithm learning stages of spending on show through Distributed Programmatic Buying (DPB).
Evaluating your programmatic advertising agency
The selection of your partners should go far beyond their "client" logos.
When choosing your partner, you will need to know how they are pricing their work. Ask to see their pricing model.
Do they work on a flat fee basis; on a percentage of the media spent; or do they take an undisclosed margin that is hidden in their media?
Hidden costs hurt campaign resources.
Request their standard onboarding workflow. A credible programmatic advertising partner will share their onboarding processes and timelines.
A credible partner will document and make available their processes for generating pixels, creative quality assurance, and ensuring brand safety.
They will also be open about how they attribute their results to their media efforts based on true incremental results as opposed to vanity metrics.
Conclusion
Scaling up requires automation; however, to achieve success through automation requires aggressive management.
Entering the world of programmatic media allows advertisers to reach far beyond their traditional media capabilities and operational complexity.
Advertisers who properly connect their first-party information from the CTV/OTT, Retail Media, and Digital OOH spaces are now able to access and reach an incredible market.
While all technologies have their pros and cons, the team who manages the technology is as important as the underlying technology itself.
Stay laser-focused on your access model and be extremely disciplined in maintaining your brand safety and full transparency on how your budget is being allocated.
Frequently Asked Questions (FAQs)
What is the minimum buy required for programmatic campaigns?
The platform minimums for programmatic campaigns differ significantly from one programmatic technology provider to another.
For example, if you want access to the top tier DSPs and self-service capabilities, you typically need to spend anywhere between $10K to $50K monthly.
You may, however, be able to leverage pooled accounts through your agency partners to avoid those high platform minimums.
How is programmatic different from the Google Display Network (GDN)?
On a broad level, Google Display Network (GDN) is essentially one single ad network offered by Google.
However, programmatic advertisers leverage Demand Side Platforms (DSPs) to access multiple global ad exchanges in real-time and thus, acquire a much larger inventory selection, utilize advanced 3rd party data for targeting, and have access to more advertising format options (streaming video and native).
Are a Demand Side Platform (DSP) and Supply Side Platform (SSP) the same?
Not exactly.
Advertisers utilize a Demand Side Platform (DSP) to purchase ad space; whereas Publishers utilize a Supply Side Platform (SSP) to sell ad space.
DSPs and SSPs communicate with each other in milliseconds to allow for a successful transaction.
Is Connected TV (CTV) purchased programmatically?
Yes.
Connected TV (CTV) inventory is primarily purchased & sold in a programmatic fashion.
Programmatic advertisers purchase advertising slots from connected television providers (streaming media and smart televisions) via DSPs, using 1st party data to facilitate precise targeting at the household level which traditional television advertising cannot offer.