Grigoriy Misilyuk
Oleg TymchyshynThe TeqBlaze team is here to deliver another gentle heartbreak: “it’s not that simple.” Spoiler: removing supply-chain intermediaries won’t magically boost your platform’s performance. But there is something far better — real engineering experience, wrapped in a slightly sarcastic article. Enjoy.
Our core point remains the same: platform owners should invest in intelligent supply path management. It’s a tougher path than just avoiding intermediaries because it forces you to examine every component of your process — and their value shifts depending on the situation and the scale. Add the human factor and the occasional engineering… let’s call them “future-feature improvements,” and the story only gets more interesting.
Boring? Definitely not.
What is supply-path optimization?
This section is unavoidable because agreeing on terminology at the start is better. Supply-path optimization (SPO) is the process of researching, evaluating, and changing the supply chain of advertising inventory. You can read more about SPO on TeqBlaze's blog; I will provide only a shortened description here.
Publishers supply advertising inventory through the supply-side platforms they work with. SSPs might connect advertisers directly (PMP, API, DSPs) or through ad exchanges via OpenRTB. When ad exchanges come into play, the SPO is required. The reason is that the number of traffic resales and duplicate bids can harm the monetization goals of suppliers and advertisers, who also want the best inventory at the most favorable prices.
The SPO takes place at the level of ad platform owners who can use the functionality of the ad exchange on their platform. Supply chain optimization tools are available to all platform owners who have built their programmatic businesses based on TeqBlaze's white-label technologies. The SPO opportunities are the guiding star for ad inventory suppliers and the ad exchanges connecting participants. We may also discuss bid route optimization features for any combinations of roles, including beyond those mentioned above.
So, SPO is a necessity for large ad ecosystems. However, I recommend gradually optimizing by eating an elephant in pieces. The second argument is that optimizing is challenging when bidding on significant traffic. The process of testing theories becomes more expensive.
Focus on intelligent management, not just reduction
What an image, huh? It looks like a battlefield. Well, no one said that programmatic would be easy. Let me explain what's interesting here.
In addition to direct interaction with the DSP, a request to sell ad inventory is initially transferred to two ad exchanges.
Each of these exchanges gives access to unique advertisers, which benefits the supply side.
However, each of these exchanges also sends the request further to other exchanges; these steps are marked with red crosses and are blocked as less profitable.
So, we performed a primitive supply-path optimization: we took the main advantages of intermediaries (rich demand), protecting ourselves from the potential disadvantages of such a partnership (reselling bid further).
But is this the best option of all? I doubt it. The best option is usually unexpected. This is a very rough application of the rules: we had to clearly define the effectiveness of each path and node in the advertising chain in the dynamics to block further transfer of the bid surgically. Sometimes, a bid that has been resold three times is still a very profitable deal.
Transparency builds trust, and trust makes money
In programmatic advertising, many layers often make it difficult to see where your budget is going. Transparency has been a recurring issue for a long time, and a clearer view of how ad dollars flow through the supply chain is essential to fixing these problems.
It all starts with a comprehensive analysis that requires real-time execution. My advice for optimizing supply routes relies on two fundamental points.
It is best to involve a separate ad operations specialist in such an analysis. In the free time from optimization, the AdOps engineer can perform other tasks related to your platform's performance analysis. The best option is to create conditions for the specialist so that all working time is devoted entirely to research and development activities for the success of your business.
Invest in developing tools for analyzing supply chains and trade performance, including particular chains, nodes, and partners. The more advanced the analysis tool, the easier it is to make optimization decisions.
These two conditions guarantee sufficient transparency to optimize supply chain settings. I don't have any particular words to emphasize once again the importance of a person with their finger on the pulse of the performance and a comprehensive tool for analyzing this performance. This is as important as it can be. Before making complex optimization decisions, a specialist can transparently see the picture of thousands of intersecting routes and give specific advice on how to improve bidding. The first steps in such optimization usually produce tangible results immediately, an unforgettable experience.
Avoiding duplication and inefficiency with an eye for tech
You're probably thinking, "Okay, so it's just about cutting out redundancy, right?" Well, not exactly. Duplication in the ad supply chain is like that annoying email thread with 20 people CC'd — a bit unnecessary and often counterproductive. But here's the catch: not every duplicated bid is inherently wrong.
Below is a short list of cases that you should avoid when optimizing your supply chains:
When the same inventory is auctioned at different prices among the same partners
Excessive bid requests for the same inventory to the same buyers
Auctioning remanent inventory alongside premium advertisers
The presence of more unauthorized nodes in the chain than authorized ones
Having more than two nodes in bid responses that are marked as direct traffic
The trick is to avoid the wrong kind of duplication that eats away at your margins without shooting yourself in the foot by blocking out all the good stuff. Advanced tech approaches can help a lot here. Suppose you're not planning to deploy machine learning models for route analysis and maintaining efficiency. In that case, you'll probably leave money on the table when the future comes.
Think of it as filtering noise from a great music track. You want to strip away the excessiveness that is harming the performance while preserving what adds value. Sure, manual decisions can get you so far. Still, the best results often come from a hybrid approach — combining tech-powered insights with human experience to weed out inefficiencies and avoid pointless bid reselling. That's how you maintain value without causing a collapse in your revenue chain.
Please, let me emphasize again, do not rely on technical solutions as a panacea. Your base in this matter is constant manual monitoring and control. Even if engineers have developed tools for SPO, few can predict how they might be misused in an automated application.
I like this meme with Leonardo DiCaprio. Because it's so real and relatable, imagine the situation: you see traffic coming to a website that has been blacklisted. Or not to the website but to some intermediary in trade. Why is this happening? Because the miracle of engineering can be wrong, and it is tough to achieve perfection in this matter. Sometimes, it seems like trying to dress nicely once and waiting for this outfit to be appropriate and match the weather during the year.
You can see how it happens: a node was added to the blacklist, but a particular chain was prioritized because it made a good profit in this part of the day, and several such rules were applied. There were even some that were mutually exclusive. Of course, machine learning algorithms tried to do the best they could, and this error is infrequent and was immediately eliminated by experts who refined the model. But I'm telling you all this because a professional would never make such a mistake. Because people understand the essence, while any automated model relies solely on metrics. The miracle is in people: they create and maintain it and are the miracle itself.
Going beyond flat solutions with a balanced SPO strategy
The best strategy isn't about cutting paths and hoping for the best. It's about flexibility, considering both direct and indirect approaches, and offering more adaptability in an unpredictable market.
Why limit yourself to one method when a combination can yield better long-term results? Direct approaches like cutting inefficient intermediaries can surely bring immediate value. However, indirect strategies, such as exploring new exchanges or prioritizing lesser-used routes, can reveal hidden opportunities.
It's like driving — sometimes, taking the long route might get you to the point faster or, at the very least, open up new paths you hadn't considered.
A hybrid strategy allows you to tailor your SPO efforts to both short-term goals and long-term market shifts. The ad landscape is far from static, and what works today might not work tomorrow — you need that adaptability. So, think about balance instead of chasing a "perfect" SPO solution. When direct cuts don't deliver enough value, your indirect strategies can step in to optimize for future gains. The result? You're not just reacting to changes in the market — you're driving them.
Conclusion
Hope this article was worth your time — here’s the essence in one place.
Effective optimization starts with rigorous performance investigation, grounded in solid technical tooling and handled by people who actually know what they’re looking at. And yes, traffic resale can still be profitable, so don’t box yourself into rigid assumptions about bidding partnerships. Creative testing — the kind most teams overlook — remains one of the most undervalued skills in programmatic.
Investing in machine-learning models will gradually make supply-path optimization more predictable, but keep in mind: automated logic will never fully replace human oversight. Real operators still carry the weight of tough decisions — no drama, just facts.
As always, the TeqBlaze team is open to thoughtful conversations. If you have thoughts on supply-path optimization, drop us a message. See you in the next article.
