One of the most important aspects of programmatic advertising is the ability advertisers have to optimize their campaigns. These optimizations can make the difference between an okay campaign and a great campaign. This is why it’s so important that you choose a partner like Megagon Ventures who has real-world experience working directly with DSPs, optimizing campaigns for some of the biggest brands in the world.
In this article we’ll touch on two main points.
Taking a page out of a Simon Sinek book, let’s start with why. As we stated above, an optimized programmatic campaign can turn an okay campaign into a great campaign. The unique capabilities of programmatic allow it to let advertisers target their users as granularly as they wish, based on inventory, channels, context, data, geo, etc. In each of these categories there are literally endless amounts of permutations of sites, users, creatives, browsers, and more. That’s what you get with programmatic out of the box–endless opportunity.
And unless you have a partner who can make the most of that opportunity that programmatic provides through consistent optimizations, your campaigns will underachieve. Optimizations allow you to pay the right price for the right message to the right person at the right time. Site X on a mobile phone on weekdays might be performing well, and you may want to keep your bids high to ensure you keep winning on that site. Or maybe your CTR is static, and you need to get your CPC down so you lower your bids on that same combination. And maybe all you really want is to buy more of that inventory combination at your current price. Skilled media buyers at Megagon Ventures know which strategies to implement when, making sure that you’re getting the maximum results from your campaigns.
If something as simple as blocking low performing sites, ad formats, SSPs can increase your KPIs by 30+%, imagine what traders with years of experience can do with your campaigns.
There are three main phases when media buyers (traders) optimize campaigns.
In the planning phase, users have the option to “pre-optimize” their campaigns using data and insights from past campaigns or from media planning software like The Trade Desk’s. Let’s say for example you had low CTR on abcdefg.com all throughout your last campaign. Before even beginning the subsequent campaign for that brand you could go ahead and block it, ensuring no impressions are bought on that site.
Once a campaign is live, traders ensure that spending ramps up, and monitors the overall health of the campaign for the first few days. Once enough data is collected (usually within 1-2 days) they can start analyzing the bid data, creating hypotheses, and experimenting. They know when to be aggressive, and when to make small changes that nudge a campaign in the right direction without losing scale or risking decreased pacing. They optimize on a cadence, repeating this process every few days throughout the life of a campaign to ensure the highest possible results.
Finally, after a campaign, traders delve into the data from a wholistic perspective, looking for high-level trends and insights that can be used as inputs to subsequent campaigns. These insights serve two purposes. First, they help the advertiser and brand learn about what worked, and to avoid next time. Second, they provide guidance to planners of future campaigns to make sure campaigns start off with as much prior learning as possible.
Now we really get to dive deep. In this final section, we’ll talk about a lot of the different types of optimizations our traders can do. We’ll introduce a few terms and provide a couple examples. We don’t blame you if you don’t stick with us for this portion. We do all of this so you don’t have to. We just want you to be assured that we know our stuff :) There’s no way we could ever list all our knowledge or examples here, but we hope it gives you a glimpse into how traders think and how they work.
The first term we’ll discuss is rails. Rails are the settings you predefine in your campaign, similar to railroad tracks that keep a train heading in the right direction. These are usually fixed. Some examples of this could be your targeting model if it’s pre-defined or non-negotiable, add-ons like fraud detection, brand safety, and contextual key word targeting. It’s any lists you choose to allow or block from the start. The rails are the initial parameters that a campaign has to work within.
The next term is optimizations. This term specifically refers to categories and items that can change, or be improved. For example, even if you have a site list as one of your rails, we can still optimize sites within that site list.
Slightly changing gears, we’ll now define bids. A bid is how much you are willing to pay for an item. Just like on eBay. You already knew that. A bid profile then would be defined as a certain price your are willing to pay for a combination of items (user who loves shoes, at 11 am on a Tuesday, in Texas, on an Android tablet).
Bid factors, or bid multipliers, are ways to increase and decrease your bid, make popular by The Trade desk. These allow you to take an item or an entire bid profile and set a price above your base bid that you’re willing to pay. For example, normally you’re willing to pay a $25 CPM for CTV ads, but you see a user who is interested in taking a vacation during primetime, and so you’re willing to pay 1.25x of your base bid, or $31.25 for that bid profile.
The next optimization term is grain, meaning the category or item that is optimizable. For example, site, browser, device, operating system, time of day, day of week, etc. are all considered grains. When you look at what you want to optimize you start looking one grain at a time, and then depending on the goals, can build models of multiple grains.
Some of the grains you can optimize on include:
Lastly, when you start optimizing multiple grains grouped together, you are creating a multi-dimensional bid. For example, you are trying to increase CTR and so you look at sites. Then you layer on sites by browser, by device, by OS, and by time of day. You create a bid model with individual bid factors for each of these distinct permutations.
Thanks for sticking with us to the end. Sorry we geeked out, but we hope you have a better idea of what’s going on in traders’ heads. We can’t speak for everyone, but we know our traders eat, sleep, and breathe programmatic advertising. In fact, we think it’s so important that we require our traders to have at least 2 programmatic advertising dreams a month. haha jk.
That would probably be overkill, but it would be awesome.
If you’re looking for a great partner who can help elevate your advertising, take the next step now and get in contact with us!
