INDUSTRY WISDOM

What does a programmatic trader do?

In plain English: What does a trader do?

“What do you actually do?”

I've been a programmatic trader for the last six and a half years, and in that time this question has arisen on a number of occasions with family and friends. So I tell them: a programmatic trader is the person who tells automated ad-buying systems how to bid in digital advertising auctions. We set campaign strategy so the right impressions are bought against an advertiser’s goals and budget.

But obviously, there's more beneath the surface. So here's my attempt to explain, as simply as possible, what that role involves, how programmatic ad buying works in practice, how campaigns are set up and optimized, and the qualities that make a strong trader.

The background of programmatic advertising

There are some things you need to know before we get into the thick of it (I'm also going to be generalising for ease):

  1. Almost every ad that you see on the internet was chosen specifically for you.
  2. In most instances there are a number of different advertisers who want to show you their ad. So, in the milliseconds while your webpage is loading, an auction takes place in ad exchanges. This is called real time bidding (RTB), and whoever is willing to pay the most gets to show you their ad.
  3. Billions of these auctions take place every day, so it's impractical for a human to physically bid to show you an ad. Instead, computers do the bidding in an automated way, with advertisers using demand-side platforms (DSPs) to buy ad space and publishers using supply-side platforms (SSPs) to sell ad space. This is called programmatic buying.
  4. Cookies (crumbs of data all over the web that ‘remember' your behaviour) mean advertisers can see if a visitor to their website has previously been shown one of their ads.

Where do programmatic traders come in?

Traders are the people who work within programmatic advertising platforms to instruct the computer how to bid in the auctions. Tech platforms provide traders with a myriad of options on the parameters they can choose for those instructions. For instance, I can make an instruction like:

“Whenever a user appears on a mobile device, on thetimes.co.uk, between 9am and 10am, in London bid £1. Keep bidding until we've spent £100, then stop.”

We would refer to this as a campaign strategy. It's also part of campaign setup within programmatic campaigns, where the trader defines the target audience and can adjust targeting and bidding inside the ad platform. It's one of the many ways in which the ads (in digital advertising they are referred to as ‘impressions') can be bought for this campaign.

So, an advertiser gives a trader the budget, timeframes, and a brief – for example, £1,000 to be spent next month targeting families interested in holidays. The trader will then set up a number of strategies that achieve that aim and provide the advertiser with £1,000 worth of display ads, while helping manage ad spend across ad campaigns and tracking impressions, clicks, and conversions to assess campaign performance. In broader digital marketing, modern measurement in programmatic advertising can also include e-commerce outcomes or closed-loop attribution that links digital ads to store traffic, with privacy-first measurement solutions becoming increasingly important.

So programmatic buying is as simple as that?

Unfortunately, not. Advertisers don't want to spend just for the sake of it. They want to get results. And those results will vary by advertiser and by campaign. The success of the ad buying process depends on optimization inside demand-side platforms, and measurement and verification tools, not just budget alone.

Each campaign comes with a KPI or goal in mind. To use the above example, the advertiser probably wants the people who see the adverts to book holidays for their families.

So the money the trader is spending comes with a target, for instance, we expect at least 20 holiday bookings for the campaign. That would mean that for the £1,000 spent, the advertiser doesn't want it to cost more than £50 per holiday booked. We would call this the CPA goal (cost per acquisition or action).

Once the trader has set up their campaign strategies, they can use cookies to work out which strategies are resulting in people booking holidays, and to what degree. They will look at campaign performance signals like impressions, clicks, and conversions, and use cross-channel analytics or incrementality measurement to judge whether the ads actually drove additional results. Their aim is now to continually tweak the strategy to reduce the CPA. We call this optimising.

Optimizing means adding new strategies based on what works the best. For instance, only targeting one website that works really well and bidding more to show adverts there. Artificial intelligence can help analyse those signals and improve targeting, and 76% of marketing decision-makers believe AI will enhance targeting. It also means removing wasteful spending, for instance not serving ads at certain times of day that we can see aren't working well or to users displaying certain online behaviours that mean we don't think they're likely to buy a holiday. It also means protecting ad spend, because ad fraud wasted 22% of paid ad spend in 2023, so traders use quality controls as part of optimization.

It wouldn't be unusual for a campaign to have dozens of strategies – I remember one particularly painful campaign with over 1,000 strategies – so it's the trader's job to determine what they are and to optimize them across programmatic platforms, using a mix of analytics and dynamic creative tools as well as DSP decisions, to meet the client's goal. (And to make sure we outperform our competitors who are trying to do the same thing.)

What makes a great programmatic trader?

There are a few qualities that make for great traders (and one big misconception).

Misconception first: You need a maths degree or similar to be a trader.

The reality is I've never used advanced calculus when trading a campaign. It really doesn't matter if you dropped maths at the earliest opportunity and never went to uni. You can still be a great trader. That said, you do need to feel comfortable with numbers. There's a fairly constant stream of arithmetic and that's daunting for some people. But if numbers don't scare you, you'll be fine.

Other qualities that make for great traders…

Conscientiousness and diligence are hugely important. It's a job where you spend a lot of real money (which is also not your own money). Mistakes will be costly, so we look for people with high levels of attention to detail and consequently lower error rates. We also look for resilience, because eventually everyone does make a costly mistake so you need to be able to bounce back quickly.

The job requires making hundreds of optimizations to campaigns per week, so we look for decisiveness. There isn't time to go back and forth on one decision for several days.

And we always look for people who display innovation in their work, and who will push the boundaries of what we can do. We're proud to be at the forefront of programmatic and we can only stay there because we have great traders who want to stay ahead of the curve.

We also want traders to be chock-a-block full of drive. Without that, nothing gets done. The autonomy of the job absolutely requires it.

Lastly, there's the unquantifiable X-factor. A great trader looks beyond the numbers. They find establishing the ‘what' easy, and are always striving to find the ‘why'. They are using the data around them to search for answers, sometimes to a question that the advertiser hasn't even asked yet.

A good trader hits CPA targets. A great trader can give an advertiser the insights that help drive their whole marketing and their business strategy forward.

At MiQ, we aim for great.

If you would like to find out more about trading roles at MiQ, check out our careers page.

share: