It will be a decade since the first programmatic auction took place this year and it can be stated that since then programmatic technologies and the ecosystem as a whole have well and truly prospered.

In 2017, programmatic display spending reached almost $33 billion globally.

According to eMarketer, in the UK alone, programmatic accounted for 79% of all digital display ad spend in 2017 (up 23.5% from 2016) and is expected to climb to 82.4% in 2018 and 84.5% in 2019.

eMarketer senior analyst Bill Fisher outlined the reason for this growth by saying “The programmatic ecosystem is growing because it’s maturing. This maturation is leading to better practices, better behaviour and better transparency.”

The programmatic ecosystem that we currently operate within can be considered as being mature. There are already advanced technologies such as header bidding and there is more focus than ever on factors such as transparency, brand safety, ad fraud, viewability and of course, performance.

The increased importance of these factors not only helps to build advertisers trust in programmatic but also increases the efficiency of media budget spent through programmatic, which in turn provides better performance, greater investment and further growth for the industry.

As a result, the point of execution, the way in which programmatic campaigns are traded and optimised has become increasingly important to ensure all of these important factors are taken into account.
Trading and optimising programmatic campaigns is complicated and carries a lot of risk if not executed correctly. In addition to the important factors mentioned above there are more variables, formats and methods of accessing supply and audiences than ever before, which has made efficient programmatic trading become even more difficult to achieve.

The Programmatic Advisory recommends the following to help you minimise risk and maximise efficiency across your programmatic campaigns.

Golden rules of optimisation

1) Make sure you have enough data – don’t optimise your campaign too hastily, always allow sufficient time to gather enough campaign data so you’re able to take a data driven approach and increase the likelihood of your optimisations being successful.

2) Don’t make too many changes at one time – stagger your campaign optimisations and only carry out those that you think will be most worthwhile first, do not carry out too many optimisations at one time as they will become difficult to manage and their success difficult to track.

3) Don’t over or under optimise – There is no set rule for how frequently you should optimise a campaign, however carrying out significant optimisations once a week and monitoring campaign performance and pacing daily is usually a good benchmark, it’s important to allow for sufficient time, budget and data collection between optimisations.

4) Monitor changes and results – it’s important to keep track of the optimisations you’ve made to a campaign, creating a campaign change log in excel usually works well. You may find that removing a certain set of domains from targeting has had a detrimental impact on your campaign, with a change log it’s easy to identify changes and revert back if needed.

5) Adopt a test & learn approach – always test optimisations in a controlled manner, it’s never worth allocating too much budget to a strategy that hasn’t already been tested. Apply a small portion of your campaign budget to the test strategy, if it works well, allocate more, if it doesn’t work well the impact on your overall campaign performance won’t be as detrimental with a small budget.

Bulk Optimisations

1) Whitelist – create a whitelist for each market activity will be run in and ensure that it has been checked and vetted numerous times before implementing. This may be a time consuming process initially but it’s very worthwhile in the long term as it helps to boost performance by focusing spend towards quality domains, limiting delivery against long tail domains (which add little in the way of incremental targeting opportunities) and helping to prevent fraud.

2) Brand safety – most DSPs provide pre-bid brand safety settings that allow for the blocking of content categories that aren’t considered brand safe such as gambling, alcohol or violence. It’s important to ensure that firstly, you are aware of the capabilities/limitations of the brand safety technology provider and secondly that the relevant brand safety settings are applied across all campaign strategies.

3) Day parting – many DSPs allow programmatic activity to be paused or scaled up during certain times of the day. This helps ensure that a greater proportion of the campaign budget is utilised during times of the day performance is strongest (e.g. 8pm) and less campaign budget is spent wastefully during times of day that performance is at is weakest (e.g. at 3am). Day parting is a simple, yet powerful tool used to help enhance the efficiency and effectiveness of programmatic campaigns.

4) Audience exclusions – always ensure that the relevant audience exclusions been applied to your strategies to help maximise efficiency and minimise wastage of campaign budget. For example, if you’re running prospecting strategies looking for new customers who haven’t engaged with your brand, it would be worthwhile excluding anyone who has visited any page of the brands website from targeting.

If you found the above useful and would like some more suggestions as to how to improving your programmatic trading operations please reach out to The Programmatic Advisory at hello@theprogrammaticadvisory.com.