This is platform's core function.

First, we gathered the campaign data requirements on a serie of stakeholder interviews. I documented our findings in a mindmap.

Next, I sketched out the campaign creation flow to get stakeholders' confirmation on my vision of the process. Many DSPs have the campaign setup process going from general to granular, particularly setting up media at the very end. At LightAd, that's not the case. Here the stakeholders wished the media to be pre-moderated before setting up a campaign, which allows to launch campaigns almost at instant. This means media have to be uploaded and get SSP approval in a separate flow, and campaign creation starts from picking already approved creatives.

With the core flow confirmed, I assembled the campaign creation form.

User has to pick one or several media pieces from his asset library. By default, the platform suggests only the approved media, though the user can select an unmoderated one as well. In that case the platform throws a warning message.

Campaign has to have specific start and end date. User can specify hours he wants the ads to run weekly.

For space-saving reasons flighting section utilizes conditional disclosure method. So the big dayparting block is revealed upon user request only.

Frequency capping, pacing are hidden under Advanced budgeting accordion - again, following the progressive disclosure method.

Custom audiences are user-created lists of identifiers (IP, IFA, etc.). Similar to media, these are pre-curated items, made before the campaign creation. Not only it allows to use reliable data only, but also to distrubute work between users: some focus on audience import, while others work with campaigns tuning only.

Due to SSP requirements, age options come in pre-made intervals.

Geotargeting supports several input modes: contextual search, manual selection from a list and bulk ZIP-list upload. On top of that, any location can be excluded from targeting, which makes LightAd geotargeting fairly precise.

This observation allowed us to develop five consistent flows already. The sixth one for dynamic native ads happened to be very different from the others. Whereas other media types arrive as holistic ready-to-use piece, dynamic native ads are being made on the go out of separate bits of information provided. So in this case it's not a bad thing to step out of the pattern.

This is an advanced-level feature which allows to get the targeting precision level that cannot be achieved with standard age/gender/geo parameters only.

We identified 5 main identifiers that are supported across SSPs. In order to process user's data correctly, we require to specify the audience type first, and then upload the list (usually it comes in a csv file). Copypasting is supported as well.