Proposal View

A quick context to this project; ADstruc, originally a marketplace for buying out-of-home advertising, made a quick pivot in direction and became a Request for Proposal (RFP) platform, allowing vendors to respond to advertiser requests with inventory that caters to their campaign needs.

We call this tool the Advertiser Proposal View as it presents the advertiser with a way to view and evaluate all proposals of units that they receive from outdoor vendors.

Our proposal evaluation tools went through 3 different versions in less than a year. It was a challenging time where we quickly learned and iterated. This is version 3.

I worked closely and brainstormed on this project with ADstruc’s former CTO, Sam Herbert, who created the high-fidelity visual mockups.

A typical out-of-home proposal consists of units and prices (usually in excel format) as well as "photosheets" (a PDF with pictures, a static map, and descriptive information). Our research focused on studying the existing processes involved in receiving proposals and evaluating them, particularly within a big agency’s environment. We heard feedback about our existing design, asked advertisers how their companies were structured, how they received RFPs, who processed and evaluated them and what made them buy.

The handmade mockups were created to be flexible and interactive. We tested them with people on the ADstruc team and advertising experts whose feedback was extremely insightful.

Here are some of our research learnings and how we used them to drive design decisions for the interface:

For large multi-market campaigns, advertisers evaluate each market separately. How we organized all that information was crucial. Since advertisers mainly care about location (neighborhood, zip code, proximity to landmarks or points of interest), it was important that our main focus be the map. Our market selector is a great way to quickly flip between locations (cities or DMAs) and control the evaluation process by limiting it to one market at a time. We built every other element around the fact that the map will always be on and visible, so advertisers can evaluate units individually while being able to see all other units (that they are potentially buying as part of their campaign) around it.

A crucial learning for us involved understanding how advertisers dealt with traditional photosheets. Many advertisers would print (yes, hardcopy!) all photosheets (often hundreds) of units received for a specific market, discard the ones they don’t like and keep the ones they would potentially buy aside. We aimed to reproduce that process by allowing people to hide unwanted units (which would remove them from the map and the list) or like (star) the ones they will request to buy. Starred units would appear differently on the map, so the distinction is clear between what has been "evaluated" and what hasn’t.

The left list pane is designed so advertisers can look at units within specific proposals, or display all units in all proposals at the same time - allowing them to get a better idea of what their campaign would look like. Units are tagged by a color that represents the media category they belong to (i.e. blue for billboards, red for street furniture). Clicking on a unit would open up an "info card", similar to the concept of a photosheet, making it interactive. The "info card" now layers on top of the left pane so the map stays in the viewport. Same thing happens when asking a question (sending a message) about a specific unit, or the proposal in general.

Apart from the typical bulk selection (checking items on the list), we also learned that outdoor advertising experts would know to discard or like units based off of location alone. Given the large RFPs we were expecting, we designed a way for advertisers to be able to draw a box on the map, around the units they wanted to select, setting the units as checked on the map and in the list view.

Finally, we knew that things like data layering (i.e. demographic data) on the map was in the future of the proposal view interface, so we made sure our designs were flexible and the map toolbar could support this feature, along with potential others.