The ads landscape at Meta

Meta offers advertising solutions to help businesses reach new customers across all its platforms, including Facebook, Instagram, Messenger, and WhatsApp.

Businesses of varying sizes use Meta platforms to promote their business, both organically and by leveraging ads.

We broadly categorize these advertisers across 3 segments:

• Basic – comprises people with little to no marketing expertise, for whom advertising is mainly a part of their job.
• Intermediate – includes those who are more comfortable with advertising but lack the time or resources to become experts in the field.
• Advanced – consists of advertising professionals who work at agencies or as dedicated digital campaign managers at medium- large sized businesses.

In order to support these users of varying sophistication, Meta has two categories of tools:

• Lightweight ads (LWI) – a lightweight tool designed for novice advertisers with minimal to no marketing experience, targeting primarily the small and intermediate user segment.
• Ads Manager – This is the full featured, robust tool with advanced feature set, which is aimed at more sophisticated advertisers across the intermediate and large segments.

Lightweight ads (LWI)

My team owns the Lightweight ads tool(LWI). This is a simple advertising tool for small businesses to create, launch, and manage ads. It's accessible within the Facebook app and also on web and mobile browsers, but it enables creating ads that can show up across all Meta surfaces including Facebook, Instagram, etc. This product has been de facto solution for small business ads at Meta for 10+ years.

As stated earlier the Basic advertiser is the primary persona our tool is deisgned for. Let’s zoom in on this user to understand their needs.

Zooming in on the Basic advertiser

The Basic advertisers are small business owners (1-5 employees) who wear every other hat in their company, from housekeeper to accountant and marketer. They are jewelry and fashion designers, bike repair shop owners, IT consultants, food truck operators, and realtors.

Ads Platform -
Best of both approaches 1 and 2

My team identified the proliferation of these two approaches and the need for a better solution.
As a result, we developed an Ads platform that can be embedded into any Meta surface, so teams don't have to reinvent the wheel and invest in creating their own ad tool from scratch.
This would effectively allow teams to build a native experience like approach 1, but leverage a consistent flow powered by LWI – the Meta standard ad tool (approach 2).

Shaping the platform

• I first mapped out the advertiser user journey, based on my experience working on LWI.
• I then evaluated what parts of the journey can the platform support building blocks for, namely, ad creation and ad management.

I broke down ad creation and management into individual steps and identified how we could break down our single monolithic application into discrete and logical "lego blocks".

This is where i identified the two primary tenets for our platform

Composability

The blocks had to work together as pluggable legos, and made available as a "menu of options".
A partner team may not to replicate the entire ad functionality within their app, and should be able to choose a selection of blocks to compose the right experience for their app.  

Adaptability

• Design: Each app has it's own design system– it's own visual language and design components. The blocks had to be adaptable to leverage the respective teams design system.
• Code: The blocks needed to be easily adapted in code, despite whatever backend or frontend technology the app was built on 

My contribution

My contribution in this phase barely involved drawing any new pixels.
‍
It was primarily strategic and systems thinking of how our existing ads tool can be abstracted into a usable platform.

The few pixels I did draw was painting the vision of how our blocks could be adapted by our partner teams to drive a consistent ads experience, while still feeling native within their own app.

The end-to-end process

Design sprint

I organized a design sprint bringing together 25+ stakeholders across 2 organizations/3 teams. This sprint turned out to be extremely crucial in building relationships and laying the groundwork for this 0-1 project.

Driving alignment

This sprint helped foster a shared understanding of our goals, unique advertiser problems on Messenger, each team's long-term product strategy, MVP feature prioritization, and early design explorations. This sprint helped us uncover differing assumptions held by each team that had been getting in the way of alignment. At the end of this process, we came together as a single team with a unified goal.

Prioritizing features

When my team collaborates with partner teams, we contribute our core ads expertise, while the partner team provides more specific knowledge in their vertical and surface areas. I worked with my team to document general ad use cases, while the Messenger team documented specific "messaging ad" use cases and North Star solutions.

We evaluated each feature by type of impact (adoption, completion, performance, retention); experience (needed, expected, nice to have); relevance to longterm strategy (key, not integral); and eng/design effort (low, medium, high). We then prioritized each item (MVP, fast follow, future) and discussed items with questions or differing opinions.

Design Principles

During the design sprint, we heard from people, platform, and product experts share insights on who we’re building for, what we’re building into, and why we’re building, we distilled this into guiding principles that will guide this work:

• Lightweight: In priority markets, connectivity is an issue. Therefore a simplified experience where advertisers can easily access, create, and manage ads is a key factor for SMBs.
• Personalized: Featuring recommendations will not only reduce cognitive load, but also help advertisers create better performing ads and build trust over time.
• Supportive: Our advertisers are usually solo entrepreneurs who are inexperienced advertisers, it will be crucial to provide guidance every step of the way without overwhelming them.

Ad creation

This is what the final version of the end to end ad creation journey looks like, as implemented in Messenger.  

As an example, I will walk you through one of the sub-journeys and how I went through multiple iterations to go from the lego block to the final version, which natively works within the Messenger design system and interaction model.

Sub journey: Set audience details

User needs

• Help me get started quickly, without needing to deeply understand how targeting works. Provide me smart defaults.
• More advanced advertisers need more control.
• All advertisers need a way to efficiently reuse (and edit, if needed) a previously created audience.

What does the Ads platform provide?

• Advantage audience: Meta recommended audience (uses an ML model to predict personalized targeting settings).
• Edit audience, create custom audience by choosing targeting locations and interests

Audience details sub- journey – As provided by platform Blocks

Audience details sub- journey – As translated within Messenger

The "audience details" sub-journey itself had 6 different states with multiple design decisions and iterations across each one of them. But lets look at the "Edit audience" screen where bulk of the interaction is happening.

Design iterations

Original - Meta platform block

Iteration 1 - Direct translation of the existing ads platform modules to Messenger

Feels busy, too many different UI controls, visual for audience estimation conflicts with the stepper navigation

Iteration 2 - Updated after aligning with the Messenger design system teams

Trade-offs 
Removed the “Address section” which showed a map view & double sliders due to lack of the components in the design system library.

Final design - A more native experience to the Messenger app

Improvements: Unified controls, less information density, easy to scan, added deletion feature

Business impact

In Q4 2023, we launched the Messenger ad creation experience. This resulted in an annualized revenue impact of approximately $20M.

By introducing the native ad experience in Messenger, ad completion rates increased from 1.3% to 4%. This is a substantial improvement over the previous experience, but it still falls short of the ~11% benchmark on the Facebook ad tool (which is to be expected, since the Facebook ad tool has been optimized over 10 years).

Next steps

In early 2024, I transitioned from this work to focus on scaling our ads platform on WhatsApp for iOS. However, the Messenger team continued to carry out funnel analysis and optimize the experience. Their planned work involved enhancing top-of-the-funnel adoption and introducing features to improve user comprehension and retention.

Cross-org collaboration

Over the past 1.5 years, I have been immersing myself in partner teams. If there's one skill I've had to develop and use extensively, it's the ability to collaborate with cross-functional partners spread across different organizations.

This requires quickly adapting to different platforms and their design systems, understanding unique user needs and challenges, and integrating the ad platform to meet these needs.

This often involves working with differing opinions, goals, and assumptions. Bringing everyone onto the same page is crucial for driving 0-1 projects as smoothly as possible.

Power of documentation

Documentation is a powerful tool that streamlines communication and fosters collaboration among team members. Building ad tools in Messenger was a significant project that required many alignment discussions. Here are a few examples of the most helpful documents:

• Design execution - I used this document to outline the project's scope, ownership areas, progress, user insights, and design review timelines. This helped my design partner and me stay on the same page and work towards a common goal.
• Variants, new components and patterns - I documented where the existing Messenger design components fell short, rationale for a variant or need for a new component. This document informed design decisions and guided conversations with design system teams.