Service Design program to develop an automated landing page builder using Adobe Experience Manager

 
 
hero_01.png
 

Internal creative teams will never meet all of their companies’ design needs. But, if we embrace enablement, we can UX ourselves out of a job so we can focus on bigger things that matter.

 
 

Why is this important?

Design democratization, streamlined workflows, and automation enabled marketers to create, build, and increase the volume of go-to-market campaigns they launched themselves.

Iterative executions & updates

100’s of team members needed to access design resources and people. By including them in research, testing, change management, we enabled smoother buy-in and adoption.

Team

Myself — Snr UX Designer
Lynnie Joseph — UX/UI Designer

Research — Phase 1

Audited hundreds of published landing pages, and interviewed our internal users

 

Design principles

  • One landing page per tactic

    Don’t try to overextend your campaign’s paid targeted tactics.

  • Ask people to do one thing

    When people do click, deliver what you promised — no more.

  • Message must align

    The message and visuals must align with what they just clicked. Keep copy short, obvious, avoid jargon and cliches.

  • Create forms you'd fill out

    Use the least fields possible that still ensure a qualifying lead.

  • Continuously test and improve

    Run experiments with content and layouts to optimize conversion. Submit is not compelling and performs poorly.

Strategy — Phase 1

Grouped the most common business needs into big themes

 

We asked our stakeholders “what do you want your audience to do?”

We envisioned a 3-step process

  • Identify your business goal

  • Create messaging to increase your chances of obtaining it

  • Select a themed landing page template that accommodated both of the above

 

Information Architecture

Created overarching content chunks that all templates could pull from.

 
LP_content summary.png

Workflow & processes

We sketched rough workflows that non-designers and project managers could deploy to get up and running.

 
WF map 1.png

Design — Phase 1

Templates by business goals

Our system needed to be flexible to accommodate different types of content to achieve the same goal.

  • Headline + messaging + form

  • Headline + messaging + download e-book form

  • Headline + messaging + read case studies + contact us form

 
 

A flexible and usable system

By simply swapping specific messaging, visuals, imagery, marketers from across product, industry, solution teams could all use the same template.

 

An Achilles heel…

We still needed to accommodate ‘occasional’ custom requests. So we had an off-menu option that had a similar design family feel, but we allowed customization of components. It triggered a traditional web process requiring all resources. This came back to haunt us…

 
 
design screens.png

Research — Phase 2

A year later, we took a hard look at the pain points and successes.

 

This was an opportune moment to address usability issues, explore fresh designs, and ultimately improve conversions.

 
01.As Is - journey flow.png

Mapped our internal user’s journey

After doing a complete review of the current ‘as is’ process some serious risk issues were identified. We not only gathered user feedback, but analytics data of highest-performing templates, and blended it all with project management reports.

 

A broken process

  • 59% of requests marked rush

  • 272 delivered late, only 31 on time

  • CTA’s not included at the start

Project needs

  • Kill root causes

  • Eliminate wrong template selection and endless debates over design

  • Reduce or eliminate custom requests

  • Clearly demonstrate high-converting templates, and why to choose them

Our vision

  • Build at scale and increase speed

  • Automate decisions when possible

  • Reduce copy and proofing cycles

  • Enable more self-authoring — not less

  • Reduce duration from 17 to 10 days

  • Explore removing UX tasks completely!

Key pain points

 
painpoint stakeholders.png

Stakeholders

  • Requests are submitted without a strategy in place

  • As everything is always a rush, marking it urgent by default

painpoints pm.png

PM’s and timelines

  • Requests marked rush without justification caused inaccurate prioritization

  • Wrong or inaccurate content causes multiple proofing and design cycles

  • Using unapproved assets (expensive Getty images) exposed a brand and legal risk

painpoint AEM.png

AEM site authors

  • As authoring is the last step, ALL the pressure falls on them

  • They bore the brunt of workaround tactics for non-approved changes

  • Authors wasted time reaching out for help, and doing endless rounds of edits 

 
painpoints copy.png

Copywriting & proofing

  • Content guidelines not utilized, ignored, or hacked together

  • Designer editing copy themselves to meet guidelines

  • Wrong content docs versions sent through proofing

painpoints uxui.png

UX & UI

  • The CTA drives template decision, it is needed upfront

  • Unclear strategy or content leads to changing templates midway through

  • No design direction caused redoing final UI comps “I don’t like it”

Design principles — Phase 2

  • Prevent premature requests

    Eliminate bad projects clogging the system by making strategies and call to actions required at submitting a request stage

  • Enable automation

    Utilize Workfront form features for information and requirements gathering. Create a matrix for recommended designs

  • Reduced choice

    Display optimal performing templates based on metrics not gut feelings and volume of content

  • Proof once

    Reduce the workload of writers and the proofing team. Utilize tools that enable constraints until dedicated writers can be sourced

  • Make yourself redundant

    UXers are available for consultation but create fool-proof designs that just work

  • Eliminate risk

    Restrict publishing ability by only proving paths to approved and optimized assets

Strategy — Phase 2

Reimagined a ‘To Be’ future state

Using all our previous inputs, we mapped out our recommendations counteracting the key failure points. We used these customer journey maps at the beginning of every meeting to remind decision-makers of the gravity of issues that needed to collectively resolve. Most of them are outside of our teams’ hands.

 
02.To be - journey flow.png

Key insights

Align Workfront process with campaign needs

  • Create steps and dropdowns that automate recommended designs

Use locked PDF docs

  • Editable PDF copy docs that are a visual 1:1 and character count match to wireframes

  • Allocated dedicated writers

Confirmation email

  • Recommended workflow

  • Links to our resource best practices

  • Expose high performing conversion rates

Fixes

Fix 1 - Updated business objectives to content and CTA options

Created a hard stop when collecting requirements, preventing incomplete requests from proceeding.

 

Fix 2 – Automated recommended designs

Mapped out each combination of dropdowns, to identify which matrix delivered our recommended template. Due to the simplest template being able to fulfill all the needs, we always exposed it by default.

 
design WF.png

Fix 3 – Optimized request steps in Workfront

Worked with the product owner to create dropdowns, required fields, and implemented progressive disclosure to increase chances of a successful submission.

 

Fix 4 – Enabled users to get started quicker

Reduced the number of steps, time for requests to be 'approved' and consolidated all information in only two locations.

 
wf 7.png

Fix 5 – Connected workflows across multiple systems

After receiving a confirmation link on-screen in Workfront, it opens a new tab and takes the user to AEM anchored to the correct template.

 

Design — Phase 2

 

Fix 6 – Refreshed design templates

A parallel design effort was happening simultaneously to update the design system and added forms to hero banners.

 
 
cta strategy 1.png

Fix 7 – Error prevention with locked copy decks

We discovered semi-editable PDFs that created a visual 1:1 match to wireframe templates and inherited the character count restrictions so users could see immediately if their copy exceeded specs.

 
editable pdfs.png

Fix 8 – All systems and tools in one place

The relationship between the template gallery and building in AEM meant users could reduce the number of tools they had to bounce between. Also, being an enterprise environment, there were security and account login steps to reduce.

 

Fix 9 – User-friendly template gallery and tools

A stunning gallery page was designed to house all of the categories of templates.

Renaming them reduced the bias from the previously ‘business objective’ named series, which caused users to think they always had to request a custom project if their project didn’t match.

Useful materials all in one place

  • See a live example

  • Download the copy content document

  • Get started in build mode

 
 
design gallery.png.jpg
 

Link to see live examples

image 1.png

Download copywriting tools

copy deck.png

Key takeaways

We released solutions iteratively, so by the time it came to present to the large volume of users, we had already started seeing performance improvements as per below.

We continue to roll out internal usability tests, created performance dashboards, start large-scale AB tests when applicable, communicate high-performing templates, and campaigns for others to learn from.

What we’re doing next…

  • Creating a localized image bank for personalization by global locations

  • Revisiting this program every 1-2 years to continue to our growth and innovation

  • Internal design school and learn resources for landing pages, so more resources can contribute to our speed and scaling

 
key takeaways 1.png
key takeaways 2.png
Previous
Previous

UX Strategy - Mapping content across 4000+ pages

Next
Next

Localization - Site speed optimization