website and product audits

my approach

To perform the audit, I used:

• MozBar to view each page’s metadata
• Google Analytics for page performance metrics
• Google Search Console for core web vitals
• Screaming Frog to uncover links and SERP previews
• SiteImprove for WCAG compliance
• Excel to catalogue the content 

the findings

The site had grown with no content standardization for many years, so unsurprisingly, the audit uncovered content issues that I categorized as a lack of clarity, consistency, and purpose.

clarity

Technical language and jargon abound, tasks seemed complex due to over-explanation, and most content was hidden in collapsible containers. 

This was the Check in page. It skipped the H2 level and had 13 headings at H3 indicating a lack of content hierarchy. The majority of the page content was in accordions, which a user may need to open several to get the info they need, increasing their cognitive load. 

consistency

The brand voice was missing in most content, products, and features were referred to using different labels, and custom elements were published on various pages, creating disparate experiences.

Compare these two table styles, the first from the design system and the second, custom coded to reflect the design of a print brochure.

purpose

Much of the content was published with no user need attached, pages were living that had no content — only links to other pages, and every program launched with an FAQ page (I have very strong feelings against FAQs in almost all contexts).

This page was published as a link farm to point to its child pages. There were many contentless pages like this.

accessibility

The aviation industry in Canada has a legal requirement to meet WCAG 2.0 AA, and the audit uncovered a number of areas were offside.

• Technical: Aria markup was missing or incomplete for many elements, links in components were coded to repeat the same CTA, and alt-text inputs were missing from the CMS.

• Design: the font size was fixed instead of scalable, accordion headings were all h3, and font styles used all caps.

• Content: alt-text was missing on most images, buttons, and links, and progressive headings weren’t used, making on-page navigation difficult.

the solution

I presented a suite of recommendations to address these issues, and these were the ones we implemented:

• I created a training program for site authors on web best practices, including SEO and WCAG.
• My team collaborated on a Content Style Guide to ensure consistency and clarity.
• I created the first content backlog and agile content creation practice to release incremental improvements to the content.
• My team collaborated with design and product to create page templates with minimum content recommendations.
• I worked with the product team to ensure platform enablement of accessibility standards and prevent future problems, like making alt-text mandatory and removing the h1 level option from components.
• I created and documented content lifecycle and governance models as formal content creation and review/revise/retire processes.

the outcomes

There’s more work to do to create a consistent, clear, and purposeful content ecosystem, but the outcomes of this audit and subsequent remediation work helped users of all abilities self-serve content on WestJet.com, and reduced calls to the call centre. The work I undertook to educate content contributors on web best practices will help prevent issues like these from overtaking the site again (I hope). 

user-generated content

the problem

With the goal of monetizing engagement content in the product ecosystem, the question of content quality was a big one. Clients aren't going to pay for access to poorly performing content generated by non-profits or other actors.

my hypothesis

Users generating content are unlikely to be expert writers, therefore they'll need stricter formatting guardrails for each content type.

my approach

I audited samples of user-generated content in several client instances (mid-market and enterprise).

• Google Analytics for conversion performance 
• Flesch scores for readability
• Alignment with web content best practices (chunking, progressive headings, hyperlinks with descriptive anchor text, strong CTAs)
• Content accuracy, clarity, and completeness 

findings

I discovered a number of content issues that could be avoided with structured inputs.

• excessively lengthy titles and descriptions
• full URL text pasted into the body copy
• CTAs to complete conversions off-platform
• text on images
• excessively long location descriptions
• unstructured, unprioritized information

the solution

My recommendations were to:

• replace the long-form "description" WYSIWYG with multiple purposeful sections with set headings, content hierarchy, and character counts
• Provide formatting tips (bulleted lists, etc.) as sample text within the interface
• Make CTA label and destination customizable 
• Improve the location tagging and map functionality
• provide image guidance (accessibility, cropping, subject matter) within the upload workflow.
• Make alt-text mandatory on buttons, text links, and images

the outcome

I was laid off shortly after compiling these recommendations, so I don't have clear outcomes for this work.

My assumption is that the additional structure of the content inputs and flexibility of the calls to action would better meet the needs of content contributors, enabling them to create and publish higher-quality content. With the improvements in quality, come improvements in performance, and thus, a content ecosystem clients would pay to have access to.

Rewards

the problem

The Spark! platform rewards program has a participation issue. User research indicated that, primarily among new users, program understanding was low, leading to reward dollars expiring before they could be redeemed, resulting in lost platform revenue.

Users didn't know critical information about their programs, including:

• they had rewards
• they could earn more rewards by volunteering and other pro-social actions
• rewards are basically cash
• they can donate their rewards to charities they care about
• rewards expire, so they have to donate them often

my hypothesis

With timely mobile push notifications and rewards information communicated in context within the platform, the percentage of users redeeming rewards will increase.

my approach

Journey map to understand the user's goals, feelings, actions, and issues as they work through the task of redeeming rewards.

Content ecosystem audit to identify relevant places to provide program information, and identify current content gaps.

Concept map for my own conceptual clarity of the program.

the solution

The project team worked with a content designer to craft new push messages at different intervals (new reward, 30 days til expiry, 1 day til expiry). 

I worked on the conceptual clarity of the ecosystem content. My recommendations were: 

Improve Consistency:

• use the same design patterns as other content types
• align terminology; depending on the user type, the same concept was referred to using up to four different terms 

Create Clarity:

• users had to search for entities that were labeled "match eligible" in order to find ones they could donate their rewards dollars to (despite the rewards dollars themselves not being match eligible). My recommendation was to disentangle these two concepts and introduce a "Donate rewards" label.
• To further clarify that rewards dollars are not match eligible, the "100% match eligible" content section needs to include a disclaimer that rewards are exempt from this policy.

Set Vocabulary:

• Basically, pick one term and use it everywhere. My recommendations were:
  • Rewards (program level) replacing Incentives, Dollars for Doers, and Donation currency, plus whatever legacy custom names were in prod.
  • Earn Rewards (action level).
  • Donate Rewards (conversion level) replacing "give". Based on consumer research comparing charitable donation conversions between give, donate, and spend. 

the outcome

This one was still in the backlog when I was laid off, but I expect the outcomes to be more sessions among new users who receive push notifications, and increased program participation (volunteering and redeeming rewards) as users understand they will be rewarded for doing so, plus greater program understanding for users overall. Platform revenue should see a lift with more users redeeming rewards, and operational costs (user support on rewards questions and reinstating expired rewards), will be reduced.