Analysing an online retailer’s shopping cart

Analysing an online retailer’s shopping cart

After many years relying on a shopping cart developed in-house, a retailer was on the verge of signing with a new vendor. Some basic requirements analysis had taken place, as well as a series of questionnaires being answered. The board were due to approve the contract, until a non-Executive Director challenged the decision. How did they know the new product would fix existing problems in the shopping cart?

We were asked to review the as-is checkout flow and provide an unbiased assessment of its effectiveness. The board decision was due in a week, with a report required before then. Our sponsor provided us with some basic research they’d carried out, and confirmation no design documentation was available. 

Step 1: The Teardown

Acting as a consumer we made several purchases. As we stepped through the flow it was documented carefully, capturing specific points to return to later. Each purchase added to our map of the flow, which we illustrated with screen shots and commentary. Using the screen shots, we put together basic wireframes that presented the major stages in the flow and the fields the customer completed.

An example page from the teardown, high-lighting issues.

Step 2: The Happy Path

From our wireframes and notes we put together the “happy path”. This was the ideal route the user should follow to complete the shopping cart. 

The “happy path” for the flow, showing how the customer should progress

Step 3: The Sad Paths

Where the user encounters errors or problems the flow breaks into “sad paths”. Some of these will come to an end, forcing the user to abandon the cart. Others loop back around and let the user continue.

In our flows we use a clear notation to distinguish happy and sad paths. Happy paths always follow a solid green line, while sad paths use red dotted ones. This makes the different routes easier to follow.

The main “sad paths”, showing where users might encounter problems

Step 4: Analytics

With the flows mapped, we applied analytics and sales data to create a drop-off funnel. This showed us where customers were leaving the flow, including whether they were being drawn down sad paths. The data exposed several points of interest we could test; including why the abandonment rate on the payment page was exceptionally high, the lack of international sales and so few account holders logged in.

The full user flow, showing where issues are being encountered

Step 5: Testing

We tested each abandonment point through a combination of working back through screenshots, reviewing notes and on-site testing. Our more detailed reviews exposed code weaknesses that were breaking the secure padlocks on the payments page, additional shipping costs appearing in the payments page without being explained, and missing signposting to account login. Other less serious issues were also found, all of which were documented with an assessment about their impact on sales.

Step 6: Report

The final report had to satisfy two audiences. For our sponsor we provided detailed analysis on the existing flow they could check against features in the proposed replacement. Our analysis was used to raise final queries with the vendor, which provided reassurance mistakes were not going to be repeated.

For the board we provided an independent check there were problems with the existing checkout that were causing sales to be lost.

Final (redacted) one-page summary prepared for the board

Lessons for clients

In an ideal world this work would have been done at the start of the selection process. More detailed analysis and design may have produced more robust requirements to take to vendors and a wider range of options. Having missed out on this, the client took the right action commissioning an independent, high level review before committing to a contract.

When time is tight, our view is a simplified analysis of what already exists is perfectly acceptable. Extensive user research and testing isn’t always necessary when what’s required is a picture of what already happens. In our experience, smaller projects can quickly become confused when excessive analysis and research takes place.

For our client what was needed was a high-level review by experts who understood how to find problems in a UX flow. It may be all you need too.

Latest

Being a “green business” is no longer a differentiator

“Sustainable” is no longer a source of differentiation. Promoting your “green credentials” in content is important, but so too is avoiding “greenwashing”

Related

How we created an investment pitch for a small business

We helped a company secure new investment by creating a compelling narrative.