Making Freshness More Awesome — Project 4 Retrospective

What does a supermarket sell? The customer in me says ‘groceries, fruits, stuff like that’. The marketing person I was says ‘a brand, a good sentiment, an association with a lifestyle’. Both are valid answers.

The great thing about learning UX design, though, is that it offers another answer, or series of answers. A supermarket sells an experience. But it also sells, or can sell, a service — something that tries to actually bridge the gaps in a user’s routine. Having been introduced to service design through lectures, I am lucky to be able to practice it for Cold Storage.


Studying the Brief

The brief we got was quite simple. Cold Storage, a venerable supermarket in Singapore, has long had an online shop, but it’s not been able to get much traction. Through the brief we could see a few of the client’s concerns:

  • They think it is a problem of the information architecture, because they asked us to look through the site map.
  • They think findability is another problem, because we were told to focus on the filtering and sorting system.

We fired off an email to the client to ascertain what he thinks about the brief, and got back some more valuable information about their target market — namely, newly married couples or singles with mid-high incomes. (Yuppies and DINKs, to use the jargon). With this information, we could then draw up a research plan.


The Current Website

After this, we needed to understand the current website through heuristic analysis. The current Cold Storage website looks like this:

Clearly, high priority is given to the deals and discounts, but there is also a lot of content that’s tangential to the online store.

  • ‘Inspirations’, for one, turns out to be a healthy eating blog. That’s not too relevant, but it’s also not clearly labelled at all.
  • ‘Recipes’ sounds interesting, and the content is actually quite high quality (including Youtube videos). But it seems to be there as an afterthought.
  • The division of promotions into several categories might make sense to the business, but it makes little sense for the user.

Comparing these points to competing online supermarkets (such as NTUC and RedMart) also brought up other issues:


Finding out about Users

Eager to see what lessons we can learn from the in-store experience, we started our research with contextual inquiry by going to the Takashimaya Cold Storage.

Some findings:

  • We saw a lot of attractive produce and imported goods, but a lot of it is not available for online ordering.
  • There’s a delivery system for in-store orders too, but it doesn’t link with the online system.

We then went into surveys (about 20 responses) and interviews with 10 users. Our initial attempt at doing guerrilla research went poorly; it turns out people carrying bags of stuff don’t want to stand around and chat.

That’s not a trivial observation, though. From both surveys and interviews, some relevant points quickly stood out that we found out about as we drew up affinity maps.

  • The primary reason people order online is to avoid carrying stuff.
  • As a result of this, people who don’t cook much — or don’t need to buy stuff for the whole household — don’t tend to order online.
  • Those who do cook often do both online and in-store shopping: online to buy ‘basic’ ingredients, in-store for more specific stuff.
  • Because of the above, users have specific things in mind for online shopping, and prefer to search rather than browse.
  • Users prefer shopping in-store partly for the experience, and Cold Storage has a great in-store experience. The air-con is cold, the aisles are widely spaced, and arrangements are neat and attractive.
  • Another advantage that Cold Storage has is its range of upmarket items and imported goods, such as cheese, ham and sauces.

Fitting Cold Storage to the Users

Based on that research, we built two personas, one of which is actually a couple.

We then drew out their customer journey maps (CJMs), which broadened their focus. Instead of simply focusing on the buying process, our research made it necessary to also focus on the planning process.

The combination of persona needs and the CJM led to some interesting issues:

  • The people who most want the convenience of ordering online (because they buy bulky things) are also most likely to have to supplement online ordering with in-store ordering (to get extra stuff for their meal plans)
  • The things that make people want to shop at Cold Storage (nice store, wide range of goods) are not available online
  • Some of Cold Storage’s interesting features, such as the recipes, are underutilised and underemphasised

Therefore, during our feature prioritisation, we settled on two main solutions that could solve these major issues.

  • We will improve the recipe feature, allowing users to shop for groceries based on recipes instead of shopping through available items one by one after figuring out a meal plan.
  • We will implement a joint account system, allowing multiple people to shop on the same list and coordinate how their grocery list.

Design Process

For the design process, we began with some sketched wireframes, and then went right into Axure to design a mid-fi prototype. As my teammates, especially Elky, had experience in design, they were more responsible for the look. My main task, on the other hand, was to design the interactions that would link the pages together.

Besides the usual problems with Axure coordination, the process was generally smooth; based on research and competitor analysis, we worked to make the interface less cluttered and cleaner, while placing the search bar front and centre. The search function can also search for both grocery items and recipes.


User Testing

We carried out two rounds of user testing, one using the mobile device and one using the desktop interface. Through this, we learned several interesting things.

During the first round, findability was an issue — our testers could find the recipe page, but did not know where to proceed from there.

As for the second round, the testers had other issues and suggestions about the recipe shopping feature which we took on board. Therefore we implemented other features in our design:

  • The button for grocery items changes on pressing, so users can keep track of what they’ve ordered
  • Better signposting of the checkout process so users can keep track of the process
  • Enabling portioning so people can adjust portion sizes for recipes and understand how much of each item they need
  • Changing the relative sizes of the pictures, to make them less distracting

Teamwork and Learning Points

I’m very lucky to be in this team, where the dynamic was generally quite equal and the skill set was evenly spread too. We took turns to take the lead on the things where we were more skilled or experienced, and the result was a smooth process where tasks were assigned by consensus.

Some learning points gleaned from the process included:

  • Think outside the box. Whether it’s ‘testing’ the staff during contextual inquiry, or monitoring a classmate’s RedMart order, or coming up with a more service-oriented solution, I felt we got a lot out of thinking up creative solutions.
  • Research must have goals in mind. Creativity is great, but one issue that we could have done better was in designing tasks for user testing. A clear idea of what needs to be tested would have made the process faster.
  • Keep everyone in the loop. I am thankful that the team generally made sure to inform each other of plans and changes. This made our process a lot smoother, and we didn’t have too many dislocations or clashes.
Show your support

Clapping shows how much you appreciated Ch Wil’s story.