Intro

GoodRx had been saving millions of American patients millions of dollars on their prescriptions for over a decade, so its not very surprising that word of mouth was GoodRx’s biggest source of new users. Surprisingly though, the users that were responsible for the most referrals were healthcare providers. Healthcare Providers, or HCP’s, had been using GoodRx just as long as consumers had, but for a slightly different reason: to facilitate the “cost” discussion with their patients. Finding the actual out-of-pocket cost of a medication could be extremely time consuming but remained incredibly relevant to a provider trying to prescribe an accessible medication to their patient. They were using GoodRx to provide themselves and their patients with a clear set of pricing data, despite the fact that it had never been intended to be used in this way

What is GoodRx?

GoodRx is a product that creates drug discount coupons that can be redeemed at pharmacies when picking up prescription medications. It has about 6.4 million monthly users.

GoodRx Provider was conceived to better solve the needs of HCP’s everywhere and give them access to better pricing tools. It was launched in an MVP state to healthcare providers and the team watched as the tool gained traction, despite only having a few provider focused features. This gave us the validation that was needed to build out a fully realized product for providers. When the decision was made to build it out, I was tapped to join the provider squad and lead the redesign effort of the drug pricing tools.

The expanded price comparison tools were clearly very useful to providers. In addition to the classic GoodRx coupon pricing across different pharmacy chains, the pricing tools included an insurance drug pricing tool that we called Real Time Benefits Checker. This combination of pricing options were exactly what providers found useful in conversations with their patients in the moments preceding writing a prescription.

The majority of providers reported that the cost of a medication is a key consideration in deciding what drug to prescribe. If they feel that a medication is priced in a way that their patient will have trouble accessing it, they will prescribe something else that may not be as effective, but more accessible. This isn’t a call the provider makes on their own, it is generally through a conversation with their patient that they determine what drug will lead to the best outcome. Our provider product had the functionality to inform that conversation, but the current design was not facilitating it. Providers were reporting that the information was useful, but not accessible at the speed of conversation. To further complicate matters, Provider Mode was becoming a victim of its own success. Drug manufacturers were excited to partner with GoodRx to provide high fidelity pricing information for their drugs and financial assistance programs, introducing even more relevant pricing information tools. The existing page layout was beginning to crumble under the weight all of this information, and the task of comparing pricing options easily was becoming increasingly difficult.

I led a brainstorm with the squad to determine our next step forward. Through discussions during this workshop, we collectively had a bit of product epiphany: The GoodRx consumer facing product was being served as content, but healthcare providers needed a tool.

What is the difference between content and a tool?

Content is meant to:

• Captivate

• Engage

• Be consumed

This reframing helped give the team a more actionable challenge: How might we shape our product into a better tool for providers? Applying the research insights from our UXR team, we came up with these answers.

• Support adjustment of the input and display how it affects the output in one experience (keep users on same page)

• Facilitate the overview of the drug family landscape (brand and generic, other drugs in family)

• Support the handoff from provider to patient (specific price solutions as well as general pricing report)

• Promote the awareness of the presence of all pricing options

From a business perspective, this would allow us to:

• display more of our partners’ sponsored content

• Convert more patients to GoodRx users

• Create a positive feedback loop of more providers using goodrx with more patients

To help evaluate our current product’s performance, I created a basic IA map to help visualize the flows that a doctor would have to engage to complete their key tasks.

I find that creating a good IA map is a great way to help the larger team collaborate in the design process before any interface pixels have hit the page.

Right away, the entire team could see that providers were having to bounce from page to page, initiating complicated flows that looped back on themselves just to get the patient a drug discount. These flows made sense for a content format that was trying to keep a user captivated and engaged with the product, but not so much for a tool that was trying to solve the challenges we had identified above. Using IA maps, I led the team in an exercise to create a new high level architecture that would be a better fit for providers. These improved flows would prioritize minimizing the amount of new pages providers would have to load in order to complete key tasks like sharing a coupon with patients. If providers could accomplish their key tasks on a single page, not only would they be able to accomplish them faster, but they’d also be able to tweak inputs without starting the task over entirely.

With a new framework for our product, the team and leadership was bought in on our vision to redesign provider mode around the support of key task completion for providers.

Early testing

Upon moving into the testing phase, as always, we were due some humbling. Providers responded very well to the content -> tool move we had made, but there was still a lot of work to do.

Performed well

• Fundamental changes to the page layout allowed us to display more pricing options on the screen at once. The test showed that providers were able to complete tasks quicker because of it.

• With the use of the new contextual action panel on the right, we were able to keep providers on a single page, improving their workflow.

Issues found

• Awareness of brand/generic alternative drugs was still relatively low.

• Providers liked the mixture of high level prices by savings category as well as access to the full list of prices, we we hadn’t hit the mark on presentation yet

The first big breakthrough came when doctors communicated to us that they were generally unaware of the existence of generic alternatives and brand equivalents for all but their most prescribed drugs. These additional options were frequently sources of dramatic pricing differences and doctors found the ability to quickly compare prices incredibly useful. To address this need, I pulled the brand equivalent options out of the dropdown they had been hiding in and displayed the full list alongside the drug configurator.

In this early batch of tests, we also learned that our pricing summary was leaving some providers confused. The relationship between our editorialized rollup of pricing options and the full list of pricing options was unclear. The concept was well received, but the execution was not. I eventually settled on an execution that toggled between our editorialized list and the full list for the sake of clarity but at the expense of our original vision to display all options on a single page.

Later testing

Performed well

• Placing the drug configurator on the left hand panel was a better fit from an interaction flow perspective and it allowed us to expand the brand/generic drug alternatives beyond just a drop down. Speaking of which…

• The expanded brand/generic alternatives was successfully providing awareness of the existence of such options.

Issues found

• Lingering privacy-driven anxiety surrounding sharing coupons with patients.

• Lack of successful handoff feedback for providers.

• More room for improvement on the pricing options list

Providers had been handing off coupons to patients through text and email (since that’s how the core product functioned) but they were somewhat reluctant to do so. They were never sure whether personal information, including personal phone number and email, would be shared with the patient. Also, they could rarely be sure that the patient actually received the message. This exchange was always happening at the speed of conversation, and its very awkward to stop mid conversation to ask a patient to access their device and confirm the receipt of the coupon. We had always known that these were issues inherent with our sharing system, but we’d never really been able to find a better alternative until we tested this version of the prototype. Then we made an interesting discovery.

Generally speaking, the core user base of GoodRx were a bit older and a bit less tech-savvy. Newer technological conventions were things that we avoided hinging important functionality on. That’s why I was pretty surprised to hear from a provider that we should be leveraging QR codes for handoff. They explained that a lot of patient-facing medical services started using QR codes extensively during the pandemic to limit person to person or person to object contact. As I investigated further, this turned out to be absolutely on the money. Hospitals, clinics and pharmacies all had begun heavily leveraging QR in their interactions, essentially educating our customer base on how to use them. This insight didn’t only give us a new and more effective way to complete our handoff. It was also relevant to our core product. A separate team tested and verified the effectiveness of QR code usability with our core audience and began to use QR codes for keys flows in the core product. Our small team was now generating insights that challenged general assumptions about GoodRx’s millions of users and delivering actionable results.

Speaking of using off the beaten path methods of handoff, we discovered that printing was more viable that we had assumed. Printing functionality had never been something the core GoodRx product really cared about. Most users were using the product on their mobile phones, and who has a printer setup to receive print jobs from their phones? That’s why it was a big surprise to get a lot of requests for print formatting from doctors. Doctors were primarily using Provider mode on an office desktop computer, and that computer was almost always hooked up to a printer. It turns out, doctors’ preferred method of sharing resources with patients was to physically hand over hard copies of these printed resources, so the ability to print was always supported. To address this, I introduced the ability to easily print a drug coupon as the primary method of sharing. I also created functionality to support the generation of a full pricing report on any drug, formatted for printing.

These later rounds of testing also revealed that we still hadn’t hit the mark on our pricing options list execution. Previously we opted for a toggle between an editorialized list of best pricing options and all pricing options. It turned out providers just weren’t finding the editorialization useful. Providers reported that patients rarely qualified for more than two categories of pricing option. They wanted to be able to quickly navigate to the relevant pricing option category and ignore the rest. This action wasn’t really supported by either screen on our toggle, so I took it back to the drawing board.

Shipping

After shipping the provider mode redesign, the Provider Mode opt-in rate quickly ticked up to 90% and stayed there. Providers were seeing and believing the value of Provider Mode almost immediately. According to our research, nearly every medical provider using GoodRx had switched over to Provider Mode permanently. That meant that over 800k medical professionals were using Provider Mode.

Final changes

• An even further expansion of the Brand/generic equivalent options on the left panel. Providers were now able to easily see and switch between all equivalent alternatives

• Finally nailed the execution of the pricing options list by displaying the full list, but with each category in an expandable container that displayed the first 3 best options.

• Included the ability to generate and share a pricing report for any drug, to hand off to the patient.

Our advertising partners were also seeing the value of the product as well. Partnership revenue increased by 170% in the first year and was projected to grow by 15% in the next.