If you hadn't heard of AstraZeneca 12 months ago, you almost certainly have now. Following a partnership with Oxford University, the pharmaceutical company was one of the first in pursuit of a vaccine to tackle the ongoing COVID-19 pandemic. It hasn't always been an easy road, with suspensions and legal challenges in the mix, but no one can deny the incredible scientific and logistical achievement in delivery in a vaccine to market in under 12 months.
The process normally takes approximately 15 years, but AstraZeneca got approval for its vaccine in the UK on 30 December 2020 and the first jab was given on 4th January 2021, just months after COVID-19 became a reality. The company aims to distribute billions of doses around the world during the course of the pandemic and has agreed to do so on a not for profit basis.
However, whilst the story is well known at this point and AstraZeneca certainly isn't the only company to get a vaccine to market (thankfully) - it wasn't just a science challenge that the pharma giant faced in distributing doses around the world. According to David Simpson, Head of Global Commercial IT (Digital, Respiratory & CVRM) at AstraZeneca, speaking at Adobe Summit this week, there was also a digital challenge too.
In addition to the company needing to adapt to distributed work, along with the rest of the world, and making use of how digital channels to engage with healthcare professionals (HCPs) and patients, AstraZeneca also knew that it would need to use some sort of digital portal to distribute knowledge to local markets, to track jabs given, and to help those providing the vaccine track information. But the company wasn't sure what this would look like, given the fairly unique nature and scale of the challenge. Simpson explains:
The vaccine has now received approval in more than 50 countries and is already benefiting patients around the world. We've committed to make that vaccine at no profit for the duration of the pandemic and to make sure that we produced it and distributed it in such a way that was broad and equitable. That we made it as widely available as possible.
We have faced some interesting challenges. Certainly in the early part of the project, we weren't really sure how we wanted to use digital channels - but what we knew was we didn't want to leave ourselves waiting and the delivery of those digital projects became a delay in rolling out the vaccine. So we had to make a lot of ground early, without the certainty that we were heading towards a clear destination.
We knew we needed to use digital, we couldn't rely on offline channels exclusively. We knew that the solution would need to be flexible, that the needs of patients and HCPs would vary significantly, regulation wouldn't go away. And at the same time this commitment was over and above our normal business, and we needed to protect that.
Specific use cases
To meet these challenges AstraZeneca knew it needed a "really robust, really flexible content management system". It chose Adobe Experience Manager, and made use of other platforms such as MuleSoft and AWS, given that it had extensive experience with these technologies. Simpson says that the approach taken was to adopt a master template and then have localized models at a country level. Some 80% of requirements could be catered to by the master template, and then local pages could be adapted in the run up to - or even post - launch.
However, this solution differs from ones that AstraZeneca has deployed previously, as it is targeting very specific use cases and user journeys. Simpson explains:
One of the realisations that we reached quite quickly was that what we were trying to do was support a small number of very specific use cases. And those were quite different from the normal conversations that we would have with patients or HCPs. So for example, a typical conversation with a HCP might be to support and inform a prescribing decision, to choose a therapy for a condition. Well, this is a vaccine programme driven by government and informed by organizations like the W.H.O. So the prescribing journey is very different.
Another journey we might typically support is working with patients to adhere and get the most out of their therapy. Again, it's a completely different challenge. It's not really relevant in this example.
The journeys that we did need to support were to provide product safety information, or to allow healthcare professionals to ask information about the product, or to submit reports, submit new questions, or to look at expiry dates for the product. So very specific.
Simpson adds that the traffic profile of this site was going to be very different. Typically across a whole year, for the entire AstraZeneca portfolio, it would work with approximately one million patients. For its COVID-19 vaccine, the ambition is to produce around three billion doses. Even on a two dose regimen, that's at a scale the company hasn't managed before. Simpson says:
And the profile of the traffic was likely to be unusual as well, so lots of single visit users from basically anywhere in the world. Different journey, different volume.
We quickly realized that what we needed here was a standalone solution, rather than putting the content into existing channels. And that we would put that destination on a different infrastructure.
As noted above, central to all of this was catering to local markets and building that into the system. This required balancing the need for personalization at a local level (such as language requirements and legislation) with the need to ensure that AstraZeneca built in enough flexibility to direct doses as and when was needed. Simpson explains:
We needed to serve a different experience depending on the country that was being accessed from. So whether that's the language, the regulatory guidelines that change from jurisdiction to jurisdiction, etc. So each market would need a different experience. However, we also knew that we would need to use the packaging to direct traffic towards the site, so we imprint the URL and give a QR code on the medicine packaging.
What we didn't want to do was lock in batches to a certain market too early on in the process. If we put a UK-specific URL on a package, that could only then be distributed in the UK, and that wasn't helpful from a logistics point of view. So what we chose to do was go with a single Global Gateway, where we have one URL, that takes every user in the world, regardless of whether they are a patient or a HCP, regardless of which market they're in. It takes everybody to the same gateway and from there you can self drive to make sure you get the right experience.
In terms of lessons learned, Simpson says that partnering with vendors and using technologies that the company had previous experience with proved beneficial in the long run, even if they might not have been the best fit for the use case from the start. He says:
We made lots of decisions early on, some of them good, some of them not so good, but that's okay. As I look back at the start of the journey, I remember we seriously considered the idea of building this site on new technology, so selecting technologies that we weren't familiar with, but were ‘perfect' for the job at hand.
I'm really pleased that we decided not to. By working with Adobe, MuleSoft and AWS, we took a lot of risk out of the project, and we were also able to move really quickly at the start of the project. Make sure you've got some strengths and rely on them. When you've got to move quickly, understand where you are strong and try to get most use out of that.