Vendor Engagement for Tobacco Company
End-to-End Built • +3.4% Conversion and +54% ATP
Vendor Engagement for Tobacco Company
End-to-End Built • +3.4% Conversion and +54% ATP
Index
Index
A web platform to interact with vendors in any OpCo
A web platform to interact with vendors in any OpCo
Digital Trade Engagement platform (DTE) is Philip Morris International (PMI) internal platform, built on Salesforce, used to engage with local vendors by helping them manage their inventory, request products, earn rewards through points and learn about the PMI brand, its products and programs.
Digital Trade Engagement platform (DTE) is Philip Morris International (PMI) internal platform, built on Salesforce, used to engage with local vendors by helping them manage their inventory, request products, earn rewards through points and learn about the PMI brand, its products and programs.
Engagement issues
Engagement issues
After reviewing the analytics for the platform, I found that most users spent their time either in the Home Page, which was the page they landed in, or at the Tasks Page, which was where they could work towards rewards and stacking their inventory. Aside from these, they also tended to visit pages and links that were prompted on the first page.
After reviewing the analytics for the platform, I found that most users spent their time either in the Home Page, which was the page they landed in, or at the Tasks Page, which was where they could work towards rewards and stacking their inventory. Aside from these, they also tended to visit pages and links that were prompted on the first page.
The IQOS page, which was PMI's largest product push at the time, was the least visited page, with one of the lowest times spend on page, and highest bounce rate.
The IQOS page, which was PMI's largest product push at the time, was the least visited page, with one of the lowest times spend on page, and highest bounce rate.
Google Analytics data from DTE Alabania
Google Analytics data from DTE Alabania
Google Analytics data from DTE Alabania
The most typical pattern showed vendors opening the app, quickly glanced at tasks, returned to the home page, then reopen tasks again. Other patterns included visiting other pages such as Loyalty, Planogram and Rewards.
The most typical pattern showed vendors opening the app, quickly glanced at tasks, returned to the home page, then reopen tasks again. Other patterns included visiting other pages such as Loyalty, Planogram and Rewards.
DTE Albania's funnel report
DTE Albania's funnel report
DTE Albania's funnel report
Using Salesforce Experience Builder to manage content
Using Salesforce Experience Builder to manage content
The web app is built on top of Salesforce through the Experience Builder which contains a CMS that the back-office and DTE team regularly maintain & update. Although it made managing the content and syncing with the rest of the organization’s infrastructure easier, it provided a rigid frontend interface with little option to customize.
The web app is built on top of Salesforce through the Experience Builder which contains a CMS that the back-office and DTE team regularly maintain & update. Although it made managing the content and syncing with the rest of the organization’s infrastructure easier, it provided a rigid frontend interface with little option to customize.
DTE's initial information architecture map
DTE's initial information architecture map
The interface was full of inconsistencies. The typographic hierarchy was missing, off-brand colors were used throughout the app and most of the content design was managed through banners which looked differently depending on who had worked on them. The UI elements displayed bugs throughout the application.
The interface was full of inconsistencies. The typographic hierarchy was missing, off-brand colors were used throughout the app and most of the content design was managed through banners which looked differently depending on who had worked on them. The UI elements displayed bugs throughout the application.
DEMO
DEMO
DEMO
DTE's initial interface & journey
DTE's initial interface & journey
Project goals vs. organisational KPIs
Project goals vs. organisational KPIs
The Product Owner's goal was to increase the rate of completion of My Tasks and vendor engagement with IQOS & other programs. In addition, they wanted to polish The Brand Corner so users could learn more easily about products and news related to them.
The Product Owner's goal was to increase the rate of completion of My Tasks and vendor engagement with IQOS & other programs. In addition, they wanted to polish The Brand Corner so users could learn more easily about products and news related to them.
Larger goals included making the onboarding easier which was happening with the help of Business Executives, encourage every POS to log-in at least once a month, and reducing the back-office support tickets by teaching retailers to solve their issues within the app.
Larger goals included making the onboarding easier which was happening with the help of Business Executives, encourage every POS to log-in at least once a month, and reducing the back-office support tickets by teaching retailers to solve their issues within the app.
DTE 2020 KPIs
DTE 2020 KPIs
IQOS was a major shift in organizational goals, so the Product Owner wanted to reflect that in the application. Being the lowest ranking viewed section, and almost completely ignored, the team needed a way to mediate that.
IQOS was a major shift in organizational goals, so the Product Owner wanted to reflect that in the application. Being the lowest ranking viewed section, and almost completely ignored, the team needed a way to mediate that.
IQOS Program Goal vs. current KPIs
IQOS Program Goal vs. current KPIs
A day in the life of vendors, the executives and the back office
A day in the life of vendors, the executives and the back office
I conducted 13 qualitative interviews with vendors which involved questions related to their day-to-day work, shop experience, and interactions with the app. My aim was to understand how retailers interacted with the company and vice versa and what were the mutual expectations. After deliberations, I would ask them to solve their daily tasks and fetch a reward if they had enough points. Lastly, I would have them pitch IQOS to me.
I conducted 13 qualitative interviews with vendors which involved questions related to their day-to-day work, shop experience, and interactions with the app. My aim was to understand how retailers interacted with the company and vice versa and what were the mutual expectations. After deliberations, I would ask them to solve their daily tasks and fetch a reward if they had enough points. Lastly, I would have them pitch IQOS to me.
Additionally I interviewed 4 business brand executives, 1 back office representative, and 1 consumer engagement executive. The executives were quite familiar with the app because they signed-up, onboarded & supported retailers on a daily basis. The back office rep mostly handled the Salesforce CMS and related issues.
Vendor interview key takeaways
Vendor interview key takeaways
The main frustrations were related to the business tasks, which were the most important ones because it impacted vendors' bottom line directly. They would loose points when they couldn’t find the right answer and as a result loose on potential revenue.
The main frustrations were related to the business tasks, which were the most important ones because it impacted vendors' bottom line directly. They would loose points when they couldn’t find the right answer and as a result loose on potential revenue.
I loose points because I can't find answers to the business tasks.
— 47% of interviewees
Why can't I refer customers to IQOS through the app?! They need to complete a form…
— 12% of interviewees
I want more relatable content and coupons. Most of the time I'm not interested.
— 28% of interviewees
IQOS on the other hand, was seen as a hassle because they needed to contact a representative to come in the store and explain to the customer how it worked. Customers would often loose interest and vendors wouldn’t even try to sell.
IQOS on the other hand, was seen as a hassle because they needed to contact a representative to come in the store and explain to the customer how it worked. Customers would often loose interest and vendors wouldn’t even try to sell.
I loose points because I can't find answers to the business tasks.
— 47% of interviewees
Why can't I refer customers to IQOS through the app?! They need to complete a form…
— 12% of interviewees
I want more relatable content and coupons. Most of the time I'm not interested.
— 28% of interviewees
I loose points because I can't find answers to the business tasks.
— 47% of interviewees
Why can't I refer customers to IQOS through the app?! They need to complete a form…
— 12% of interviewees
I want more relatable content and coupons. Most of the time I'm not interested.
— 28% of interviewees
Inference and assumptions from the available data
Inference and assumptions from the available data
Retailers bounce back and forth between the Home Page & My Tasks and I had two working hypothesis for this. The first one related to their day-to-day. They started working on a task, but then customers would walk in and they had to leave that aside to cater to the them.
Retailers bounce back and forth between the Home Page & My Tasks and I had two working hypothesis for this. The first one related to their day-to-day. They started working on a task, but then customers would walk in and they had to leave that aside to cater to the them.
The second hypothesis rested with the way tasks were supposed to work. Business task sometimes involved fetching information from around the app and in different sections, so they would have to leave the page, but then they’d get frustrated, distracted or busy, so they’d leave that aside or answer haphazardly.
The second hypothesis rested with the way tasks were supposed to work. Business task sometimes involved fetching information from around the app and in different sections, so they would have to leave the page, but then they’d get frustrated, distracted or busy, so they’d leave that aside or answer haphazardly.
With IQOS and the referral process, the main issue was the out-of-the-app form and lack of education, which often required the assistance of a coach to be present in person for less tech-savvy vendors.
With IQOS and the referral process, the main issue was the out-of-the-app form and lack of education, which often required the assistance of a coach to be present in person for less tech-savvy vendors.
Focusing on what matters
Focusing on what matters
Armed with the necessary data and context, I started working on the new experience keeping in mind that the Salesforce constraint while providing the necessary flexibility to reach our goals and have a shot at reaching the KPIs.
Armed with the necessary data and context, I started working on the new experience keeping in mind that the Salesforce constraint while providing the necessary flexibility to reach our goals and have a shot at reaching the KPIs.
I reworked the architecture to focus on 4 key pillars; the Home Page, IQOS, Tasks and the Brand. The Home Page would function as a hub for all the relevant content Philip Morris wanted to steer vendors towards.
I reworked the architecture to focus on 4 key pillars; the Home Page, IQOS, Tasks and the Brand. The Home Page would function as a hub for all the relevant content Philip Morris wanted to steer vendors towards.
DTE's new information architecture map
DTE's new information architecture map
DTE's new information architecture map
Wireframes for the interface and new design components
DTE's new information architecture map
To think outside the box you need to know where the box is
To think outside the box you need to know where the box is
With the deadline looming, I was tasked to find and hire a developer to figure out how to actually build the new experience. All of DTE was built on top of Salesforce so that made development challenging because we were told from the beginning that we wouldn’t have access to the back-end. Any change we would make had to be done on the front-end with the available data points.
We proposed the new version of DTE to be implemented as a Single-Page Application (SPA) and provided tools & resources to do so. An SPA would remove all the delays and improve the loading speed which was a major drawback, with users taking more than 15 seconds to reach the Home Page. It also meant that the codebase would be contained within a single project, which would make the work of future programmers much easier.
Vue components & example of how GLUE makes use of the Javascript API provided by Salesforce
Vue components & example of how GLUE makes use of the Javascript API provided by Salesforce
We used Vue 3 to build the whole SPA and GLUE to fetch dynamically all the (~25 functioning) components that contained data that pertained to our requirements for displaying on the page. The Salesforce Javascript API was used to retrieve that data and feed it to the SPA components.
We used Vue 3 to build the whole SPA and GLUE to fetch dynamically all the (~25 functioning) components that contained data that pertained to our requirements for displaying on the page. The Salesforce Javascript API was used to retrieve that data and feed it to the SPA components.
Tags in Salesforce CMS in the form of simple HTML elements and/or unique word patterns which will be fetched by the SPA and used to add extra functionalities that are not supported without backend
Tags in Salesforce CMS in the form of simple HTML elements and/or unique word patterns which will be fetched by the SPA and used to add extra functionalities that are not supported without backend
The decision was made because Vue comes with great documentation with as well as the extremely easy learning curve it has. As an added bonus it allows us to organize the code in reusable components, which makes the whole project more maintainable and easy to comprehend. The whole process translated into a faster experience as well as allow easier code management.
The decision was made because Vue comes with great documentation with as well as the extremely easy learning curve it has. As an added bonus it allows us to organize the code in reusable components, which makes the whole project more maintainable and easy to comprehend. The whole process translated into a faster experience as well as allow easier code management.
Declutter and restructure to increase engagement
Declutter and restructure to increase engagement
The home page would remain the main interaction touchpoint becoming a hub for all the relevant content we wanted to funnel vendors towards. I added a weather, currency converter & horoscope widget, fetched through APIs and designed to fit the app’s visual language as per the team's request.
Tasks and IQOS Page were some of the most reworked pages as these were also the focus of our shared KPIs. We decided to implement a bottom navigation instead of the side menu to reduce the steps needed to reach relevant content. Since most users installed DTE as an app because it was built as Progressive Web App (PWA), my intuition was that these changes would increase overall engagement and discoverability of the app’s content.
Tasks and IQOS Page were some of the most reworked pages as these were also the focus of our shared KPIs. We decided to implement a bottom navigation instead of the side menu to reduce the steps needed to reach relevant content. Since most users installed DTE as an app because it was built as Progressive Web App (PWA), my intuition was that these changes would increase overall engagement and discoverability of the app’s content.
Home Page, Tasks and IQOS Page - key screens and flows of DTE app
Home Page, Tasks and IQOS Page - key screens and flows of DTE app
Home Page, Tasks and IQOS Page - key screens and flows of DTE app
When you don't involve all key stakeholders things fall apart
When you don't involve all key stakeholders things fall apart
Halfway through the implementation and with one month left to our deadline, we were informed that due to contractual implications between Philip Morris and Salesforce we couldn’t use the SPA approach. This came as a surprise to all the parties involved.
Halfway through the implementation and with one month left to our deadline, we were informed that due to contractual implications between Philip Morris and Salesforce we couldn’t use the SPA approach. This came as a surprise to all the parties involved.
The SPA was a workaround to the lack of backend access which allowed us to provide a better experience and more flexibility in development. Any change that implicated the backend would have to undergo a lengthy and costly process.
The SPA was a workaround to the lack of backend access which allowed us to provide a better experience and more flexibility in development. Any change that implicated the backend would have to undergo a lengthy and costly process.
⚠️
⚠️
If you don't involve legal, legal will involve you!
If you don't involve legal, legal will involve you!
At this point, we agreed with all the parties involved that we had to forego a bunch of features poised to enhance the experience, and work only with the components provided to us through the Salesforce platform which luckily, were all modifiable using plain HTML & CSS.
At this point, we agreed with all the parties involved that we had to forego a bunch of features poised to enhance the experience, and work only with the components provided to us through the Salesforce platform which luckily, were all modifiable using plain HTML & CSS.
Reframe and reiterate
Reframe and reiterate
Given the new restrictions we quickly re-evaluated what was feasible to implement until our deadline. Implementation was tedious and manual, working only through HTML & CSS.
Given the new restrictions we quickly re-evaluated what was feasible to implement until our deadline. Implementation was tedious and manual, working only through HTML & CSS.
Most of the changes we suggested had to be scrapped, the architecture was to be reversed and our priority shifted to delivering on time, with minimal bugs, and provide a clear handover to the Gogel team, which would implement the changes we originally wanted to make further down the road.
Most of the changes we suggested had to be scrapped, the architecture was to be reversed and our priority shifted to delivering on time, with minimal bugs, and provide a clear handover to the Gogel team, which would implement the changes we originally wanted to make further down the road.
Final screens for the DTE solution reworked only with HTML & CSS
Final screens for the DTE solution reworked only with HTML & CSS
A successful launch and lots of lessons learned
A successful launch and lots of lessons learned
Although it was a really close call, we managed launch on time to the entire customer base without disrupting the service. There were still a few kinks to iron out regarding responsiveness for a small portion of devices & browsers, but the Back Office didn’t report any complaints from the vendors in the next month post-launch.
Although it was a really close call, we managed launch on time to the entire customer base without disrupting the service. There were still a few kinks to iron out regarding responsiveness for a small portion of devices & browsers, but the Back Office didn’t report any complaints from the vendors in the next month post-launch.
In the following quarter, the DTE team reported a 3.4% increase in registered POS, a 0.63% increase in successfully completed tasks, and a 0.71% increase in active POS, although this could have been also due to the new program Gogel launched called Shoplife. IQOS unique page views increased to 874 and average time spent on page went from 39 sec. to almost a minute.
In the following quarter, the DTE team reported a 3.4% increase in registered POS, a 0.63% increase in successfully completed tasks, and a 0.71% increase in active POS, although this could have been also due to the new program Gogel launched called Shoplife. IQOS unique page views increased to 874 and average time spent on page went from 39 sec. to almost a minute.
Overall, I did not make the impact I was hoping to make, however I was happy to hear that our initial roadmap started being implemented 6 months later!
Overall, I did not make the impact I was hoping to make, however I was happy to hear that our initial roadmap started being implemented 6 months later!