Bundle of joy: Optimizing the purchase experience for add-ons and upsells

Amelia Northrup-Simpson, Tessitura Network, USA, Amanda Abernathy, John G. Shedd Aquarium, USA, Amie Wong, California Academy of Sciences, USA


Upselling, cross-selling, add-ons, bundling... Whatever you call it, it means more engagement for your patrons- and more money for your institution. But how do you design a purchase path that offers the right mix of experiences to visitors? In this paper, two organizations which have recently refreshed their purchase path in order to optimize user experience and offer relevant experiences that deepen engagement delve into the challenges of uniting all experiences in one place: segmentation and dynamic content, user flow, tracking and analytics, and the strategy behind the ideal purchase experience.  Learning objectives: - compare methods of segmenting users on a purchase path - discuss how to track user flow through the purchase experience - describe an ideal purchase experience from the point of view of the visitor and the institution

Keywords: purchase path, user experience, visitor engagement, analytics, data-driven decisions

Many organizations look to upselling, cross-selling, add-ons, and bundling to help build visitor engagement and raise more money for your institution. But how do you design a purchase path that offers the right mix of experiences to visitors? The Shedd Aquarium and the California Academy of Sciences have recently refreshed their purchase paths in order to optimize user experience and offer relevant experiences to deepen engagement.

This paper explores the decision-making behind each organization’s ideal purchase experience; discusses tactics used, including segmentation and dynamic content; and shares about the challenges that come with uniting all user experiences in one online pathway.

About the institutions in this paper

Shedd Aquarium is a public aquarium located in Chicago, Illinois that opened its doors in 1930. Home to over 1,500 species, Shedd sees about 2 million visitors annually. Shedd began investigating the redesign of its website in mid-2017, and launched on August 27, 2019. The Shedd Aquarium aimed to streamline the digital purchasing experience, including a mobile-responsive design that reflected their recently simplified pricing and policies.

The California Academy of Sciences is a renowned scientific and educational institution dedicated to exploring, explaining, and sustaining life on Earth. Established in 1853 and based in San Francisco’s Golden Gate Park, it is home to a world-class aquarium, planetarium, and natural history museum, as well as innovative programs in scientific research and education—all under one living roof. It welcomes over one million visitors each year and houses over 38,000 live animals, nearly 46 million scientific specimens, and over 380 full-time staff members.

The California Academy of Sciences undertook their website redesign as part of the much larger project of transitioning their core systems, while also introducing date-specific, dynamically priced tickets. With the ecommerce portion of the project, they aimed to simplify the checkout process, optimize the experience for mobile devices, and provide a seamless, intuitive user flow throughout the full website experience.

Online purchase pathways, add-ons, and upsells

Purchase pathways offer a myriad of options. There are three broad approaches to consider: using a prebuilt platform, building a fully custom transaction path, and a hybrid of the two. A prebuilt platform can offer numerous efficiencies, stability, and ease of maintenance. However, platforms vary widely, and different technology platforms can open up both opportunities and roadblocks. A white-label platform may not be flexible enough to meet all of an organization’s specific business needs.

Building your own ecommerce site via an application programming interface (API) offers far greater flexibility, but is a more resource-intensive process and requires more involvement in long-term maintenance. Both Shedd Aquarium and California Academy of Sciences opted for a hybrid approach, selecting Tessitura Network Express Web (TNEW) to underpin the online transaction path, while building custom elements to meet specific high-priority business needs.

One thing that both organizations wanted to improve in their new websites was the ability to promote upsells and add-ons. An upsell is an offer to upgrade a customer’s selected product or order, replacing it with a different product. An add-on is an offer to add an additional product to a customer’s order. Both of these can be highly contextual, driven by the products already in a visitor’s cart: for example, a person buying general admission might be offered an upsell for a membership, or an add-on for a special experience. A cart with a combination of adult and child admission might be offered an upsell for a family membership or an add-on for an age-appropriate experience.

Shedd Aquarium: purchase path redesign

By Amanda Abernathy and Aaron Senser

Overview & goals

In September 2017, Shedd Aquarium began the process of redesigning our former website, which had previously launched in 2014. Shedd had recently updated our mission and vision in May of 2017 to focus on sparking compassion, curiosity and conservation for the aquatic animal world and wanted a site that highlighted the Shedd experience of bringing people closer to animals and their ecosystems. Additionally, that July, Shedd had rolled out a streamlined single general admission to simplify the purchasing process for all guests.

The old site had been created to handle our now outdated price structure, and with our new goals of accessibility in mind, we hoped to create a new digital purchasing experience that matched our improved and simplified policies with an emphasis on responsive mobile-focused design.

Our four core goals for the new site were:

  1. Evolving our brand under the organization’s new strategic plan
  2. Providing the best possible pre-visit user experience that is both seamless and delightful
  3. Simple, clear, streamlined transactions under the canopy of the new pricing structure
  4. The need for a coherent code base that enables ongoing development

Project team

The in-house implementation team included staff from Marketing, Technology and Design and other departments and individuals were called upon as needed. Marketing held the reins and we worked with an outside vendor to design and implement the site in conjunction with our in-house developers. Once launched we planned to transition bug fixes and maintenance in-house.

Past pain points, new strategy

When Shedd first began the process of laying out what the new site would look like, we knew that one of the top features we wanted to improve was our purchase pathway. A frequent pain point with our old site was that it did not have a persistent cart; a user was required to go through a purchase in one sequence, and if they navigated away, they would lose their progress. This also meant that packages could not be sold together in a single transaction. For example, a guest who wanted to purchase tickets to a penguin encounter was unable to bundle that with a behind-the-scenes tour. Our goal for the new site was to simplify our entire purchase flow into one transaction while also creating the opportunity to let guests know about optional add-ons for their visit.

Previous experience-focused market research we had conducted showed that guests on our old site became lost when it came to learning more about our Extraordinary Experience events (premium add-on encounters to general admission). While each of these experiences had dedicated landing pages and promotion across the site, guests who only entered our main purchase path often missed out on learning about these additional opportunities.

Our new ecommerce path was designed to give users access to all of our offerings in one location. Upon selecting a date and the size of their party, they are presented with a full list of offerings for that day. As guests move forward in the path, additional add-on opportunities like 4-D Experiences are highlighted, as well as membership upsells customized around cart composition and party size. The overall goal was to make sure that our guests were given the ability to fully customize their order as well as make them aware of everything going on at Shedd that would otherwise be buried.

Process and implementation

With a comprehensive ticketing software such as Tessitura, you have the flexibility of building events an infinite number of ways, so looking at what financial and data outputs we wanted as the result helped guide the decisions on how to create these events. We had to outline our needs and test all the options from there. We knew we wanted aquarium admission to be the base of our offerings. We knew we wanted to split revenue between add-on experiences and general admission. We also knew that age restrictions would be a challenge. For the safety of the animals and guests, all our experiences have different age restrictions that would need to be mapped between admission and add-ons. Some features like packages and zones were sacrificed in order to focus on the add-on feature as a priority for launching. 

With that as the basis, the database manager built and segmented pricing in a way that would allow the web developer to handle the enforcement of rules and restrictions online. Price types were found to be one of the easier elements to identify in standard web calls and gave us something on which to base the rule enforcements within the web app. Each event had nine price types that were made up of three age groups (adult, youth, child) and three price groups (standard, resident, member). We worked with programmers to standardize events to this structure. Adults would always be ages 18+, youth 12-17, and children minimum-11. The child price could have any minimum age, or not exist at all.

Add-on events were given separate but coordinating price types. These add-on prices were given a flag in the content management system (CMS) to indicate whether or not these should have “co-sell” restriction with admission in the cart. The events themselves also have a flag in the CMS to determine whether or not they should be displayed in the event calendar, or only after another specified product has been selected. 

There are two ways for tickets to go into the cart. The patron can start with admission and then be prompted to add an upgrade such a Beluga Encounter, or they can start on the Beluga Encounter page and when that ticket is added to the cart, admission is automatically added for them. Because we wanted to accommodate both paths, we had to build in some logic to prevent admission from being removed from the cart when a co-sold item is present. 

Some adjustments were made to allow for outliers like free resident days, then a plan was made for configuring the existing events on implementation day and building new events moving forward.

Once the logic was in place, staff from diverse departments were asked to join testing sessions and compile feedback for refinement. This got stakeholders involved in the process but also gave us some fresh eyes to try out our new purchase path. A simple Google Doc was used in these sessions to quickly compile issues and track bugs. 

Challenges and opportunities

One of the first challenges we encountered with our new purchase pathway was that some of the improvements had come at a cost to our reporting visibility. The entire pathway had been consolidated into a web application running under a single URL, which gave us greater flexibility to add future features and improvements but hindered our tracking in Google Analytics (which tracks users across different URLs). In other words, Google Analytics could see users entering and leaving our pathway, but every action they took during the actual purchase process was now hidden.

When it came to our new upgrades and upsells, we could not track how (or if) users were interacting with them. A drop-off could happen at any point in the pathway, but without detailed tracking, we were unable to know why or where that drop-off occurred. For teams starting on this journey, our recommendation is to have a plan for how your new pathway will interact with Google Analytics right from the start. How will you track what users do? Are there any blind spots you are missing?

We had further challenges when it came to tracking information in our database. While our site differentiated between purchasing an event by itself and adding the same event as an add-on to another product, these were counted as the same type of transaction in our customer relationship management (CRM). An administrator might see we had fifty Beluga Encounters purchased online, but there was no indication which were add-ons and which were normal purchases. Similar to the Google Analytics issue above, this hindered our ability to really know just how well our upgrades were performing. We saw that many events were selling better, but we could not drill deeper to understand the cause. Just recently, we implemented new product names in Google Analytics to differentiate between the two, but our Tessitura data remains a challenge at this time.

California Academy of Sciences: ecommerce project

By Amie Wong


In May 2018 the California Academy of Sciences kicked off a project to replace two key business systems (Raiser’s Edge and Galaxy) with a unified CRM and point of sale (POS) system, Tessitura. The launch date was about a year out on May 14, 2019. This kick-off commenced after an internal team conducted a several-month-long research and request for proposal (RFP) vendor search process, evaluating several types of solutions to meet the Academy’s complex ticketing, constituent management, and business needs. This paper will primarily focus on the web purchase (i.e. “ecommerce”) component of this initiative- and, more specifically, on the strategies implemented to present a more dynamic, intuitive user experience.

Project team

The core working ecommerce team was composed of an in-house team that included web developers, technologists, a visual designer, a web editor, and project manager. We worked with an external vendor who facilitated the initial kick-off, discovery, and strategy development phases of the project, and also assisted with quality assurance (QA) and testing.

Stakeholders included representatives from all departments responsible for products sold online: Marketing and Sales, Membership, Development, and Education. At least one key representative from each department (responsible for representing and providing feedback on behalf of their department), and the ultimate final decision-maker was the Chief Revenue and Marketing Officer.

Goals and guiding principles

The transition to Tessitura presented an opportunity to rethink and redesign a new web purchase experience, one that can leverage the flexibility of the Tessitura ecommerce platform (TN Express Web, or TNEW) and address some of the constraints experienced with Galaxy. With the May 2019 launch, we would also be introducing our customers to dated, dynamically-priced Daytime Admission and NightLife (our weekly Thursday evening event for adults) that would need to be presented in the online purchase path. Though the opportunity was boundless, with the aggressive project timeline and scope of work across all system migration elements, we decided to take a “hybrid” approach to our new web purchase flow, building out a part-custom and part-off-the-shelf ecommerce experience.

Our guiding principles for the redesign of our online purchase experience were to:

  1. Simplify the checkout process
  2. Create a mobile-optimized experience (more than half of the Academy’s web users visit via mobile device)
  3. Keep things user-friendly and intuitive
  4. Provide a seamless, consistent experience between our custom-built web pages and Tessitura (TNEW) cart pages
  5. Optimize overall revenue for the Academy

Overall direction and strategy

The Academy team partnered with an external vendor, LifeBlue, to facilitate kick-off and discovery and define our overall direction and strategy within the first six months of the project. Based on information collected from interviews with key ecommerce stakeholders, understanding our overarching business goals, and identifying the most critical ecommerce pain points to overcome, we formulated the guiding principles mentioned in the section above.

The two biggest opportunities were around simplifying the purchase process and creating a more intuitive experience. We were transitioning from over 30 separate and distinct Galaxy carts (which all had to be built in order to meet our business needs) to a single, more unified cart in TNEW. This presented an opportunity to create a more consistent, standardized user experience and vastly reduce the amount of backend technical oversight.

In addition, with a hybrid (part-custom, part-standard/off-the-shelf) cart experience we had an opportunity to build in contextual upsells (i.e. an offer to upgrade a customer’s selected order and replace with a different product) and add-ons (i.e. an offer to add an additional product to a customer’s order) where it made sense in the purchase process and only present users with relevant offers based on what they have selected.

With the previous structure of over 30 different product carts, we could not upsell or promote add-on products effectively or consistently, and were completely unable to serve up contextually relevant offers. In our new purchase path we wanted to be able to serve a user an upsell or add-on offer only if it brought value to or complemented the user’s previous selections. More details on how we executed this are in the section below.

Process and implementation

Based on our discovery, gathering of business requirements, and strategy work, we developed wireframes that presented a more guided experience that would offer “smarter,” contextual upsell and add-on offers (based on what the user selected)at the right moment in the purchase process. 

We worked with the Membership team to define a logic that will only deliver a membership upsell offer message if our membership levels provided a value-add based on the Daytime Admission ticket quantities and levels selected. Here is some example logic for when/what membership levels we upsell to users based on their Daytime Admission quantity and ticket type selections:

  • 1 Adult/Student Daytime Admission ticket selected = Individual membership in upsell message
  • 2 Adults/Students Daytime Admission tickets selected = Individual Plus membership in upsell message
  • 1 or 2 Senior Daytime Admission ticket selected = Senior membership in upsell message
  • 2 Adults + One or more Child/Youth tickets selected = Family membership in upsell message
  • 3 or 4 Adults + Zero or more Child/Youth Daytime Admission tickets selected = Family Plus membership in upsell message
  • All other Daytime Admission selections do not result in a membership upsell message

Separately, we only display an add-on offer message if it complimented the tickets selected in previous steps of the purchase path. We worked with the Sales team to ensure we were defining what and when certain add-on products would appear in the purchase path. For instance, if a user selects Daytime Admission (and declined the membership upsell offer if one was presented based on logic outlined above), they will see an add-on opportunity for our VIP Tours. Similarly, if a user selects NightLife, they will see an add-on opportunity for our planetarium shows. If the user selects a separate product that did not have a complimentary add-on product, no add-on offer appears.

An additional element that informed the upsell and add-on experience was qualitative user testing that we conducted using basic interactive prototypes (built in InVision). We scripted prompts and tasks for users to complete and we wanted to hear user feedback around the following general areas:

  • Usability: How easy/difficult was this purchase process for you? Is this the length of the purchase experience what you expect? 
  • User experience (UX): Does the order of steps and where elements (i.e. calls to action buttons, offers, etc.) appear presented make sense to you?
  • Content: Is the information provided sufficient and/or useful? What type of upsell messaging is more compelling?

In order to build this custom experience, close collaboration between our web developer and Tessitura/TNEW administrator was required to ensure our custom pages were pulling in the correct products from the Tessitura API. QA was a critical phase in this project—given the backend complexities, we needed to make sure that our logic was implemented correctly and that all use-cases were accounted for and conducted thorough testing.

Challenges and Opportunities

The Academy team celebrated a successful and on-time launch of the new web purchase experience on May 14, 2019. Of course the work is ongoing and—especially given the ever-changing nature of digital platforms—never quite done. The larger migration team has been using the time since launch to stabilize the system and address business-critical needs and issues.

One challenge has been around measuring how effective the new purchase path is. We rolled out a lot of changes at the same time (new ecommerce experience, dated and dynamically-priced tickets) and want to better understand what is working well and what is not. Overall, the Academy has seen some softness year over year in admissions and related visitation products, which makes it difficult to isolate the variables to confidently measure the results of our ecommerce optimizations. We believe that some of these sales trends are due to external factors beyond our control, such as decline in leisure tourism to our market.

While this softness is impacting all channels, including web, there have been some directional bright spots. For example, revenue from our planetarium pass upsell for NightLife has increased 48% year to date (our fiscal year started July 1, 2019), and most of these purchases are online. And during certain peak periods when our dynamically priced daytime admission prices are higher, we have seen an increase in conversions to membership sales, including on the web channel. But overall, it has been difficult to tease out the variables, and we are working to gather more detailed data and analysis.

What’s next?

By all authors

Looking into the future, the Shedd Aquarium’s goal is to implement greater event and data tracking across our purchase pathway to really dig in and understand how these new changes are performing. Our hope is that by tracking our users’ flow we can find better ways to promote and sell our upgrades dynamically with customized user segmentation. The team has only just scratched the surface, but are looking forward to what 2020 brings.

The ecommerce team at the California Academy of Sciences is currently working on a cart optimization project with LifeBlue, and has implemented Google Analytics tracking to better measure the purchase path’s effectiveness in generating revenue across product lines. Much like Shedd Aquarium, their aim is to develop an in-depth understanding of how the new ecommerce purchase paths are performing and what opportunities there are to make improvements to the user experience to ultimately optimize total overall web revenue.

Both organizations currently find themselves asking similar questions with regards to their upsell and add-on experiences. They would like to gain insight into:

  •       How effective are the upsell and add-on steps? Are those ultimately helping increase overall revenue or resulting in lost revenue? More specifically, are they losing more revenue from people who drop off when they see the upsell, compared to the revenue coming in with accepted upsells and otherwise completed purchases?
  •       What UX, messaging, or other updates/improvements can they make to the purchase path and/or process to increase overall web sales and revenue, with overall revenue being priority?

Related to that, the California Academy of Sciences notes that while it has been efficient and in some ways more intuitive to be able to combine 30+ separate web carts into one singular cart, it has also been a challenge to feel confident that their new path and messaging is streamlined and relevant for all products and customer journeys.

Both institutions hope that in 2020 (and for California Academy of Sciences, by spring of 2020) they will have enough data to shed some light on these questions, learn what’s working well and what can be improved, and then build those improvements into their purchase paths.

Cite as:
Northrup-Simpson, Amelia, Abernathy, Amanda and Wong, Amie. "Bundle of joy: Optimizing the purchase experience for add-ons and upsells." MW20: MW 2020. Published January 14, 2020. Consulted .