When developing minimum viable products (MVPs), companies need to prepare for traffic peaks. If
a product hasn’t been designed to scale to meet increased demand, the application will experience
outages that frustrate users and may lead customers to choose a competing product.
Even leading brands take blows to their reputations when they experience outages due to traffic
spikes. Ticketmaster is facing legal fallout due to website failures that occurred when Swifties tried
to buy tickets for Taylor Swift’s The Eras tour. Fans experienced long wait times and were
sometimes kicked out of the queue due to extreme traffic.
Outages are common and devastating for companies. Downdetector reported 10 global outages
that had happened by the middle of 2023. These outages negatively impacted popular platforms
such as Reddit, Microsoft 365, Instagram, and YouTube.
While many types of outages are difficult to prepare for, companies can develop their products to
scale for seasonal traffic peaks because these are predictable. Using Amazon Serverless
technologies, such as Amazon API Gateway, AWS Lambda, and Amazon DynamoDB, enables
companies to hyperscale to prevent outages by predicting and meeting performance expectations
during seasonal spikes in user traffic.
Successfully launching an MVP depends on understanding when products in your industry
experience the most use. Typically, these spikes in user traffic occur during seasons or around
special events related to your industry.
Because these spikes occur at predictable times of the year, companies can prepare by
anticipating how high the traffic will surge. Examining some specific e-commerce use cases helps
illustrate how companies can predict and prepare for seasonal spikes in traffic.
The retail industry experiences huge spikes in traffic during the holiday shopping season. Black
Friday launches the season, as e-commerce retailers offer special discounts. According to Digital
Commerce 360, online retail customers spent a record $9.2 billion on Black Friday in 2022.
Food delivery and online grocery sites see huge increases in user traffic around the time of the
Super Bowl as consumers stock up on snacks, such as pizza and wings, to serve at their viewing
parties.
From January until the middle of April, tax service applications and accounting software, such as
Quick Books, undergo a surge in user traffic as taxpayers rush to get help with preparing and filing
their returns.
Around New Year’s, people start feeling remorse about their weight gain over the previous year,
resulting in health-related resolutions. These resolutions create user traffic on sites that offer
wellness subscriptions, such as gym memberships and weight-loss programs.
Preparing for increased user demand during seasonal spikes starts in the application development
process. To meet seasonal demands on their products, companies need to rethink the traditional
monolithic approach to application development, which requires a full release every time a new
functionality is added to an application. With this approach, if an error occurs in one module during
a seasonal period of increased demand, the entire application may fail to function, leading to costly
downtime and lots of unhappy consumers.
A dramatic increase in the number of users due to seasonal traffic may cause the application to
perform poorly when it can’t handle increased transactional demands. As more consumers use the
application, they may experience inconsistencies in response times that reflect poor performance
and negatively impact customer experience. Increased traffic can even block access to the
application and cause data loss.
To get away from the monolithic approach to product development, companies need to divide the
functionality of their retail and e-commerce applications into independent feature-specific
components.
By using Amazon API Gateway to access data, business logic, and functionality, a business can
understand how users of e-commerce applications access data and behave during transactions so
they can scale successfully during seasonal traffic, i.e., in the use case of e-commerce
applications: catalog searches, loading carts, and moving to checkout generate data related to
purchasing behavior.
Amazon API Gateway is used for creating, publishing, maintaining, monitoring, and securing REST
and WebSocket APIs at any scale. With Amazon API Gateway, developers can create APIs for
use in their own client applications or they can make Web APIs available to third-party app
developers. Amazon API Gateway enables developers to provide a secure and highly configurable
entry point to other resources and applications inside AWS or other platforms.
To scale successfully during seasonal traffic, a company needs to track key metrics that reflect
business priorities. These metrics allow the company to analyze throughput by starting with
assumptions about the amount of throughput per transaction and the expected rate of increase in
transactions during the season in question.
For e-commerce applications, the company may predict adding a certain number of users during
the season, with an average number of transactions, and an average duration. Transactions may
include catalog searches, adding products to a cart, and completing checkout.
Using these metrics, the company can calculate how many transactions to expect during peak
traffic times and how much throughput time is required. To handle the seasonal growth, the
business needs to find a way to manage data more efficiently.
To analyze data access patterns for an e-commerce application during seasonal traffic, a company
may need to focus more on reads than writes because catalog searches outnumber purchases.
Other types of applications may require the company to focus more on writes than reads
transactions.
Separating writes from reads and moving them to a database allows companies to record and
review data related to seasonal transactions so they can see where time is being spent. Analyzing
data access patterns and transaction times shows how seasonal spikes in traffic may cause
database transaction management issues.
When Amazon DynamoDB serves as the No-SQL database for companies developing e-
commerce applications, the solution keeps writing and reading times constant and keeps the
schema flexibility of the data stored. The consistency provided by Amazon DynamoDB is a crucial
factor in designing asynchronous transactions that help scale volume substantially during traffic
spikes.
By scaling with microbatches, companies that experience seasonal traffic spikes can optimize
throughput for transactions. Microbatches allow the same amount of information to be sent using
fewer transactions because companies can insert records into the database in bulk. With
microbatches, companies can develop less expensive products with potentially higher profit
margins.
When companies can predict projected seasonal growth in the number of users, they can meet
capacity for needs for resources and costs. Load testing using a defined number of additional
users makes estimating the demands of seasonal traffic and its costs easy.
With load testing, the company can predict the seasonal scalability of the solution to achieve the
linear, logarithmic, or exponential scalability needed to accommodate the computational complexity
of the application and the transactions it performs.
Predictable scalability has economic benefits for companies. For e-commerce applications, a
calculator can be created that allows the company to plug in numbers of users and assumptions,
such as number of catalog searches and average items per cart.
The calculator can be used to create pricing for cloud-native e-commerce solution packages. By
calculating pricing, companies can ensure that their applications are both affordable and profitable.
The good news about seasonal traffic spikes is they can be predicted. Unlike outages caused by
breaches, downtime from seasonal demands can be prevented when companies put in the work
during development to understand data access demands during transactions.
Carrying out data access analysis and using a calculator for load testing will help your company
prepare for and manage seasonal user traffic. IO Connect Services can set up a demo for load
testing. As an AWS Advanced Tier Services Partner, we have the expertise needed to put together
the right architecture for developing applications capable of hyperscaling using Amazon API
Gateway and Amazon DynamoDB.
Learn how hyperscaling with AWS can help your company prepare for seasonal user traffic.
Contact us for a free consultation
