How ISVs can handle peak messaging volume during Cyber Week

Cyber Week is the event ISVs prepare for all year because it compresses a significant amount of operational risk into just a few days.

A single retailer preparing for Cyber Week has one traffic spike to plan for, one set of channels, and one audience. An ISV has to prepare for the combined demand of its entire client base, with many of those traffic spikes occurring at the same time.

Every retailer may want its doors-open campaign delivered at 9am ET on Black Friday. Their highest-traffic week runs on your infrastructure, and any delay or queue in delivery, reporting, or escalation can affect multiple clients at precisely the moment performance matters most.

During Cyber Week 2025, the Infobip platform carried 12.2 billion interactions from thousands of businesses sending into dozens of markets, without significant delays. That outcome is built in the weeks and months before the spike, through forecasting, capacity planning, prioritization, channel planning, and monitoring established before the first seasonal message is sent.

12.2 billion interactions were carried on the Infobip platform during Cyber Week 2025, without significant delays.

This blog walks through what your messaging layer needs to be ready for, in the order you should be addressing it:

• forecast the load

• provision and test capacity

• establish prioritization and queuing rules

• plan the channel mix

• put the visibility in place to act in real time

Monthly volume forecasts are important for commercial and capacity planning, but Cyber Week requires a more granular view. Traffic concentrates into a small number of days and hours, and within those hours, often into the first few minutes.

ISV-level capacity planning therefore needs to work at three levels of resolution:

• Monthly, for commercial and longer-term capacity planning

• Daily, to identify which days will carry the greatest load

• Hourly, to identify the periods when aggregate demand may exceed available throughput

The hourly view requires input from your clients. Well before November, collect expected volumes, target send windows, channels, and campaign types so individual plans can be viewed as a single platform-level forecast.

The distinction between volume and throughput is important. A client planning to send three million messages “at 9am” requires 50,000 messages per second if the entire batch is released within one minute, but approximately 830 messages per second if it is distributed across the hour.

The total volume is identical. The throughput requirement is not.

Multiply that difference across every client targeting the same send window, and you begin to see the real shape of Black Friday morning.

Campaigns are frequently scheduled on the hour. When hundreds of campaigns begin at exactly 9:00:00, demand can exceed available throughput within seconds, creating queues that affect multiple campaigns simultaneously.

One of the simplest ways to reduce this pressure is to work with clients on send scheduling:

• Schedule away from the full hour. A campaign launched at 8:53 or 9:23 can avoid the concentrated demand created by campaigns scheduled for exactly 9:00, without materially changing the experience for the recipient.

• Spread large sends across a defined window. Delivering a campaign over 30–60 minutes can significantly reduce peak throughput requirements. It also provides an opportunity to observe early delivery and engagement data before the full audience is committed.

• Stagger clients at the ISV level. Individual clients can see their own campaign schedules. An ISV can see those schedules in aggregate. Use that visibility to identify overlapping peaks and smooth demand before campaigns begin.

None of these adjustments changes the total message volume. They change when that volume enters the network. Small offsets in scheduling and wider send windows can materially reduce peak-minute demand and help prevent queues from forming during the busiest periods.

Once the forecast is established, capacity planning becomes much more concrete.

Translate each client’s send plan into expected peak-hour and peak-minute throughput requirements by channel, then provision headroom above that forecast to accommodate unexpected demand. Confirm those requirements with your messaging provider well before peak season so the necessary capacity can be provisioned and tested in advance.

Load testing should validate not only overall platform capacity but the paths that traffic will actually take under peak conditions.

The final step is a configuration freeze in the period leading up to Black Friday. Routing configurations, sender setups, integrations, and deployments should be stabilized rather than changed immediately before the highest-volume period of the year.

Carriers follow similar practices and may freeze sender and campaign registrations around holiday periods, which is why registration should be completed during September and October.

If you are onboarding new clients for the season, workflows such as Infobip’s APIs for number provisioning and sender registration, together with Provisioning MCP, can streamline sender setup across channels and help ensure onboarding is completed before registration windows close.

Even with headroom, there will be minutes when demand briefly outruns throughput. What happens in those minutes should be decided by rules you set in advance rather than by queue order, and the right unit for those rules is the use case as well as the client:

• OTPs and login codes go first: A one-time PIN that arrives several minutes late can result in a failed login or checkout at exactly the moment the customer is trying to complete a transaction.

• Order and shipping updates go next: Transactional messages carry an expectation of immediacy that promotional traffic doesn’t.

• Promotional blasts absorb the delay: A flash-sale message that lands four minutes later still converts. This is a controlled slowdown instead of a stalled platform.

These rules also need to operate across clients, not only within an individual account. One client’s high-volume promotional campaign should not prevent another client’s OTP traffic from being delivered promptly.

That requires prioritization at both the client and campaign level, with the ability to adjust priorities as conditions change. Static rules established before the event are useful, but real-time control becomes important when actual traffic differs from the forecast.

This is one of the infrastructure layers Infobip manages for ISVs. Traffic can be prioritized and queued across clients and use cases, with priorities adjusted as traffic patterns change during peak periods.

In the US, messages are delivered through the recipient’s carrier, which means a slowdown on a particular carrier path cannot always be solved by simply rerouting the same message elsewhere.

The flexibility ISVs do control is the channel used for a message and the fallback path available if the preferred channel cannot be reached. Channel mix therefore becomes a capacity and resilience consideration as well as an engagement decision.

For promotional traffic, branded messaging is becoming increasingly important. Rich channels such as RCS allow retailers to present their brand name, logo, and verified sender identity directly within the user’s native messaging experience.

RCS also supports capabilities such as product carousels, quick-reply buttons, URL previews, and delivered, seen, and clicked analytics, without requiring the recipient to download another application.

Reach, however, still depends on the recipient’s device and network. Every branded campaign therefore needs an SMS or MMS fallback, supported by device-aware routing that checks capability and moves to the appropriate fallback when RCS is unavailable.

During peak periods, this means maintaining two operational paths per client, both provisioned and tested before registration and configuration freeze windows.

Oversized images can fail or be degraded differently depending on the carrier. Peak week is not the time to discover those differences.

Infobip can automatically resize and optimize MMS media according to carrier-specific requirements before sending, reducing media-related delivery failures without requiring ISVs or their clients to manage those differences themselves.

Check out our docs on MMS resizing

Email can absorb significant high-volume promotional traffic, but only when the sending infrastructure has been prepared in advance.

IP and domain warming takes time, so new sending infrastructure should not be introduced immediately before Cyber Week. Mailbox providers may throttle sudden increases in volume, making it important to begin ramping toward expected peak volumes in October and to send to clean, engaged lists.

Allowing clients to suddenly send large volumes to dormant addresses during Black Friday can create reputation and deliverability issues that persist beyond Cyber Week itself.

Everything above is preparation, but during the event itself, visibility becomes critical. If a queue begins building on a particular carrier path, an ISV should be able to identify it across clients, senders, and carriers, including queue size and estimated drain time, before individual retailers begin reporting delivery problems.

In practice, that means establishing alert thresholds in advance, providing reporting granular enough to be useful during a day like Black Friday, and maintaining real-time visibility into carrier-level queues.

The objective is to move from “a customer told us delivery dropped” to “we identified the queue and acted before it materially affected the customer.”

Visibility should also enable action while campaigns are still in flight to allow for pausing a send that is contributing to a queue, reprioritizing traffic by campaign or entity, or shifting eligible traffic between channels.

Before shopping season begins, confirm which of these controls your infrastructure exposes in real time. Visibility is valuable, but the ability to act on what you see is what makes it operationally useful.

Forecasting pipelines, load testing, prioritization engines, carrier-level monitoring, and shopping season operational support are all critical to the success of your clients’ Cyber Week messaging strategies.

Standard support models are designed for steady-state operations. Peak season is different. What matters during the busiest periods is access to a cross-functional team spanning routing, connectivity, reliability engineering, and product—one that understands your environment before the event begins and can respond in real time when conditions change.

ISVs already make this trade by using Infobip as the infrastructure layer beneath their products, with:

• direct operator connectivity

• transparent commercial terms

• compliance capabilities designed for scale

For shopping season, that relationship extends beyond an SLA.

In the months leading up to Cyber Week, Infobip teams work with ISVs on forecasts, throughput requirements, connectivity, and prioritization so configurations can be planned and tested before traffic peaks.

On Black Friday, ISVs can access granular reporting, alerts based on defined thresholds, real-time visibility into queues and estimated drain times by carrier, and the ability to reprioritize traffic by campaign or entity while traffic is in flight.

Behind those capabilities is a cross-functional team spanning routing, connectivity, product, and reliability engineering, supported by additional data centers for load balancing, stability, and resilience at high volumes.

The objective is straightforward: absorb the complexity of peak-season messaging at the infrastructure layer so ISV engineering teams can remain focused on the products and experiences that differentiate them.

If your retailers are asking for deeper visibility ahead of Black Friday, you experienced delivery delays during last year’s peak, or you have high-priority traffic that needs to move ahead of promotional queues, talk to us about preparing your messaging infrastructure for Cyber Week.