Email Bandwidth

Definition

Email sending bandwidth, more accurately described as throughput capacity, measures the rate at which emails can be transmitted from your sending infrastructure to recipient mail servers. Unlike network bandwidth measured in bits per second, email throughput is measured in messages per second (MPS) or messages per hour. Effective throughput is determined by a combination of your MTA capacity, the number of sending IPs available, and the receiving limits imposed by each destination mailbox provider.

Typical throughput limits vary significantly by destination. Gmail typically starts new senders at approximately 100 emails per second per IP, gradually increasing as reputation is established. Outlook/Hotmail is more conservative, often limiting new senders to 30-50 emails per second per IP. Yahoo and AOL are generally more generous at 200-500 emails per second per IP. Smaller or less well-known mailbox providers may have limits as low as 5-10 emails per second. These limits are not published and can change without notice, so senders must monitor deferral rates to detect when they are approaching a provider's threshold.

Bandwidth planning is essential for campaigns that must be delivered within a specific time window. A campaign of 5 million emails sending at 200 emails per second across 4 IPs (800 MPS total) will take approximately 104 minutes to deliver, not accounting for warmup, deferrals, and retries. In practice, total sending time is typically 1.5-3x the theoretical minimum due to deferred deliveries (temporary failures that require retry). For time-sensitive campaigns, bandwidth must be provisioned to ensure delivery within the desired window after accounting for expected deferral rates.

Best Practices

Calculate required throughput capacity based on largest campaign size and desired delivery window. If your largest campaign is 2 million emails and you need delivery within 4 hours, you need sustained throughput of approximately 140 MPS (2,000,000 / 14,400 seconds), accounting for a 1.5x buffer for deferrals. If your current capacity is insufficient, add IPs or increase per-IP throughput through warmup.

Provision sending IPs at a ratio of approximately 1 IP per 100,000-200,000 monthly sends. This rule of thumb provides adequate throughput for most campaigns while maintaining concentrated sending for reputation purposes. A volume of 5 million monthly sends would need 25-50 sending IPs. Fewer IPs means higher per-IP volume and potentially stronger reputation, but lower throughput capacity for time-sensitive campaigns.

Implement gradual throughput ramp-up at the start of every campaign. Do not begin a campaign at full throttle. Start at 25% of your maximum throughput and increase over 10-15 minutes to full rate. This gradual ramp reduces the likelihood of triggering ISP rate limits or appearing as a sudden burst of suspicious traffic. Many MTAs and ESPs support automatic ramp-up configuration.

Monitor per-ISP deferral rates as indicators of bandwidth saturation. If deferral rates from a specific ISP increase during a campaign, you are likely hitting their rate limits. Pause sending to that ISP temporarily or reduce the per-IP throughput to that destination. Continue sending to other ISPs unaffected. Real-time per-ISP throttle management prevents campaign-wide delivery delays.

Plan for burst capacity during peak periods. Holiday campaigns, flash sales, and product launches often require 3-5x normal throughput capacity. Maintain reserve IPs that can be activated during peak periods. These reserve IPs should have their own warmup schedule and reputation profile so they are ready when needed. Cloud-based sending services typically offer on-demand burst capacity.