Sinch Mailgun has published research showing that nearly 18% of business emails fail to reach the inbox, putting up to 20% of potential email return on investment at risk.
The findings come from the Email Impact Report 2026, which combines data from more than 400 billion emails sent in 2025 with a global survey of more than 1,200 email senders. The report sets benchmarks across 10 high-volume sending sectors and highlights weaknesses in measurement, deliverability practices and the use of artificial intelligence.
Email remains an important commercial channel for many businesses. Sinch Mailgun found that 78% of respondents said email is critical to business success, while 79% plan to maintain or increase investment in the channel.
At the same time, the data suggests many companies are struggling to turn that reliance into consistent returns. Fewer than half of organisations surveyed said they could confidently measure email return on investment, even though 60% of those that do reported returns of more than USD $10 for every USD $1 spent.
More than one in 10 companies that measure email return said they achieve returns as high as 40:1, according to the report. Yet weak inbox placement appears to be reducing the value of those campaigns before recipients even see them.
AI Use
The report also points to mixed results from the use of artificial intelligence in email operations. While 46% of respondents said AI improves speed and efficiency, 23% said it has not improved their email programmes at all.
Use remains concentrated in basic tasks. The research found that 41% of teams use AI to generate email content rather than applying it to areas such as segmentation, optimisation and deliverability.
That gap matters because the report links stronger results to broader use of AI in email workflows. Organisations that apply AI more effectively are significantly more likely to report improved email performance, although the report does not provide a precise comparative figure.
Across the survey sample, 49% of respondents said email performance had improved year on year. Even so, the overall picture suggests that gains are uneven and often constrained by operational issues rather than a lack of spending.
Kate Nowrouzi, Vice President of Deliverability at Sinch, said many organisations are not set up to make the most of the channel.
"Email delivers exceptional returns, but many organizations are not set up to capture its full value," Nowrouzi said. "The gap between performance and execution is where most are losing out.
AI adoption is widespread, but its impact remains uneven. Many teams focus on basic use cases such as copy generation, while higher-impact applications such as optimization, segmentation and deliverability remain underutilized. Organisations that use AI more effectively are significantly more likely to report improved email performance."
The findings reflect a broader issue in digital marketing: businesses can measure clicks, opens and conversions in detail but still struggle with more basic operational questions, such as whether messages are consistently reaching the inbox. In practice, that can reduce campaign performance without being immediately visible in headline reporting.
Deliverability has become more prominent as mailbox providers tighten rules around sender reputation, authentication and relevance. For large senders in particular, small declines in inbox placement can translate into a meaningful drop in revenue, especially for retailers, software groups and other businesses that depend on email for acquisition, retention and transactions.
Sinch Mailgun's dataset is drawn from high-volume senders across 10 industries, giving the report a broad base, though the company did not break out performance by sector in the summary findings. The combination of observed sending data and survey responses offers a view of both technical performance and how teams assess their own practices.
Nowrouzi said the biggest gains from AI are likely to come from operational uses rather than writing copy.
"Using AI to generate copy is a good starting point, but it's not where the biggest impact happens," she said. "Organisations that apply AI to optimization, segmentation and deliverability are seeing stronger results."