Email Deliverability - Is 98.5% Good Enough?
24 Jul 2020
The DMA’s Email Benchmarking Report 2020 is the UK’s definitive set of performance metrics. Taken from a broad set of email service provider (ESP) contributors. The report illustrates “what good looks like” for UK senders when it comes to delivery, open and click rates.
The average delivered rate for 2019 was 98.5% and the report celebrated the fourth successive year-on-year increase for this metric. The analysis rightly pointed to the GDPR “halo effect” with better data quality creating a delivery dividend for senders.
So a good news story at face value, and vindicating the 23% of a typical email marketing budget spent on deliverability – as we saw in our research with the DMA. But let me pose a contrarian question – is it good enough?
- List Churn
A 98.5% delivered rate means each send generates 1.5% bounces. Let us assume half (0.75%) are hard bounces which must be suppressed – a program averaging 2 sends per week will churn its entire list in just over a year, and that’s before accounting for opt-outs and complaints.
This carries a big opportunity cost – the DMA’s Marketer Email Tracker report calculates average customer lifetime value of an email address at over £34, which is lost with each suppression! The lost addresses also need replenishing, which also comes at a cost. The ANA/DMA Response Rate report in the US calculates average email cost per acquisition at $15.75 (± £12.50).
- Delivery vs Inbox Placement
Many program owners still view deliverability purely in terms of bounce rates. This only considers whether emails were accepted by the receiving MTA. Just as important is whether the emails are then placed – in the recipients’ inboxes or their spam/junk folders. Validity’s 2020 Deliverability Benchmark report shows average inbox placement rates (IPRs) for the UK are ± 88% – meaning 1 in 8 emails effectively fail to deliver.
This trend is highlighted in Econsultancy’s Marketing Census report. While 30% of respondents use delivery rate to measure email success, only 10% use inbox placement rate. The DMA’s ‘Email Deliverability: A Journey into the Inbox’ report shows that while 11% of respondents rate their ability to measure deliverability as “poor”, this increases to 19% for measuring inbox placement.
- Bounce Processing
Another surprising learning from the ‘Journey to the Inbox’ report is that 19% of respondents fail to suppress their bounces at all. This flies in the face of deliverability guidance from the major Mailbox Providers (MBPs):
- Automatically unsubscribe users who have multiple bounced messages (Gmail)
- [Our] email filters are influenced by a number of factors related to . . . list accuracy (Microsoft)
- Remove invalid recipients from your list promptly (Verizon Media)
The negative impact of poor bounce handling is seen in Validity’s Sender Score Benchmark report. Top senders (scoring 91-100) generate average unknown user (hard bounce) rates of 0.9%. Just one tier down (81-90) this increases to 3.2%, an increase of 2.3%. The difference in IPRs between these two tiers is ± 10% meaning each bounced address also results in 3 good addresses failing to deliver because of the degraded reputation.
So, while 98.5% is indeed a decent delivery performance, there is plenty of room for improvement. What should email program owners be doing to further move the deliverability needle? Here are 5 tips:
- Monitor your reputation using a tool like Sender Score. Most major MBPs use reputation as a key factor in their spam filtering algorithms, and a poor reputation means increased rejections.
- Ensure you are capturing good email addresses by using double opt-in as your permissioning model. Or implement a list validation solution such as BriteVerify, meaning incorrect addresses are remedied at point of capture.
- Adopt the most robust bounce-handling strategy your program will support. This means hard bounces being suppressed on the first occurrence, while soft bounces are suppressed on a “3 strikes and you’re out” basis.
- Implement a rule set to ensure “false positive” hard bounces are not inadvertently suppressed. MBPs are known to generate hard bounce codes where the accompanying text indicates a non-permanent condition (e.g. 550 – “Reasons for rejection may be related to content with spam-like characteristics or IP/domain reputation”).
- Monitor inbox placement rates and tabs placement rates using solutions like Return Path or 250OK. Understanding where emails are being placed means senders can then start to reduce the “1 in 8” going to spam.
While the 98.5% delivered rate is indeed encouraging, deliverability is a much broader discipline than just monitoring bounce rates. Make sure you have visibility of all the factors that ensure your subscribers see their emails. Out of sight is out of mind – improved inbox placement is the starting point for better engagement to maximise your ROI.