I’m a great believer in not biting the hand that feeds you, and the world of marketing puts food on my table, so I try to be as objective as possible. I willingly volunteer to respond to marketing surveys, and I’m uber-resilient to hitting unsubscribe these days. Obviously, as product manager for Unica Deliver, I have a vested interest in doing this as it gives me an opportunity to compare and contrast many different messages from many different marketers, and this keeps the creative juices flowing (yes, really).
From time to time, however, I receive something in my inbox that sets my teeth on edge and prompts me to commit metaphorical pen to paper. This time around, I’ve combined my obvious lack of literary talent with the more credible musings of Laura Villevieille, our resident deliverability guru here in the Unica team.
A few days ago, I received an email invitation to participate in a market research survey by a respected organisation. My son is a member of the said nameless organisation, so being the ever-helpful parent/participant, I clicked on the Complete Survey button.
The email linking to the survey helpfully told me that the survey is “short” and should only take twenty minutes to complete. I don’t know about you, but some days, I struggle to find ten minutes to take a quick shower, never mind contributing to worthy marketing causes.
You won’t be surprised to learn that. This time, I just didn’t bother completing it, and this set me off a-thinking about why some marketers still feel the need to trawl for every scrap of information like it’s 1997.
In this blog, Laura and I will explore why this attitude may still exist and highlight some of the recent changes in data privacy and how this is impacting the way we need to communicate with our customers.
In the 1990s, I worked in the data warehousing industry, and the modus operandi at the time was to gather EVERYTHING you could about your customer with a view to creating a single, unified customer profile. At the time, futureproofing anxiety led to the general fear of “what if we need it later?” Vast sums were invested in creating analytical data models that were inflexible and always at least 24 hours behind because of ETL batch lag.
Fast forward to 2004 and my first experience with Unica Campaign; I very quickly realised that the days of gathering EVERYTHING were numbered as a Campaign can easily process data from multiple sources, including operational data sources, thereby eliminating much of the batch lag problem and giving rise more fresh datasets to work with. It was this, combined with the rise in popularity of agile development methods in the mid-2000s, that prompted me to jump ship and join Unica in the first place. If you can’t beat them, join them.
The gather EVERYTHING paradigm hasn’t completely gone away, I know a number of large banks and retailers that still work this way, but we all know that there many ways to approach data integration that cost less and deliver value more quickly.
More recently, GDPR dictated that you should only gather data if you have an immediate need for it. Gathering EVERYTHING isn’t just inefficient– it’s now illegal in the world’s largest market.
It’s also worth mentioning that, in contrast with the mammoth survey I mentioned in the introduction, micro surveys are so fast to respond to it almost feels miserly not to feedback. Delivery companies such as DPD are particularly good at using this method, and more often than not, I respond.
The data privacy issue isn’t going to get any better – in fact, it’s getting tougher and tougher in the industry to track and analyse customer activity. Third-party cookies are out, and Apple MPP is already having a huge impact on email marketing. This is where I hand over to Laura for her take on the current state of email marketing.
GDPR has forced us to completely rethink what data we collect to use in our marketing communications. Recipient attention spans also dictate that it’s completely unrealistic to operate in this gather EVERYTHING mode. There’s no getting away from the fact that marketers NEED data to drive personalisation. So what can we do to collect the data needed to drive personalised and targeted campaigns?
Instead of asking for user interests in a boring 20-minute survey, we can look at user behaviour. Did a user click the ‘finance’ tab on the website? Did they read the article on NFTs in a recent newsletter? Did they make a purchase after a long hiatus? Using open and click data from emails and websites is a great way to gather data about user interests. However, there is a new ‘gotcha’ in doing so. Open and sometimes even click data gathered from email metrics can be inaccurate. Not only are we not able to gather much data about user interests by asking them for it, but also gathering it by observing user behaviour is even tainted. Let’s dive in and look at what exactly is going on.
Apple recently released a new feature that protects users’ privacy by masking open tracking. The idea is that users who value their privacy don’t want marketers to see when they’ve opened a message. In today’s GDPR world, users value their privacy and are no longer ok with something they call the ‘spy pixel’. Marketers use an invisible 1-pixel image that, when fetched with the rest of the images in an email, triggers an open event in reports.
The way these feature works is that Apple pre-fetches open tracking pixels for all messages instead of fetching them only when a user opens the message. So, users opening their mail from within the Apple mail app on an Apple device will show an open event even when they did not open the message. This means open rates for mail opened in Apple will be reported at higher rates than it actually is. How inflated are the rates? Among our senders, we saw that, on average, a sender’s (estimated) true open rate was about 90% of the (inflated) open rate in reports.
But before Apple even released this, other mailbox providers like Yahoo and Gmail were pre-fetching all images in an email in order to provide a better user experience. By loading images prematurely, they are able to show emails in the inbox much faster. So actually, opens in email have been inaccurate for a while.
Add to this that there are corporate email filters out there that pre-fetch links in messages in order to scan them for security. Due to the growing volume of phishing scams, this is a prominent and growing feature incorporating spam filters. And it’s a total nuisance for senders! Click rates can be totally inflated when this happens. We’ve seen click rates as high as 90%.
All this means that open and click data gathered from emails is unreliable. This open data can’t be relied on for important segmentation, automation flows or A/B tests. We need to be careful using opens and clicks as the basis for determining a user’s interest in a thing. And marketers need to re-evaluate current automation and segmentation to work around the unreliability of open data from Apple and others.
There is an interesting movement coming from a few ISPs who built mechanisms on sharing open metrics with senders at an aggregate level. Yahoo has recently made available to senders, for a fee, a data feed with aggregate open rates. Over the years, other ISPs have made similar reputation portals with this kind of aggregate level data, like Mail.ru or QQ Tencent. These portals usually share a sender’s open rates per day and per sending domain. In the case of Yahoo, these can be available at a campaign level.
The data is wonderfully accurate, as it’s not relying on an open pixel. Yahoo has immediate access to whether a user has actually opened an email. Even though we only have data like this from a very small portion of senders, the hope is more will follow suit. Over time, this may be the way we view engagement data for email campaigns.
A lot has changed in the marketing ecosystem since the 1990s. GDPR spawned a real change in mindset around user data and privacy, and because of Apple MPP, we have to work in a world where we have fewer response data. Add the fact that the sheer increase in the volume of email has impacted user attention spans and expectations. Marketers will continue to evolve, and here are some tips for doing so.
When it comes to requesting feedback from your customers, the principle of only gathering what you immediately need should apply. There are ways that will consistently motivate a response, and there are ways that will achieve exactly the opposite effect. Some delivery companies are particularly good at distilling the major points of interest down into a painless exercise that barely lasts five seconds. Like the DPD example, think of using quick polls or short in-email surveys.
Where a more detailed market research exercise is needed, getting to the point is key. If the market research exercise is something that you plan to perform every year, then you really need to hone it down and trim every ounce of fat from your survey.
We know we need data to base our segmentation and personalisation on. It’s not so much about gathering this data but about being a great data wrangler. Look at data that has already been captured, like website visits, email content link clicks, landing page visits. The work has to be done to detect user interests and needs from this data and combine that with what we know about a user to form intelligent communications.
User privacy protection has changed the way we use open data. Rethink automation flows, journeys, re-engagement campaigns, segments and KPIs that use open rates to indicate user interest. Most likely, these will have to be modified. There are still a great many metrics available to senders that indicate user interest or disinterest, like link clicks, unsubscribes, complaints, landing page visits, form field responses, website visits, purchase data and opt-in date. When using open rates as a KPI, look at them as a directional indicator. Threshold values, like 30% is good, 15% is bad … are no longer helpful. Instead, look at whether rates are increasing or decreasing from past performance.
If you would like to learn more about Unica Deliver, the latest industry trends and Apple MPP, please feel free to get in touch.