You spend large sums of money on your marketing automation platform (MAP), and get some lovely reports from the system, but something just doesn’t feel right. You know that although the numbers look credible, they don’t tell you the whole story. What’s going on?
The reality is that, despite those percentages having two digits after the decimal point, MAPs cannot give you precise data. There are many challenges that make it hard to get accurate information from your marketing automation system. Although it’s frustrating that you can never have the full picture, if you understand the issues, you can get a much better understanding of which campaigns are working and those that you need to revise.
Don’t Believe Email Opens
Email opens are just not accurate. Probably the most publicised reason is Apple’s decision to hide whether users of Apple Mail open emails or not. An email open is recorded when the recipient downloads an image (often called a tracking pixel). Email opens have never been a totally accurate metric, as if you don’t download images, the open isn’t recorded.
The bigger problem today, however, is that Apple is automatically downloading images for emails, whether or not the recipient actually views the email. There is no way of knowing whether users who have enabled Apple Mail Privacy Protection have actually viewed the email, and this is likely to be a technique adopted by more email systems in the future.
Those Email Clicks could be Bots
This is a particular problem in B2B marketing, where companies install malware prevention systems that analyse incoming email. If the system decides there is a risk, it will follow the link to check for the presence of malware. Following a link is recorded as a click by your MAP. If you’ve ever seen that there are people who click on links as soon as they receive an email (or click on all the links in the email), you’ve seen a malware bot in action.
Although most MAPs try to filter out the obvious bot clicks, they are nowhere near 100% effective in trying to give an accurate picture of the clicks made by real people. This makes sense – if it was easy to detect a malware detection bot, then those evil scammers would be able to serve benign content to the malware detectors while delivering viruses and other nasty content to you and me when we click. The need to confuse the bad guys means that there is little prospect of eliminating bot clicks from your MAP reports in the near future.
Landing Pages Aren’t Immune to Problems
We’ve seen that the two major metrics we use for email reporting – opens and clicks – are inaccurate. The good news is that landing pages are more immune to errors, but they’re not perfect either. Unfortunately, you’ll get traffic to the landing page that isn’t from people responding to your marketing automation campaign. You can reduce this by ensuring that your landing pages request that search engines don’t index them and making sure that there are no links to them from any other pages, but there is no guarantee that all your visitors will be genuine.
If you are running a campaign with a public landing page and using marketing automation emails to drive traffic, it does help to have two versions of the page – one that is public and the other that is not in the search engines’ indexes for your emails.
People Reply to Emails
The problems with reporting are not only due to technical issues. People also reply to emails, particularly when they are a sales connect email that feels personal. This type of email is often the most effective email in any campaign, but this doesn’t always show in the MAP reports.
As marketers, we’re trying to make the email feel as “real” as possible – i.e., that it was sent from a salesperson who was interested in the prospect. But this means that the recipient is likely to respond by replying to the email, rather than clicking links. If you use a personal reply-to email address (which will improve the performance of the campaign), you won’t be able to track the replies. You’ll need to trust the salesperson to report the interactions, and we all know that sales teams are not always super-keen to give marketing the credit they deserve!
Attribution: It’s not all about Marketing Automation
Marketing is messy. These pesky prospects engage with multiple assets from a range of campaigns during their customer journey. To make it worse, very few prospects follow the idealised customer journey you created when planning the campaign.
The problem of attribution isn’t new, but it can be a real challenge around marketing automation campaigns. People respond after the campaign ends because they are not in-market at the time of the campaign. Alternatively, they see and respond to another marketing asset, but this does not mean that the impressions from the automation campaign were not instrumental in driving a conversion.
Although tools, including marketing automation systems, are getting better at attribution, we’re going to be facing this challenge for a long time to come.
Scoring Isn’t the Solution
Scoring is seductive. The MAPs offer the ability to create some magic algorithm that predicts who is most likely to become a customer, and sometimes the systems will even use AI to create that algorithm. It just feels so right. But scoring is a million miles away from perfection.
Although there is undeniable logic to scoring, the underlying assumption is that everyone follows more or less the same customer journey. Yet we know they don’t. So a score of 100 for one contact might be equivalent to a score of 200 or 50 for other contacts who are following slightly different customer journeys.
It’s true that, provided you have a lot of reliable data about prospects and conversions, the use of AI can help make better models than the simple manual “points mean prizes” approach. Even if AI could create the perfect scoring system, your prospects will change the customer journey over time, and you’ll run different campaigns. However good the AI is, it has to run on historic data, so it can never generate a scoring system that matches the behaviour of prospects today.
Your Data isn’t Up to Date
If you want accurate information about performance, you need good data. Unfortunately, it’s hard to keep data in a marketing automation system, so your results will be skewed by any out-of-date or other data quality issues. Many of those email addresses will accept the email, and it will just disappear into a black hole, rather than bouncing back. Making sure your data is current and accurate is an important part of getting the best quality report from your MAP.
Marketing automation systems are amazing and offer great insights through their reporting. But the reports can’t be perfect. By understanding the main things that cause issues in the reports, you can improve your chances of getting far better insights into your campaigns.