I had a boss once who introduced me to the phrase, “in position to be in position.” He would use it to mean we had made enough headway on an initiative to almost be ready to see some results. It would be nice if all initiatives in business worked like a light switch, and you could just switch them on and they’d go. The reality though is that most initiatives – whether they’re a product launch or a marketing promotion or a new business partnership – have a discovery period that you enter with a hypothesis, but one you should be prepared to modify if your initial assumptions don’t bear out. In my experience, I’ve found myself “in position to be in position” after this period, where I have some conviction in my hypotheses and can start to execute on the strategy.
Now that we’ve started using and testing Social Magnet, we’re in position to be in position to learn a lot about email + social. Combining email and social in the ways Social Magnet allows is all new, and I don’t yet know what we’ll learn. In fact, we’re in the stage of learning what we’ll be able to learn – another version of being in position to be in position. Here are some of the assumptions we have going into this process, what we’re testing, and what we hope to learn:
Impact of social media on email results: One of our assumptions is that there is a lot of crossover between the email and social audiences of many companies, particularly in the B-to-B space. If that’s the case, then messaging in all channels increases frequency among the people who are attentive in all channels. So what happens when the same message is promoted across email and social? Will social messages lift email results from the added frequency? Or will the results from social cannibalize the results from email if they reach the audience first? We ran a test earlier this week promoting a webinar to see. We ran an A/B test of the email list, mailing to the first half before we promoted the same link on Facebook and Twitter, and then mailing to the other half after the social messaging. We saw a difference, but not one that was statistically relevant. Not yet anyway – if we see the same small difference over the next 4 or 5 tests, we might be onto something. Or we might have to revise our assumptions. Or our results may change as our list and audience sizes change (our email list is significantly larger than our social audience currently). As you can imagine, there are a ton of variables that can’t be isolated for a perfect testing environment. So for now, we’re not testing our results as much as testing our tests.
Impact of email on social media results: Social Magnet allows us to measure the number of clicks and (if we’re using Real Magnet’s events module for registrations) even conversions that come from all of our channels. Just as social messaging may impact email’s results, it’s equally likely that people who see something in their inbox may be more inclined to act on a message from the same brand on Facebook or Twitter. It’s plausible, at least. So we’re looking at ways to test that, and see if we can identify some situations where email messages that may not drive inbox engagement still lift results measurably in other channels. Intuitively, it makes sense. It’s only now that we’ve had a tool to put some serious thought into how to measure it.
Performance of different message types by channel: Like most marketers, we promote everything in all our channels. If something goes onto the blog, it may also show up in an email newsletter, on Facebook, Twitter and in LinkedIn. Some of what we do I’d classify as informational – like this blog, for example, or Tuesday’s blog about Subject Lines, or Chris’ blog earlier this week about Deliverability Numbers. Other messages are promotional or direct response, like a webinar registration reminder or a whitepaper download for lead generation. In Social Magnet, you can categorize any message – email, Twitter, Facebook and soon LinkedIn. It works very similar to a blog, where you determine what content categories you would like to group messages within, and then simply select one when you publish your message. We currently have a number of informational message categories, including Blogs (where we measure how much lift our own blogs get from our social channels), Articles (similar to Blogs, but measuring the clicks to articles we publish or are quoted in such as MediaPost and Mobile Marketer), Images (used to measure the pickup of any kind of images we post, principally socially) and Pass-Along (where we group any social messages that pass along links to articles that have nothing to do with us, but which we find interesting or influential enough to share). We have another category for Webinars on the promotional side, and one more for Newsletters. Once we have enough data, we might have some insight about which channels are best for different message types. It could be some highly actionable learning, as marketers could then fine-tune their messaging by content, improving targeting by channel while at the same time reducing clutter by limiting messages to the channel they perform best in.
You’ll notice I haven’t once said “best practices” in this whole piece. My hypothesis does not allow for them. I don’t expect to find any universal truths about combining email and social, and fully expect that what I learn is only partly applicable to other marketers. Instead, what I hope to learn – and share – is the process by which we’re discovering what works for us. The findings may not be replicable from one company to the next, but learning what you can learn from email and social is something that we can all do.