Not Sending is Scary!

Since the beginning of time (well, since the beginning of email service providers), everyone just wants to send emails. A lot of emails. Too many emails. nonstopemailspleasestop.

One of our cofounders, Aaron, ran a brand, Modify Watches, from 2010-2019. He loved Mailchimp. He loved it so much that he would send multiple emails per week to his entire audience. (He also loved the Mailchimp monkey t-shirt he got as a customer). As a brand owner trying to build his business, Aaron would ask:

"What’s the problem with sending another email? It costs $0.001. And I’ve got awesome stuff. Of course they’re going to buy!"

Well, over 85% of the emails sent globally are spam. And because of that, Gmail, Yahoo, Outlook and every other inbox provider has teams of data scientists and machine learning engineers protecting our inboxes from spammers, phishing attacks, and stranded Nigerian Princes.

Orita is a team of machine learning engineers and commerce fanatics. We build custom models for every single client, in order to help them identify which customers are engaged, and when. Why? Because no one - not even your most loyal customer - is always engaged with a brand. You might want to hear from a business one day, and not again for six months.

At Orita, we talk a lot about deliverability because it’s important. But our mission is not “to improve deliverability.” Our mission is to help all of our clients achieve profitable growth, so that they can thrive for the long run.

And we’ve built our model to reflect that. Orita is a weighing machine. We take into consideration the risk of negative events - someone clicking spam, clicking unsubscribe, or even just not clicking - and balance them against the likelihood of positive events, namely if a customer is likely to click, learn, and hopefully purchase.

What’s scary about that, we ask, self-referencing the title we chose for this-here article?

We’ll tell you what’s scary. On a regular basis Orita will temporarily suppress thousands, tens-of-thousands, and sometimes hundreds-of-thousands of email profiles. And remember, brands loooooooooove sending that one incremental email. Just like your author looooooooooooves adding extra letters to words.

But, hold on - depending on the week, we’ll sometimes reactivate thousands, tens-of-thousands, and hundreds-of-thousands of emails. You know, likely during your big seasons and, if you’re like most brands, right before Black Friday / Cyber Monday.

Why the big shift? Well … people aren’t always engaged! Brands have peak seasons, shoppers have buying cycles, and depending on the product and market, that can cause huge swings in customer behavior. We analyze not just their behavior, but also the seasonality of the brand, the engagement window of past customers at the brand, etc. Again, we’re building a custom model for every single brand, fine-tuned on their proprietary data.

Back to improving email deliverability. It’s an output of what we do, but not the output. Sometimes you actually want to risk letting your deliverability sink a bit, when there’s a high-enough likelihood of the positive event of a sale to outweigh the risk of no-click; someone clicking unsubscribe; or someone clicking spam.

Just like sometimes you want to send 100,000 fewer emails because the risk is high, and those 100,000 emails are 100,000 opportunities you might have created for your loyal customers - customers that would have bought from you in the future - to unsubscribe from your list altogether!

During times when we suppress a ton of contacts, it’s because our model is saying that for those 1,000, 10,000 or 100,000 folks you would be taking a risk in emailing them that is not offset by the reward of a likelihood of them buying on account of receiving an email. Not that they won't buy, of course, but that email will not be the causal event that leads to a purchase.

A few times brands have asked us to just send to a much bigger list. And each time we end up receiving an email like this after they see a seriously depressed click rate and drop in deliverability:

Hi,

I had a chance to review performance and while the Orita segment did drive some additional revenue, the engagement wasn't strong enough to include it foing forward.

Can you go ahead and have team turn profiles back of when they have a moment?

One final thought: Not sending an email doesn’t mean that person won't buy. Think about yourself as a consumer … you don’t only buy when you receive an email, right? You buy because it’s your season, or you saw a friend wear something, or you saw an ad, or an influencer’s post, or just because you’re the boss and you can do what you want.

Email is a tool - a crucial, impactful, wonderful tool! - but it’s just a tool. Use it wisely, and don’t waste it. When you spend that $0.001 that feels so good, you might just be losing a future customer.